"Before starting the IPO process, we spoke to many business leaders who had gone through a similar process in the past. One piece of advice, which we heard in slightly different forms from multiple people, stuck with us: never let your company be worth a dollar more than it is actually worth. What they meant is that it is very dangerous when people buy your company's stock based on expectations you cannot fulfill or on incorrect predictions about how you will behave. This is in fact very similar to one of Anthropic's guiding principles: underpromise and overdeliver."
(narratology, fandom slang) To intentionally call attention to the improbable, incongruent, or clichéd nature of an element or situation featured in a work of fiction within the work itself.
Their AI is only dangerous in that it doesn't have a moat, and they're using it as a weapon to coerce government into giving them one. "Oops, we "accidentally" left a hole in the sandbox and when it was told to do things it did them on the internet."
They don't deserve a moat; they deserve criminal prosecution for these flagrant acts. Meanwhile, the rest of the world can go about treating software in the way it always has.
And it also seems to me that prosecuting the people who warned about this 6 months ago for unspecified "flagrant acts" will probably not fix the problem.
This really reminds me of that scene from Wolf of Wall Street where Leo's selling penny stocks out of a strip mall office. "I never ask people to judge me on my successes, I only ask them to judge me on my failures, because I have so few" or something like that.
I think it's probably fine in anthropic's case but I flinch at the "underpromise and overdeliver" line. Heard way too many people say that and what they really mean is "underpromise and deliver."
To be fair, I spent unbelievable amounts of money on Opus tokens for OpenClaw. Easily my biggest expense this year.
The risk is that other models are pretty comparable to Opus since around July.
Unless they have cheaper electricity costs, I can switch models interchangeably.
They don't have a moat.
Gosh I wish it was easier to short stocks. Whatever platform I used made me sell/buy 100, and I wanted to do this short 3 years long, and buy every 6 months.
We know Tesla and SpaceX are going down at some point. But "The Market can remain irrational longer than you can remain solvant."
And bear ETFs are not the same as shorting. I learned this the hard way with $80.
To be fair, this account is 4 hours old and created 2 hours before the S-1 post.
With the amount of money involved here and the current agentic tooling context, how am I ever to take comments like from accounts like this on topics like this at face value?
I'd like to think I don't just take hype as a reason to buy a stock.
I went to a birthday party for a kid with first generation Indian Americans that thought they could make money off SpaceX. It was extremely showoffy, they had their cleaning lady carrying around drinks like a butler lol
I warned them it was overvalued.
They thought they could make money quick before it collapsed.
> We know Tesla and SpaceX are going down at some point
How do we "know" that about SpaceX? Still seems to me like they're basically maintaining their decade lead on everyone else on rocketry. And remaining competitive in their other endeavors as well. To me it doesn't seem at all unreasonable for SpaceX to never be meaningfully cheaper than it is today. But I'm not exactly following SpaceX closely. What am I missing?
If the rocket company was in the stock market as its own thing, not as an extremely complicated bundle, it would not be priced anywhere as high as it is now. I would expect that company to be healthy for quite a while too.
The actual difficulty we have with the modern tech company, which is at its worst with Tesla and SpaceX, is that we are really seeing a price or large conglomerates, so they are extremely hard to price. If they were all split into individual companies, how much of the market cap would go to each? I find the bundles to be rather anti-investor.
And no, it's not just those, and not really just tech companies, but that's where the problem is at its worst, as that's where expensive investments some of us consider worthless are tied with business worth something. Take Amazon. You are buying AWS, a logistics company, an ad company, and as part of this, there was always a devices division selling Alexa. I found Alexa to be a real loser from very early, but there was no way for investors to say what they thought the division was worth. The Amazon es execs could at any time decide to invest billions, or invest zero, and given the size of the conglomerate, the price change was going to not be all that high.
SpaceX has many components I would price near zero, where a lot of money is being spent. But as a pile of risky bets, the fact that they are a pile means it's harder to bet against any one, or ten of them. And if it merges with Tesla, that makes it worse. The difficulty in pricing the bundle makes it really easy to hide big mistakes for a while, until they are so big it all crashes. I don't think that is good for investors, or for having accurate pricing, which is one of the goals of a stock market.
Biggest issue Tesla, SpaceX, Anthropic and all these companies have is you can't inflate your value beyond the total market. It's the same thing as bitcoin pushers trying to imply the value would 100x again pushing beyond the total valuation of the Earth.
These companies have the benefit of a real product at least, but eventually the rubes get fleeced.
You are missing his Elon political bias. Given every analyst on Wall Street is bullish and there last earnings was a huge beat its safe to say OP is a left wing warrior
Could you explain what the joke is? It's very good advice, and should be followed. Also consistent with the beliefs that Anthropic's leadership has demonstrated.
They might be very wrong, but they undoubtedly believe what they wrote and what they put in the S1
There has not been a company that has overpromised and underdelivered as strongly as Anthropic in the history of humanity. They literally talk about their chatbots being conscious, model welfare, blah blah blah. Whether or not they believe their marketing is one thing, but it is nonetheless a fact that they market themselves as having created life.
“Beleive” in the sense that Tom Cruise believes in the cosmic DC8?
Anthropic [employees] huffing their own koolaid fumes aside, it’s extremely difficult to believe the claims that these mega valuations are substantiated on (*one* company somehow assuming a meaningful fraction of global research and production)
Look there's largely diversified customer base paying them at rates snd growth that justify the valuations.
Unless "cosmic DC8" (whatever that is) has S1 with data thus string, what possible comparison could there be?
These financials are very very strong. They alleviated many doubts I had about Anthropic's viability (namely I wasn't sure training+experiments could fit in the compute budget and keep them profitable).
We have the numbers right here. Comparisons to Scientology and Elizabeth Holmes are more similar to Scientology than what we see in the S1.
Everything comes back to whether the demand is really there.
Anthropic thinks it is. The got the demand / datacenter match right in Q2, got great margins, and committed hard.
They've got:
- $114b committed to Google
- $110b committed to AWS
- $161b committed to Broadcom
- $120b committed to SpaceX
If the party continues, they'll go from $50b in revenue to $500b and be fine.
But if they don't, it's going to hurt. And that's their only choice, because otherwise, they end up with a frontier model and 50% of the capacity and get eviscerated on either availability or last-minute datacenter leases.
It's a big business, but it's not an easy business.
These commitments aren't annual though, most of them amount to 10 billion a year commitments for 10 years. Put another way Anthropic has committed 52 billion dollar a year to these companies and their S-1 indicates they earn 67.2B a year in revenue.
I'm not going to predict whether Anthropic succeeds or fails, just that it's very misleading to claim that Anthropic needs to grow to $500b in revenue without making it clear that it's $500B in revenue per decade rather than $500B in revenue per year which is what most people reading such a number would assume.
Thanks for clarifying, that’s a huge difference. Obviously they will have other revenue sources, but will likely make a small portion of their revenue. If we approximate their growth rate by these numbers, it is quite a slowdown.
Meanwhile the Chinese are going to undercut them every chance they get.
Today's prices are the ceiling. But AI capability is getting less of a scarce resource with every passing day and it would be a big mistake to plan that your current margins are going to be your future margins as well. Long terms AI will converge on energy costs + capex write-off + margin and the latter will be a race to the bottom.
The problem for me is that it's really hard to project what this industry is going to look like 6 months from now, let alone 6 years.
What's the endgame once the dust settles? Is it going to be a utility like water or electricity and the cheapest offer will prevail, in which case margins will be crushed? Or is Anthropic (or some other big AI lab) going to swallow everything and become the Google Search of LLMs, with everybody else fighting for scraps?
Or is it local compute, models becoming more efficient and hardware becoming cheaper (again) and companies hosting their own AI clusters like they would run SAP?
I think there is a plausible path where this valuation actually works out, but it's a narrow one IMO. I also want to acknowledge that I personally would prefer if it wasn't the case and a more distributed, open model prevailed.
I just look at technology over the last 50 years and see: decreasing prices, increasing capabilities. Barring temporary hiccups. Altman & co did a very smart thing by locking up the GPUs and the RAM, they effectively created a temporary near monopoly. At the same time: that incentivized the rest of the world to look more closely at efficiency negating some of the gains of that expense.
It's just like a utility with the slight difference that it can be sold at a distance without significant cost for transportation, in many ways it is the ultimate digital good. Unlike Spotify copyright is not an issue, unlike your electricity, gas, internet or water company there is no last mile problem. Unlike Google there are a ton of qualified contenders already (Google had to be better than Altavista, which wasn't hard, but they were so much better they managed to lock up search for long enough to cement their position).
Anthropic has none of the advantages and all of the disadvantages. On top of that all AI companies seem to be run by megalomaniacs.
> On top of that all AI companies seem to be run by megalomaniacs.
I think this might genuinely going to become a problem for them in the long run.
People just absolutely hate their guts. That's a very strong motivator to do everything possible to get rid of them.
We'll of course see in 5-10 years if the doom PR will work out, but that always was a rather high-risk gamble, and it already looks like it might not.
There is a significant amount of businesses and people that are very willing to switch to literally anyone else just out of spite alone.
It's not necessarily spite as well, it goes without saying that international relations between the US and its allies have deteriorated to the point yet another critical dependency Washington can turn off for geopolitical reasons isn't a very attractive idea for many. While I dislike a lot of Anthropic's behaviour, it was not their actions but the fact their models were subject to export controls which really motivated me to not be dependent on US providers alone.
My country has no frontier models; since I've no national grounds to support one provider over another, a strong degree of provider diversity is an obvious benefit to me. I also think a highly competitive global market is a good thing for mankind in general, any hegemonic power having a monopolistic or oligopolistic position in AI provision is probably the 'bad ending' for most of us since it represents an enormous concentration of democratically unaccountable power regardless of the power which 'wins'. I'd argue it's rational, not spiteful to favour provider diversity on those grounds.
As a French person myself, I think this antipathy is less strong in non-English speaking countries, where the population (even if they understand English) is less exposed to the antisocial rants of these megalomaniac idiots.
Someone like me who spends a lot of time in the anglosphere puts faces on companies like OpenAI or Anthropic. I doubt that's true for most people, especially outside of the USA. They're just big companies, like 3M, DHL or Toyota. Who knows what the management of these companies think of the future of humankind?
I would expect than personality-driven rejection of American tech from abroad nowadays is driven mainly by Trump and, to a much lesser extent, Musk. Everybody else is just "some guy" to 99% of the population.
I'm not so sure if that assessment is true, but regardless of that, it also doesn't matter.
Decisions in tooling are not made by the 99% of the population, but a subset that is likely a lot more tuned-in to all this.
After all, that is kinda their job so that they can make good decisions
The only thing protecting America's AI industry is the cartel of hardware and embargoes against sharing resources. Once those are cracked, they have zero moat. We're talking 1's and 0's at the end of the day, not nuclear engineering or nuclear fusion or some other revolutionary way to increase the value of real property.
>not nuclear engineering or nuclear fusion or some other revolutionary way to increase the value of real property.
Is engineering the factories that make the hardware not nuclear engineering? Even if not, considering there exists only 1 supply chain in the world capable of making them, it seems pretty revolutionary. And that moat seems pretty big, given that it hasn't been crossed for many years.
The long-term is considerably easier because in the short-term there are limits on supply so prices aren't really able to move in a way where balance will be found.
In the long-term, the marginal cost of a request is zero so prices will plunge and a lot of the fixed investment will get stuck. The difference between the cost of training a model/infra and serving a request from that model is a clear problem for economics.
What happens with Anthropic specifically is less clear because they, as every company in the same position in the last 150 years has done, are going to push for regulation. A lot of the risk has also been shifted onto neoclouds. The value of the customer relationship is, however, effectively zero. A core part of their strategy has been to build stuff out like CC to differentiate that but, ultimately, that won't work.
The other strategy they have employed is trying to get other companies to stop training new models. Again, seems unlikely to work.
There is no scenario in which they become Google search outside of government intervention. Obviously this is a key part of their strategy now but at best, this will only work in the US and Europe (I would actually say they are much further along in Europe where there is a strong tendency towards corporatism and centralisation of power, they could be "Europe's AI").
The only thing regulatory capture in the USA will do is to improve the situation for the rest of the world to compete. I wonder how long it will take them to figure that one out, for all their smarts they seem to be incapable of reading a history book.
> improve the situation for the rest of the world to compete
The risk is that USA and China will come together on a joint "AI non-proliferation" agreement in the same way that nuclear technology became restricted by the countries that already had the capability. This will lead to a de facto techno-colonization of the rest of the world just between themselves.
This was a topic of discussion at Trump and Xi's AI summit just last month.
Nobody cares about hobby models running on a 5 year old GPU cluster in the basement. This is about industrial-scale AI which, to date, only USA and China have accomplished.
Although I agree inference will become cheap (probably not zero), it's quite possible that there is a moat they can build around understanding customer context.
With each query OpenAI and Anthropic get more context about their users. When I compare answers between AIs the ChatGPT answer often includes useful things it inferred from earlier conversations. Losing that context could be a meaningful barrier for a customer to switch to a cheaper service.
> With each query OpenAI and Anthropic get more context about their users.
Not unless they break their own TOS. The temptation must be very large to do so and is probably why those settings keep mysteriously shifting back to 'yes, I give all my data to you to do with as you please'.
Top models, capable of discovering and inventing stuff, held back from public, and available only as for-rent research service (or whatever the relevant bureaucratic term is) to big corporations and government agencies like NASA, DARPA, NSF, NIST, DOE, etc. only.
I have a feeling that we, the public, will stop getting access to more intelligent models starting next year.. i mean, not the incremental ones, but, major ones.. like the one you got going from opus 5 to opus 5.5.
this suffers from the same prisoner's dilemma we already see at play. Any player can defect and in theory capture a "trillion dollar market". Once they IPO, then their service is to maximize (short term) shareholder value. Saying "we reject all the billions in recurring subscription revenue because we are hoping all AI providers will do the same, therefore creating a moat" isn't going to go over well.
There is a good chance of that. This will incentivize the rest of the world. Keep in mind that the world is much larger than the USA, that the USA has alienated a lot of parties that were its allies and friends not that long ago and that there are far more talented people outside of the USA than in it. Silicon Valley had the world by the short and curlies and then Trump came along and pretty much burned it all down. It will take another decade to know how big the damage is but it is probably in the trillions of $ of lost income and that's before we get into intangibles.
I don't want to make this discussion politicized.. but man... Trump truly burned everything down that the US built over decades of hard work and careful planning. And these things are extremely difficult to build back, especially in a world like today's.
I agree with you but as they say "markets have short memories". After the Ukraine war is over let's see how fast Europe goes back to Russian gas. They're making noises about it now (and parts never stopped).
The world is going to be very different in another two years.
yes, you are no longer trusted. Europe is lucky that the UK and France essentially told the US to fuck off when they tried to stop them developing their own nuclear deterrent. If they had, then right now the alliance between Russia and the US would put the whole of europe at risk of Putins Russia. As it is the French and UK capability is more than enough to counter the russian nuclear threat. As in we all fucking die if russia launches nukes, but it at leats lets russia know that too.
fuck trump and republicans. and yes whatever goodwill you had built up over the last 60 years was up and away like a fart in the wind when you elected a corrupt pedophile for the second time. And it cant be built back, ever. No one trusts you no to pull the rug if we became reliant on you again. What we will have though is a relationship with allies built on equality rather than one main player calling the shots.
I haven't been back since Trump I. Missed a funeral of a friend. I still have more friends on the other side of the Atlantic than I have in Europe, even so I don't think I will ever go there again. Enough is enough.
The evidence we have so far points to LLMs becoming commodities. I can’t think of a technology in the last 30 years where so many companies are within ankle biting distance of the leader. There seems to be no data aggregation moat like with Google.
Distilling just works too well. It costs much more to aggregate the data in the first place and train the first model, but once that exists and you are selling access, the second guy can train their model by using the first model to generate the training data.
The only ways to fight this is trying to either:
1. Block it by cutting off access, which is pretty tough because it basically means you cannot offer unmonitored access which is a no-no to a lot of customers, and even then the workload looks very similar to actual real workloads and if the company doing the distilling is a bit sneaky and using intermediaries etc it's almost impossible to stop, or:
2. Try to use IP law to block the practice, which is pretty gross given how they got the data in the first place, and so far no court has supported them on this. And even if they can make it illegal, it won't actually help because there are countries that see having top-tier AI as a national priority and will actually support their companies breaking laws to get there.
People misunderstand why Anthropic is being so stingy about granting access to their top models. It's not some kind of marketing ploy, they know that if they release them fully openly then there will be a Chinese open-weights model in 3 months that is just as good. Also the reason why they try to hide chain-of-thought and the probability distributions.
Given Anthropic is basically just “distilling” the IP of others they’re in no position to cry foul when others run the same playbook on Anthropic IP. They can pound sand on that front.
Indeed, the hypocrisy is cringe worthy but the lack of proof makes it a transparent play to stop giving their users the ability to audit the token counts. "Here is your bill. You spent 20% more than last month, but we're not going to show you what you spent it on."
It's very possible given the current environment that LLMs go the way of airlines. Low margin businesses that are essential to world economics with heavy price competition.
And this is why I think companies will wise up and stop sending their information to the likes of Anthropic and OpenAI: they will not be able to maintain their edge if either one of them decides to defect and start using all of their uploaded user data to improve the model. So they have a massive incentive to break their own TOS.
That's the push factor. The pull is, say there's a giant repo like windows or Adobe. LLMs produce spaghetti code because the repo doesnt fit in the context and it wasn't in the training set. If Adobe fine tuned it the LLM would be much more useful.
It wasn't a theoretical thing, it was a practical thing. Just like when we talk about 'infinite tape' when looking at Turing machines. As opposed to 'clear boundaries to how much context you can have because that is baked into the model'.
When google became the king of search the incumbents were so far behind that catching up wasnt possible and they thought their position was entrenched anyway. With LLMs openAI's competitors are at or close to their capability already and everyone knows they need to keep spending to remain relevant. So unless the cash pipeline dries up for competitors that is unlikely to happen. The reason for that is anyone who has invested knows that either they give up and hand everything to their competitors or someone else does. Which means you need to keep spending to outlast your competitors. These companies are too rich and there is too much VC money around and enough billionaires with more money than they and their descendants could spent in many lifetimes, and a huge amount of greed. If the money runs out that is because some very bad shit has occurred and LLMs will be the leats of our worries.
Given the trillion-dollar spend, the theory is that every many woman and child in the western world will devote a substantial portion of their income (like 5-10%) towards some sort of AI subscription.
or...
That doesn't happen. All the AI investments go south and every man woman and child in the western world will devote a substantial portion of their income (like 5-10%) to bailing out an economic system that went all-in on AI. Either way, the money has already been spent.
I don't see a path forward for valuations like Anthropic, either the hardware requirements fall off a cliff (and why would people pay Anthropic so much if it runs on a cheap laptop?) or either they don't and they will never have the capabilities to serve all their customers (and why would they be valued so much then?).
It's a lose/lose situation here for Anthropic (and OpenAI) whatever happens.
> Or is Anthropic (or some other big AI lab) going to swallow everything and become the Google Search of LLMs, with everybody else fighting for scraps?
Of all scenarios, based on the data we have now, this seems to be the least likely one.
The Chinese labs are not providing their open-weight models for free because they think it is good for humanity. They are doing it because they are severely chip constrained due to the Nvidia and ASML embargos and all available GPUs have to go toward training with very little left for inference [1]. So they are giving their models away for US and EU hosts to serve inference in order to still have a presence in the market.
It's not a great position to be in and it will still take a long time for the compute gap to close. In the meantime, Anthropic, OpenAI, and Google are essentially unchallenged.
You should. Maybe they have it open now just to destroy competition and then they close it. Now you have a closed system, with no alternatives.
Besides...
"If passed, the rules would require Chinese companies to obtain licenses to export the core technology behind their leading models.
They would also ban the export of A.I. that can be used to launch cyberattacks, as well as models trained on what China calls “core data.” The government broadly defines that as data that concerns national security, the welfare of Chinese citizens and the public interest. Experts say the scope includes data on the military or critical infrastructure, state secrets and other sensitive information. "
> You should. Maybe they have it open now just to destroy competition and then they close it. Now you have a closed system, with no alternatives.
What makes you think there would be no alternatives, though?
Put it like this: Kimi K3.1 is downloadable and can be run locally. (If you have a datacenter.) This is a very capable, ~2.8T parameter model.
Now let's say that every Chinese firm goes closed in six months or a year and jacks up pricing... What's stopping a bunch of enterprising upstarts from updating all 2.8 trillion parameters, extending them, distilling the model into very competent flash models, or applying targeted RL?
Once the principles are understood and there are models that can be pulled off the open internet, it doesn't seem that hard to spin up your own LLM. You just need some financial resources and a reason for doing it. And that would come into play if the Chinese kill every closed American provider and then go closed themselves.
Training is really really expensive and open models are not getting funded by VCs. Distillation also doesn’t improve performance. Until training gets exponentially cheaper so that you can do it for a few hundred grand to train a competitive model, open models are labs trying to flood the market to get market share / undercut more successful labs, not a sustainable strategy that will extend indefinitely necessarily. It’s like claiming Uber and DoorDash are fantastic because their prices are so low - once they eat the market the prices become higher than ever.
> What makes you think there would be no alternatives, though?
There could be, they can actually come from open American companies too. Ask yourself why we aren't seeing a thousand or more American AI startups spring into life overnight to build "open" models. Follow the economics.
> What's stopping a bunch of enterprising upstarts from updating all 2.8 trillion parameters, extending them, distilling the model into very competent flash models, or applying targeted RL?
> And that would come into play if the Chinese kill every closed American provider and then go closed themselves.
What's stopping American or European tech companies from doing this today? It's economic and technical capability. I'm very suspicious of any suggestion that it is in any way not incredibly difficult to build and scale up LLMs. Maybe in the future it can be much easier, but in that case you run into a sort of tautology where it becomes easier but the leading companies continue to lead and do so by a significant margin because they're the ones that have the most talent and access to capital.
The premise seemed to be "an open China" versus a "closed USA". I'm not sure that's a great premise. If China can create open-weight models so can the US. If the US can close their models (as clearly evidenced by the article) so can the Chinese.
They're already useful. People are shipping real products with them.
Also, it's entirely plausible that an previously open model (that then became closed) will be much easier to distill than having a bunch of closed models.
They'll be just as useful as they are today. Which is plenty useful. Just like my 10 year old laptop is still useful. I could replace it, but why would I, it serves me well.
I mean how useful are GPT-3 weights today? That's the kind of useful they'll be relative to the frontier.
I feel like the so-called "Deepseek moment" broke too many people's brains and they've lost the ability to reason about why frontier advantages matter. I think we have to learn some new core concepts and rate of improvement over time is going to be a big one, and not understanding that causes people to exaggerate the significance of downstream distillation that's done in response to it. It's cheaper and open and that matters, but it's not the kind of mattering that replaces the relevance or advantage of the frontier, and I feel like I'm not seeing people demonstrate that they know how to really compare the relative merit of these two things so much as zoom in on one to talk past the other.
I just disagree, I don't think your 10-year-old laptop is very useful either.
It's like sure yea you can access the Internet and tinker around or whatever, and with some old LLM model you can have it do fun things, but on a scale of absolute usefulness and if we include what people actually pay money for and what is highly adopted, your laptop and model are severely outdated and by and large irrelevant. That's why the iPhone sells like hotcakes all day every day and twice on Sunday and Apple is a $5 trillion company. Society disagrees with your measure of usefulness.
> I don't think your 10-year-old laptop is very useful either.
Technology never was about the tech an sich but about what you do with the tech. For one person that 10 year old laptop is useless, for another it is a source of $500K / year in income. You, the user are the biggest factor in how much value you can extract from the technology at your command. I'm not sure how much longer that will be true but it definitely is still true today.
The laptops generating income are the relevant metric here; most software developers would struggle with a 4GB Pentium these days. It matters to your company and your economy if your engineers can work.
Compared to laptop quality, I suspect model quality will be a much bigger competitive advantage in the tournament of the marketplace. High-quality laptops are still limited by the quality of the software developer.
When you're not in the lead, your best strategic move is to neutralize trade secrets advantages of the leaders. I think a strategic approach is being misinterpreted as a demonstration of virtuous ethos, which is crazy given (as the other commenter mentioned) the norm of secrecy when it strategically suitable in basically every other context.
China bans all sorts of things, so this doesn't really matter. I actually view it as a good thing. Americans get the best access to the best technology and we can go on to use it to maintain or gain a leading edge in the progress of our society and businesses. Why would we want to give that away for free so people from countries like Russia or China can build competing companies or use these tools to undermine our society's aspirations?
There are dozens of Chinese domestic AI accelerators: Huawei, Alibaba, and Baidu all have chips, there are a bunch of smaller startups too. The chinese mainland is behind on leading edge fabrication though, SMIC only does ~7nm class since they don't have EUV.
This might just be the thing that tips the scales in China deciding invasion of Taiwan is worth the risk. They're eying the Russian war warily though, if they can keep the United States distracted long enough and the EU at bay by continuing to bankroll Putin then they just might risk it. The window for that is closing though.
The online crowd harping on about Chinese models is disconnected from reality imo.
I won't use a Chinese model. Why would i bother? I pay $100/m for the frontier model, understanding it's marginally better than much cheaper alternatives. I get so much value from the product it's not worth shopping around
Think of it like apple and android. Plenty of people pay 2k for their iPhone every few years when android at 1/4 the price is just as functional. Why? Because they get so much excess value from the device it's not worth shopping around. For most people it's their only personal computer. From booking airfare, to filling out their kids permission slips, it's a device they use 10+ hours a week. They don't bother, until a paradigm shift happens and something is 10x better than the iPhone. I imagine llms will follow the same trajectory
First off, $2k is the price of the new iPhone Duo. Suggesting that that is basically the same thing as a $500 Android phone is... strange.
But also, I don't think you should use individual consumer behavior as a proxy for how the overall LLM market behaves. Current estimates are that individual users represent a pretty small share of overall genai revenue. Most of it comes from enterprise contracts. Those might be more price sensitive because the CFO gets a say in which vendor they choose.
For a couple years now I've been accustomed to choice of foundation model being a heavily scrutinized design decision because it's just about the easiest way to manage COGS. I can certainly imagine scenarios where, given the option to pay 1/2 as much in inference costs but with a 5% reduction in output quality, the boss would say, "Nah, that's a bad deal." But I have yet to encounter one in the wild.
> First off, $2k is the price of the new iPhone Duo. Suggesting that that is basically the same thing as a $500 Android phone is... strange.
Okay fine, it's $1,400 for a latest iPhone Pro with 512gb of storage. Point still stands.
> Most of it comes from enterprise contracts. Those might be more price sensitive because the CFO gets a say in which vendor they choose.
I think enterprise contracts are even less sensitive to price than consumer. That's the whole idea behind enterprise pricing. You'll see companies charge multiple times as much just for SSO for instance.
There are limits to this obviously, but to think that major commercial enterprises will switch over to some Chinese models to save some money is laughable.
> Okay fine, it's $1,400 for a latest iPhone Pro with 512gb of storage. Point still stands.
So you switched away from the specialty product, but you're still comparing a top of the line 6.3 inch iPhone to a mid-market Android phone. Why not compare it to a Pixel 11 Pro with 512GB. That way they're at least comparable devices. But also, that would cost $1,220, which I suppose doesn't go well with the intent of suggesting iPhones are several times as expensive as Android phones.
The reason why I'm digging in on this is, when you start your argument with such a transparently silly comparison, it gives the impression that you're dealing in superlatives and willing to say just any old thing. Which undermines your own credibility.
Yeah. People have very strange ideas about competition and choice and moat. If all you needed to unseat the entrenched was a competitor then Bing would be going toe to toe with Google, and Edge with Chrome. They have a billion+ devices they can stick those options as defaults and still nothing. There are many examples.
People are naturally sticky. If they're satisfied with a product, they'll stick around unless something comes along that is so much better it gets people talking. Open AI's 1B+ customers won't give a shit about Chinese models. Enterprise customers won't move to Chinese models, especially just to save money. They are a non issue, especially if the M.O is just going to be playing catch-up.
I would happily bet that > 50% are already doing just that. Whether you'll be able to collect the data is another thing but they'd be fairly stupid if they haven't at least checked out what is available. It would be next to irresponsible not to do so and given that they're in the S&P 500 you would have to at least give them some credit.
At $100 / month (or $1200 / year) it may not be worth it, for you. But many of us would be spending a (large) multiple of that and then thinking about how much you spend, where you spend it and who you spend it on can make a big difference. For companies there are other considerations at play as well, for instance trust.
You aren't paying API prices. Anthropic (or OpenAI, SpaceX, etc) are almost certainly losing money on your subscription. How long they'll be able to continue to subsidize your usage is an open question, but they definitely won't be able to do it with everyone (currently, their solution is to make enterprise users pay per token, but that might change). What that means is that for the customers that the frontier labs need to be profitable, going with a closed weight provider isn't "marginally better intelligence for marginally higher (or even lower) prices" but "marginally better intelligence for much higher prices".
I get (sometimes) frustrated with people that don't work in the tech industry over this concept, and having them rationalize that their usage is heavily subsidized by the distributor of the model.
The costs in compute, electricity, cooling, etc. per token is much higher than most people (I feel) who work in the industry realize. Which, cognitively makes sense: they just want the 'magic 8-ball' to give them a tasty dinner recipe that meets their specifications, for example, and don't care how the 'sausage' is made.
However, economies generally equalize with supply and demand, and this (most likely) will become cheaper as the technology progresses and optimization passes/new algorithms are discovered.
There's been a headline circlulating lately: 80% of Anthropic's revenue comes from 1% of their customers. The $100/mo or however much individuals are paying doesn't seem to be material.
But if you’re a heavy user you’re probably paying $100 for something that costs $1,000. Hence Anthropic’s problem. If they charged real prices to the masses the massses would flee.
The LLMs don't have the moat that Apple has. Apple has carefully built their walled garden to keep people in. Their messaging, their cloud services, their media store, it all keeps users in.
LLMs don't have that. For at least 90% of users, the model in the back end is completely irrelevant to them. Switching to Kimi won't lose them their iTunes music.
I spend extra on a smartphone because I use the thing all the time, see it all the time, hold it in my hand all the time, it's my camera, music player, note taker and so on.
For AI? Prompt comes in, tokens come out. The value of the 'ecosystem' is roughly zero to me. Switching costs between models are essentially zero. The UI and UX is the same whichever LLM runs underneath.
If you do use llm some, then the $20 tier makes sense, but that's not where the money is. If you use it a lot then a $100 sub probably makes sense. But if you're a BigCo that heavily uses it and pay API prices and suddenly you pay $200 million a month, well, maybe it's worth buying a bunch of hardware and stick it in a basement somewhere, and you won't upload all your secrets to your future competitor to boot. If US labs didn't care about the competition from Chinese models they wouldn't lobby so hard for banning the evil, unaligned, super dangerous communist models that want to steal our freedoms and make our gas expensive.
Ultra-cheap models like Deepseek are also very nice if you need to burn billions and billions of tokens.
When the chinese or any other party releases a model that is Opus 5.5 level, there might be a shift to other providers as especially people in the EU are trying to get rid of their dependency on american stuff.
You won’t use a Chinese model, and I won’t either. But both of our bosses have much higher incentives than we do to routinely shop around, and unlike us they can convince an Anthropic sales rep to give them a discount.
That's a great way of putting it and something I have come to realize as well. I'd rather just pay for the frontier model if I trust that it is indeed at the frontier and works reliably. Same with the iphone.
> Why would i bother? I pay $100/m for the frontier model, understanding it's marginally better than much cheaper alternatives. I get so much value from the product it's not worth shopping around
Wait until you are forced onto an enterprise plan...
It's really marginally better and in some cases actually worse. If 100 dollars is throwaway money for you, and prices don't increase sure why switch. For heavier users, and for people with less money(most of America is struggling with bills and food) it's not attractive at all.
"I am spending a provably untenable amount for a services that will have to degrade to meet my price point, the cheaper options and processes are beneath me"
Anthropic doesn't care about sub users. At all. The real target is pay pig enterprises that give their employees access to the API and allows them to go buck wild spending $1000 on fable in a day.
Wish there were more case studies on what people were actually doing. Building out an entire factory sounds amazing. Are they actually getting that level of value out of AI?
Sounds like this will be another one of those things where everyone makes their money and it then will eventually become something experienced people 'blog' about.
> Wish there were more case studies on what people were actually doing. Building out an entire factory sounds amazing. Are they actually getting that level of value out of AI?
I don't doubt it.
> Sounds like this will be another one of those things where everyone makes their money and it then will eventually become something experienced people 'blog' about.
Or not. The problem is that right now you and a whole bunch of others are trying to outrun a pack of lions. The lions have already taken down a couple of your group. The people further ahead have seen what the lions can do and they are attempting to speed up to the point that they can outrun them. Whether that will work or not depends to a large degree on what the other people are doing and they are not going to make it easier for them to keep pace.
That partially explains the lack of deep insight blog posts, besides everybody is just too bloody busy.
Don’t know why this was downvoted but basically correct. OpenAI is in a price war with Anthropic right now for enterprise api consumers and that’s the biggest revenue line item for both.
Are tdoay's prices actually the ceiling? OpenAI added a $500 tier pretty recently while lowering capacity for their $200 tier. I wouldn't be surprised to see Anthropic doing something similar soon, particularly once they are a publicly traded company beholden to making shareholders happy.
I think you’re right - it’ll be interesting to see how the behemoths navigate when new and cheaper memory manufacturing capacity comes online. Currently predicted to happen starting 2028. That is where most of the current hardware costs are concentrated.
On that front, I think DeepSeek is a dark horse. They are really odd - not chasing revenue, care more about unit profitability, are obsessed with driving down costs (like Elon back in the day). As the founder says, they are content with where they are, being the model that does 80% for a fraction of the cost.
I mention them because they are playing a different game and might be able to out-innovate everyone else.
Deepseek has been a game changer for me. I don't use them right now but the model is very capable.
What a lot of people don't seem to realize is how much of the edge of the 'frontier' is in the tooling, the prompts and the harness. The LLMs themselves are already all reasonably close if you keep all the other stuff constant. Sure, one will have an edge in one domain or another. But that edge mostly translates to a degree of performance rather than some absolute. And if you allow for a few more toolcalls, a web search or a longer context then that difference is much smaller than it might seem at first glance.
Anthropic is not in the model business. They were smart and they are building productivity ecosystem business. Look at what claude artifacts are doing now -- inline multi-user editing, comments that users and/or models can respond and take action on, import/export from all common formats. It's at least a Google Docs killer if not an M365 killer.
And claude code as a harness and skill/plugin/mcp ecosystem is far more valuable than any one model.
And coward for long-running tasks, though they're slipping on that one because they're really trying hard not to make Claude an assistant. We'll see if they react to Muse and change course.
But the Chinese free-weight models? Not a competitor. If/when those companies (or others) build vertically integrated productivity stacks on those free models, then we'll talk. Microsoft certainly saw that opportunity but couldn't get out of their own way and are basically roadkill at this point.
>Anthropic is not in the model business. They were smart and they are building productivity ecosystem business... It's at least a Google Docs killer if not an M365 killer.
Google Workspace revenues are something like $25bn/q, and that's the entire workspace division, (including enterprise gmail sales, so the productivity suite is meaningfully smaller than that figure). It's taken Google a long time to carve that slice out of Microsoft's entrenched market position. It's not small money, but they'd basically be going to war with two of the largest and most profitable companies in history on their turf.
I don't think they can credibly take that business plan to the market and say it'll be worth $1.75tn. They're all-or-nothing on being in the model business, and everything else is superfluous.
I mean, a lot of people before them were "not in the model business" (Cursor et al) and it seems they got disintermediated by the model companies. Step 1 is still to have the best model, or hope for model progress to taper out.
Even outside of the Chinese models, Amazon, Google, Microsoft, Meta all of them are offering significantly more coherent and competitive internal models and reasonable price.
What's it like co-editing a combo document/workflow with multiple coworkers and an AI agent on Amazon, Google, Microsoft, or Meta?
Like if you want to design a survey you can send out, and collaboratively edit and test the survey, ask for changes in layout, wording, etc, and design the statistical analysis to be used, before sending the survey out where end users can fill it out right in the doc, while internal users/admins can ask the agent about its results?
Honest question -- that all works pretty well in claude. Amazon, Google, Microsoft, Meta all pretty similar?
That depends on whether "frontier" models can maintain an expanding frontier. If the frontier settles, they're dead, and it will be commoditized. Otherwise, there are still plenty of pathological behaviors of these things to address that will keep expanding that frontier (if that's possible and the limitation isn't the approach itself).
The reception of saying things like this is interesting. AI stans think I'm way too pessimistic but AI haters think I'm mindlessly optimistic.
Cloud compute has managed to have pretty good margins. That seems like a really similar business model to what Anthropic is doing - inferene+value-adds as-a-service. Why wouldn't they also be able to get cloud level margins or a bit higher?
Lots of people seem to have pretty short memories (or are too young to have seen it), but the flight from "local service" to "cloud service" was intense.
Companies don't mind paying extra to have a another company that does nothing but said thing do said thing for them. It even actually often ends up being cheaper/more profitable.
I don't think AI will be able to sustain cloud compute-size margins. Inference is a commodity and value-adds can't be sold for anything. Why pay for an advanced harness when you can just ask the model to build one for you?
Observing SaaS vs using an open source stack in the beforetimes, the answer might be some combination of, "So I never have to think about it, and if anything goes wrong, no one can blame us because we went with the biggest & most trusted provider."
there is no point of saying 'the chinese'
everyone will do it because it was proven that any political frills might cause an attempt to damage your system.
I dont know when, but i think if you keep menacing of taking technology of people hands for any political reason will be a reason for people to not use it or to rely in a more reasonable party.
The point in mentioning the Chinese is that they are doing this today. In Europe there is Mistral but I think they're too far behind the frontier to really matter but in time they may catch up.
I don't think today's prices are the ceiling, considering they are subsidized. Once businesses have grown dependent on AI (and their institutional knowledge is lost), they will pay much more just to stay afloat.
Unfortunately it's now extremely likely that our governments will manipulate markets to keep frontier domestic AI afloat. They are financially and militarily dependent on domestic AI now.
Chinese models are being banned, will continue to be banned by USG. Even the Chinese govt is restricting external access to models now. It starts with banning Govt access to International AI, but will continue to consumer and business access.
> Meanwhile the Chinese are going to undercut them every chance they get.
I wish they were, but I always wonder when people chime in with Chinese models what they are actually doing with those. I have an OpenRouter account and give various models a shot every other month, wishing for an alternative, but I always come back to Claude. Even more so since the switch from Opus 5.1 to 5.5. To me, personally, it's unmatched. Like a year ago it was very different, and I would use other models all the time. But honestly at the time they all felt kind of "meh" still. I just hope everyone is at a good baseline in a year from now on, so I actually have a choice. And then, as you stated, things will get interesting.
I use Opus 5.5 daily at work and a mix of GLM 5.3 (Flash or not), Deepseek v4 (flash or pro) at home and while Opus 5.5 is more capable, in my opinion it matters for less than 5% of the work i do.
These chinese models are simply good enough for the vast majority of tasks I run and when they're not I use Astra or Opus 5.5.
Yes, leaving Opus 5.5 on all the time is comfortable when you pay the 90€ plan but it's just not necessary.
I use Opus 5.5 at work and I use Sonnet 5.5 at home. Sonnet 5.5 medium according to AA cost per task is $0.48 per task, while GLM 5.3 (max) is $2.01 (opus 5.5 high is $1.82). GLM 5.3 flash is $0.28 per task and Deepseek 4.1 flash 1017 is $0.26.
So it seems like Sonnet 5.5 medium, GLM 5.3 flash and Deepseek V4.1 flash 1017 are all in a similar space in capacity and price, however the $20 anthropic sub will get you like $200-$300 in sonnet 5.5...
Another bonus is that sonnet uses a fraction of the overall tokens and finishes much faster than the average chinese model, so your work goes faster (Sonnet 5.5 medium in 2 minutes, Deepseek 1017 in 6.5 minutes, GLM 5.3 flash in 13 minutes...)
I still don't get why paying for chinese inference makes sense. I should try Haiku 5.5 high/max because it's posting higher numbers than Deepseek V4.1 flash for half the price
Yeah, I spent the last week with Deepseek Flash 4.1-1017, Kimi 3, and GLM 5.3 (and 5.3-flash). None of them were literally useless but also none of them were capable of making progress on the main project I am working on where Opus, Sol, and even Gemini Flash have all done very good work. Benchmarks might have you believe open models are only barely behind the big US labs in intelligence, but it feels much worse in practice.
I am running GLM 5.3 Flash locally at home for coding and from my point of view it is indistinguishable from the frontier models of 6 months ago. It can handle almost every task I throw at it.
I have 2 DGX Sparks, 256 GB unified RAM, so not a cheap set of hardware. Luckily I bought them before the recent price increase.
However even with a high tier gaming machine you can run one of the Qwen models and it will be good at coding... like having an intern with boundless energy that will sometimes make mistakes but in general save you a lot of time. And for automating tasks involving scanning documents or creating summaries it will be brilliant.
Comments here are not considering what Anthropic is doing with Claude as a product. The models matter, yes. But they are building Claude into the only tool you need to get business done. In the past few weeks they have added slides and docs - a shot across the bow to Microsoft's dominance. If they can pull off building the best interface to get work done, $500b in revenue starts to seem not too crazy. It's more about selling a complete product than it is about selling tokens.
I somewhat agree to what you said, but also, the whole point of agentic coding is that "code is cheap". So I can't imaging companies like Google not having the resources (whether human or token) to build a better agentic assistant/harness than what Anthropic can. I mean Google has a pretty good starting point with Google Docs, etc.
the market for their valuation is just not there imo. they need to significantly expand into non-coding products and outshine their coding offerings to make it work.
as in, if they capture every single professional software dev and extract $1000 from them every month, their yearly revenue will be like $500B. do your own back of the envelope on that with ongoing costs and consider how unrealistic a goal this actually is.
even worse, the very product they offer is self limiting. if it becomes so good as to actually replace software devs, they will have less customers to extract money from.
their valuation essentially relies on a small percentage of mega users having unheard of levels of spending.
the part about a small percentage giving them most of their money is already true (their rumored $100+ bil revenue as of mid 2026 already outshines all software devs having a $100 max subscription), but they are two orders of magnitude or so away in terms of the actual amounts to make it work.
I agree, I would say $500 billion valuation is fair, they don't have a moat and open models are catching up, they have a limited products with strong competitors, if they have a step change model in the pipeline now would be a good time to release it.
>as in, if they capture every single professional software dev and extract $1000 from them every month, their yearly revenue will be like $500B. do your own back of the envelope on that with ongoing costs and consider how unrealistic a goal this actually is.
The market for Claude Code is much bigger than professional software devs. It's really anyone who uses a computer. I already know normal people who have been using it for personal software.
Not really. Random non tech users won't pay $100/mo. Nor should they, most people will be served fine by whatever free thing MS bundles in Windows next year.
You don't have to skip the plumber or the mechanic many times before $100 becomes a plainly obvious need.
It's not gonna install a shower for you, but it will confidently walk you through that leaky faucet. And I don't know if you have paid a pro recently, but they don't charge $20 for a 20 minute fix.
Indeed. On the other hand, I fixed my underfloor heating with the free claude version, so unless they remove the free tier, I still have no reason to pay for it.
It's quite a leap to assume any kind of issue would require a $100 subscription. Plumbing, mechanics, electricity ... The issues are usually far easier to trace and debug than in software, because the systems have been designed by smarter people.
You're not buying some arcane and complex solution, you're buying the convenience of not watching a couple of youtube videos or articles of someone troubleshooting the exact issue you're having.
To boot even if you skip the free models, you could probably do a year's DIY assistance with 50 cents worth of GLM or Deepseek tokens. And the really costly jobs are ones where an AI assistant won't really help since, shockingly, most of the cost is hours of labor which you would have to do yourself.
Also, most people don't want to build software. They just don't care. There'll be tinkerers like with anything, and I'm sure some people will create solutions to their niche problems with (as you say) 'whatever free thing MS bundles in Windows'. But like the programming trope - they'll build the 'ultimate' todo list and feel like it's an amazing achievement, but be back to pen and paper/notepad/todoist within 6 months.
In most office jobs, at western salaries, you don't need much productivity improvement to offset $100/mo
Though I would expect the majority to do fine with $20/mo. That's the same effective cost as replacing a $500 computer every two years, which is the sort of expense most companies are comfortable doing without much second thought
You don't need to spend $100/mo. They have other price points. And I think you and others are underestimating the appetite out there to be able to create personal custom software. It's empowering. Like I said, I see tons of people I know who can't code who have started doing this (and been very happy with it).
The vast, vast majority of people are happy enough using free chatgpt. Openai could just add a kind of online app development/marketplace thing to chatgpt and that'd entirely steal the personal custom software thing. But I don't really think most people care that much.
>the market for their valuation is just not there imo
I tried to look for the valuation in the linked document, did you see it or any reference to a valuation elsewhere in anything published by Anthropic? what is "their valuation" that you refer to?
These were ~Q1-Q2 commitments based on what Anthropic THOUGHT token maxed compute demand was going to be over next few years. Then Late Q2, and Q3 PRC models dropped floor for token prices. It may very well be that Jevons demand which Anthropic though required 500B of compute now only requires 100B, and unless demand is 5x what they thought, those additional 400B non hardware cancellable commitment are going to be balance sheet concrete shoes.
> they will keep doubling down, because stopping is simply not an option
Of course it is. Apple is taking that strategic approach. If Anthropic believes it’s approaching saturation, pulling the plug when it’s at steady state and its rivals are over-invested is the winning move.
Well, unlike Apple a lot of very aggressive investors in Anthropic are riding on its trillion dollars valuation. They just can't on their own decide to stop or slow down.
Apple has a lot of actual money, Anthropic does not. Apple sells a lot of products, Anthropic doesn't. Apple have a platform multiple players want a piece of, Anthropic fight to stand alone.
There's very little comparison to be had here. Also, Anthropic likely have investors with way too much leverage for them to stop.
I have a hard time believing they are likely to get to $500bn in revenue. That’s another $450bn in revenue growth that needs to happen before demand tapers off.
But I see every reason to believe demand will taper off. It seems it’s already happening. Enterprises have been deciding that maybe corporate chatbot and coding agent budgets shouldn’t actually be infinite. AI startups are starting to successfully outcompete each other, effectively killing off members of Anthropic’s customer pool in the process. And consumer demand for AI products seems to be dropping. None of these are signs that suggest that Anthropic has thus far met only 10% of plausible long-term demand.
Total global wages ~$12 trillion/year. The "what you have to believe" here is that AI tool spend becomes at least 10-20% of it, my belief here is we see net slowdown in hiring as human + AI = super productivity so as businesses grow we'll continue to see increase in AI spending.
Then if inference does grow to a $2-3 trillion/year market that Anthropic is a big winner, as big as NVDA is now (they're going to be over $500 billion in revenue in calendar year 2027).
At this point, though, “human + AI = super productivity” is looking more lime dogma than reality.
The pattern seems to be that it makes people more productive when you measure it in terms of raw work product volume. But not all work product is equally valuable, and we have yet to see a clear pattern of increased economic productivity. Instead we just keep finding new and interesting bits of jaggedness in the frontier.
Of course nobody can prove those economic gains will not manifest at some point in the future. I can’t prove that Jesus isn’t coming soon, either. But I also can’t prove which millenarian denomination, if any, he would choose to bless in the event that he does come back.
> Everything comes back to whether the demand is really there
And unit economics.
HY1 2026 revenue less compute and infrastructure turned positive (45% gross compute margin; $8.9bn C&I to turn $16bn revenue) for the first time; it was -176% in FY 2025 with $7.3bn C&I on $4.6bn revenue.
It looks like a lot of costs of revenue are in sales, marketing and partnerships, though. Backing those out shows H1 2026 margin at 26% ($4.2bn) after -72% for FY 2025. Still positive.
If you owe the bank a million dollars you have a problem. If you owe the bank a billion dollars the bank has a problem.
The same logic applies here.
Google, Amazon, Microsoft, Meta, Broadcom, Nvidia, SpaceX and others are spending hundreds of billions a year each on fast-depreciating infrastructure to serve this potential AI demand. If the demand never materializes then they know they are in big trouble. So they have every incentive in the world to prop up Anthropic (and OpenAI) and keep the party going.
Can they keep the bubble inflated indefinitely? Probably not, but all these companies combined have heck of a lot more influence than Anthropic alone.
They aren't "fast-depreciating infrastructure" -- research shared by a16z team, rental rates are up across the board, even for GPUs that are 5+ years old. The demand is real right now. The proverbial tide could go out but we aren't seeing it yet.
That’s because of current supply shortage, not inherent value. Is there any reason to believe that when fabs can sufficiently scale up production these GPUs will continue to be valued like gold and not something you hold on to for a couple years and then toss in the bin for a newer faster model – aka how computer hardware economics have worked for the last 50 years?
Yeah the fab shortage is really keeping lifetimes long, far longer than in the past.
I've brought my old GTX 1070 out of retirement for for some molecular dynamics these days because its "free" except for eh electricity and have Ben pleasantly surprised at what it can get done chugging along in the background. I'm considering buying some more, as it's kind of a steal in pricing.
The OP post is misleading because it makes you think Anthropic has to earn 500bn per year in revenue, but those commitments listed are not yearly commitments, they're commitments over 10 years with contracts that run until 2036.
Anthropic needs to earn 52B a year in annual revenue to honor those specific commitments plus obviously its own operating expenses... and yeah that will require Anthropic to grow as a company, but their S-1 indicates 67.2B a year in revenue and sure maybe they stop growing and this is their peak... or maybe they do manage to grow further to 80B a year, 90B a year...
are those 67.2B revenue or profit? Because you know there are costs associated with running the business, so of those 67B if they only keep 30, they they're in the red, no?
Yes its hard to predict for sure. Especially as the dust will only settle properly when Frontierlabs reach some kind of slowdown or ceilling, which I do not see right now.
Unless anthropic becomes an entertainment platform like netflix or something, I really can't see how they gain significantly more _paid_ users. The vast majority of people will use the free tiers of openai and claude.
Currently every business is building the Agentic Layer. A layer between the user and the product.
With tx to an LLM, this layer will close the gap to full or borderline full automatisation.
Your Ticketsystem will have agent support out of the box. Your incident system too. Your UI for SEO, SEO optimization, article creation etc. every single on of them wijll be there and consume tokens.
Future invesment from 10 people in your l1 support might be 5 + token budget. Your SEO optimizer collegues? 1 Person + token budget instead of a team.
The SpaceX thing interests me. They are basically paying SpaceX at a $10B per gigawatt equivalent rate right now, yeah? Because they only get 0.3GW for their $3B.
So what that means is that if the demand exists for any AI data centre, it's a solid business to run one — they cost maybe $50-60B to build out, they cost maybe $10B per year to run. You get your money back fast if the demand is there? And you can project being able to replace with better GPUs in the future without having to borrow for all of it.
But it is extremely difficult to believe that Anthropic is not projecting that all of its customers will ultimately use their service to the extent that their outlier customers do right now. They probably won't be projecting that their customers will find the need or ways to use them less.
And also, "no customer is more than 10% of our revenue" means that one customer is. If that customer is Microsoft (can it be? I honestly don't know) then things get tricky, because Microsoft clearly want an unmetered, on-device AI future.
For kicks - lets take the optimistic side here (because the whole thread will be reading pessimistic). What if they are under built and this growth continues to go parabolic. We then still have supply constraints and pent up demand for frontier model and you have a company thats currently at 1-1.5 T and levels up to 4-5 T.
I'm not saying I think thats what is going to happen, but it really isn't a stretch to think that this keeps running given that it hasn't really hit anywhere near as broad as we think it has (being HN community based).
Their entire goal is to get to AGI via RSI ASAP. If they do they win. But if not they’ll be a legendary story of a flying too high too fast.
That they are IPOing tells me that the massive amount of funding they need is no longer available privately and that this is their 1 opportunity to get the funds they think they need to get there.
Biggest risk is the growth stalls as everyone moves to open weight models. They certainly like telling stories.
> If the party continues, they'll go from $50b in revenue to $500b and be fine.
You really have to be insane to think that an AI company will ever reach $500B in revenue. That's almost the same as (2025) Amazon, Walmart, etc., companies with a MUCH wider market.
It is MORE than Apple, it's 5 times more than NVidia even. There is absolutely no software seller that is even remotely close to $500B in revenue.
I’ve spent more money on Claude this year than I’ve given to a lot of other tech companies, combined. Well, I bought a new phone too - but amortised over the phone’s lifetime it’s still much cheaper than I’ve spent on my claude subscription and what I’ve gotten my company to pay in api usage credits.
There are two questions. Will the value keep scaling? And how much of that value chain will anthropic hold on to? Right now software development is being eaten alive by LLMs. Will that happen to other industries too? And in the long run, what do the models we run look like? Are people going to end up self hosing opensource models themselves, or are we going to collectively hitch our wagons to frontier models provided by anthropic and their competitors?
If ai keeps getting better - which I think it will - and if anthropic can corner a large part of the market in an enduring way - which seems less likely to me - then yes. I can easily imagine them becoming a huge company.
"Subsidies" don't stop because it's not "a subsidy". It's a subscription service.
Like Netflix, Comcast, T-Mobile, etc, etc. There are millions of subscription services in the world. They all "subsidize" higher-usage users with plans for lower-usage users using a fixed price plan and rate limit. The profit comes from lots of users who don't hit their cap. The higher priced plans are actually more profitable.
This is called a tiered usage-based subscription business model.
The thing to be remember about those companies you listed, none of them are software companies like Anthropic is.
* Amazon is a logistics company
* Walmart is a logistics and retail company
* Apple is a hardware company (that uses software to set their hardware apart)
* NVidia is a hardware company
The closest you get is Microsoft, at 281B in revenue. Some might argue Google/Alphabet, which is in the same area, but they sell Ads, not software.
Anthropic's expectations is that they will be able to fast-track their revenue generation to almost twice what Microsoft does today. In less than a few years.
> The thing to be remember about those companies you listed, none of them are software companies like Anthropic is.
Yes exactly, that's why I listed them. It's much easier for them to capture a higher revenue compared to software companies, and OP (and Amodei) want to make us believe an AI company could ever reach that revenue.
Let's posit that that's possible, for Anthropic to get to 500B in revenue, they'd either need 100% of the market or the market has to be even bigger.
OpenAI, Google, the many Chinese labs all have products and several of them likely have better financial positions. At the moment Anthropic have one of the best models but it needs to stay sufficiently above competition for them to capture a large share of the market.
At the same time this year we've seen a move for model diversification as cheaper models become "good enough". If this IPO was 6 months ago I think it might have gone better for them.
I think the market has the potential to be 1 or 2 trillion in the US, 10% to 20% of labor costs. I say this because I believe most companies would not blink at spending that money to supply their workforce with AI tools assuming the AI actually did the work.
The winner take all network effects are probably not there in this case, but still if Anthropic is a large player and has 20-30% of the market they could have a revenue in the hundreds of billions.
Double the size of the market to include most of the rest of the world and the 500B is suddenly under 20% market share.
The "good enough" problem is a real biter. The personal PC market overtook the mainframe market long before PCs were even close to beating the top supercomputers. Cheap and good enough is the threat.
What does good enough look like for various use cases, when do we hit it, and can these companies guard that territory?
As I understand it, AI is only really getting traction in IT and software development. It's been mostly irrelevant elsewhere overall, all things considered. That could change, but basically there's only one product that is selling. Dunno whether there is $500b/year of IT and programming to replace.
People are using it to do contract negotiation without a lawyer, argue with their doctors over health matters, complain to teachers about their methods, etc. On the flip side, lawyers are using it in supplement to paralegals, doctors transcribing patient notes with it, and teachers are using it to generate worksheets. It's being used fairly widely already
Maybe, but most people aren't paying for their chatbots at all, yet alone the 100$+ plans. Out of the non-technical acquaintances of mine who do use some form of AI (which is a tiny amount of people), they just use the free tiers on claude and ChatGPT, none of them ever have or ever will pay a single cent for any of this. Even the technical folk are barely paying, and they're certainly not wasting 200 bucks on it.
That's true, I do all of that, but I'm using the free Claude version. It's been more than enough to handle the few home renovation projects I'm currently working on.
That's quite a leap to assume there's a ton of money to be made with the casual users sending perhaps a couple of messages a day to real an email and draft a response.
I recommend listening to the LegalTech Week podcast. Generally lawyers are finding it's not delivering efficiency overall. They spend more time reviewing the output. The courts are drowning in slop and deepfake evidence is a growing concern. The jury is still out on all of this.
My hypothesis is not that the AI is going to eat up that 12T dollars. But there will be additional investment to provide AI tools to the workers to the tune of 10% to 20% of their salaries.
I agree with you from our perspective it is ridiculous. I also agree that getting the consumer market to willingly pay for AI to the point where it sustains such an amount of revenue is not going to happen. (The knock on effects of folks using AI, and turning others off to AI because they used AI instead of writing emails themselves, for example).
But.
They may peak at 500B before this bubble bursts, simply because other major industry players are betting on AI as a means to get folks to believe in their own stock.
It’s belief all the way down, until you get to the point where “the market” starts asking where the return on investment is, and why they haven’t seen it yet. By that time, the investors have gotten their return and cashed out, and it’s the pension funds and the retail traders that are holding the bag.
And of course, at some point AI’s promise will be delivered in a form or fashion as yet unknown to us, but that is still years away, and until folks recognize that LLMs as they exist to day are not the path to that, they’ll continue to believe in this valuation.
What feels ridiculous today might be the normal of tomorrow. Welcome to the new world.
What you are saying is the same thing that has been said about anyone trying to do anything massive and at scale. Can't be done - thats ridiculous - you are out of your mind.
Of course such growth cannot be sustained and the word "MAY" is funny there. You cannot have infinity growth, it is mathematically impossible. There are no unlimited (monetary) resources on this planet.
It's a misleading number, the 500 billion in revenue is needed over a 10 year period, it's not a yearly figure. Anthropic isn't paying 500 billion a year for the commitments that OP posted, it's paying 500 billion between now and 2036.
FWIW I max out my pro max plan basically every week, not even using Fable any more. I think it's just a matter of people figuring out how to apply tokens effectively.
I don't think they are betting on that many people subscribing to their $200/mo plan, there's pretty much no upside in using a frontier model for the average consumer.
But on the flip side I don't think aiming for 1M businesses spending $40k a month in token in average is going to be more viable. And I'm not sure having 1000 corporations spending $40M a month is a option either.
This is the issue, for two reasons: Is there trillions of dollars of demand for LLMs, which is what it takes for Anthropic to make $500 billion in revenue? Is that demand going to be met by frontier cloud LLMs, or by open weight models run by the organizations using them (and cheaper cloud alternatives)?
There are a lot of parallels to the WeWork S1. They’ve made absolutely massive forward cash commitments, on a business that’s increasingly commoditized, and with little hard evidence that they’ll generate the revenue (let alone the cashflow) to pay for all these future liabilities. All while everyone is discovering there isn’t really a secret sauce moat to demand much of a a premium over cheaper options that trade a few points above the direct COGS (cost of compute in this case).
If everything goes just perfect they might eek by. If it doesn’t this goes from high flying startup to insolvency and bankruptcy real fast.
Yes it is a newer business but doesn’t change the fact that models are trending towards a commodity. And that’s a huge problem for the business model on offer.
Right – they can't argue for acceleration factored in their own value, if it's a property of the whole surrounding market.
IMO, US AI providers are not leaders in efficiency anyway, they're still running the original ~2010 playbook that led to LLMs ("What if we took ML researchers' work, but funded it at 50M/yr instead of 250k/yr?").
They're currently arguing for the continuation of that approach: Scale funding first, and the rest will follow. It is high-risk and WeWork-like, by nature.
It's high-risk in that it's proven to not work anymore. Scale isn't a differentiator or won't be in the short term.
Meanwhile, everyone else is working on making these models more useful through a ton of ecosystem around them.
I don't see how the frontier labs will have any control over that to deliver on what investors expect. As mentioned in another post here, if they could, they would already be spawning hundreds of companies to diversify their reach but they aren't, because this era of AI is awesome but still doesn't allow a few individuals to replace millions of others that easily (or efficiently).
Selling subscription productivity tools is a business that has existed for decades if not centuries. Just because their product is new does not mean their business is.
Someone should do the same thing Karen Hao did for OpenAI with the furniture costs of their office, if Anthropic's HQ decoration and office equipment value stops growing then we start worrying
Nobody was saying subleasing offices wasn’t a business. It was just WeWork’s setup that was untenable and ultimately insolvent.
Similarly nobody is saying AI can’t be a successful business, but here too the big labs have dug themselves into a giant hole such that a bankruptcy reset may ultimately be the only way out of it. Time will tell.
was it though? sounds like wework had the last laugh. is bankruptcy part of the strategy? people like you call insolvency the end of the journey. the journey is whatever you want it to be.
The corporate entity called WeWork still exists, sure. If you were a pre-bankruptcy investor or equity holder in WeWork you got wiped out. The management and capital that went into the first iteration is gone.
That seems like pretty good evidence of current and ongoing explosive growth. Much of this revenue is pre-paid (re: cashflow). Meanwhile, operating margins went from -184% to -$4%.
Seems like pretty good evidence. Will growth continue apace? Unknown. But the arguments for Jevon's paradox applying to intelligence seem pretty convincing, so the TAM here seems likely to be enormous.
> All while everyone is discovering there isn’t really a secret sauce moat to demand
Also not convinced by this. The Chinese models seem to have distilled their way to current levels, with Anthropic and OpenAI now getting serious about defending against this.
I don't think you're definitely wrong in all this. Rather, I think your claims are too confident.
Shorting a company is not a trade against the future of a company, it is a trade against the market. You can know a company is going to zero and the company can go to zero and you can still lose your shirt from shorting it.
A lot of people will, if this IPO actually happens. But as is frequently pointed out a short is a market bet not a bet against the company and the market remains wild and irrational. You can be right about the company ultimately failing and still lose your shirt on a short.
However the price of CDS on bonds is already a strong market indication of where the market thinks all of this is heading. And none of that looks good if you’re betting on the AI exuberance continuing.
One thing that bothers me is that if the AI maximalism turns out correct (which is needed to justify the valuations), why would AI companies respect money and ownership?
So far, even major disruptions were still affecting relatively small segments of the economy. Car lobby in Germany, tech in the US, "Big Pharma/Food/etc”. They needed the rest of the economy.
But if some AI company figures out a way to monopolize everything: new companies, old companies, physical manufacturing - all the things they say they will, please buy here for lots of money - why would they be bound to pay out dividends or otherwise share the fruits of their economic labour?
I’m not saying specifically Anthropic would be looking to screw its investors. Rather, if you control an almost omnipotent genie, wouldn’t the temptation just be too much?
Or temptation aside - will money even have any value? If some AI company rebuilds the economy from scratch, and control everything from accounting to shoe making, why would they respect the value of some fiat tokens? It’d be like the conquistadors respecting the shells and beads of their new Aztec subjects.
Of course, perhaps none of this happens, but then the question is, how good an investment all of this even is.
The value of money, and the ability for average people to have influence on future events, ultimately is tied to the fact that the labor of an average person has inherent value.
Once that is no longer true, all bets are off: All promises of compensation become meaningless, and it's easy for those in power to rationalize to themselves why they no longer are beholden to their earlier promises. They will say to themselves "The situation ended up more nuanced than we expected"
(Which is exactly what Google said in February 2025 when they reneged on their pledge not to use their AI for military applications)
Then we regress back to relying on the fact that the ability of an average person (not) to be violent has inherent value. This has been the dominant factor in many moments in history.
Although if you have killer robots, then all bets are truly off, and we go the way of the dodo, literally.
(I don’t actually believe in this LessWrong-esque rhetoric, reality is more complicated, but it’s fun to muse about.)
EDIT: If oligarchs with magic AGI really don’t need our labor, nor are they concerned by our violence, then perhaps we become more like ants rather than dodos: we organize to help each other, creating a somewhat traditional economy around the scraps that don’t matter to the oligarchs, relying on obscurity and harmlessness to avoid getting toyed with a magnifying glass. I suppose that’s slightly better.
> Kings didn’t have robots that gather resources and make more robots, that make killer robots, exponentially.
We still live in a finite, constrained, physical world that has to obey the laws of physics. We have a looong time before we get fully autonomous self replicating decentralised robot armies.
I think people conflate the real world, sci fi and their wet dreams a bit too much.
We're at the "this thing is kinda useful sometimes but still mostly annoying" part, not at the "skynet in 48hrs" part
I suspect any intelligence powerful enough to pull this off would quickly realize that it had no use for the oligarchs, perhaps even more quickly than the oligarchs realize they have no use for us.
The raw resources you need to build robots are constrained and have a very very long and complicated supply chain.
In the second world war, Germany had impressive science and technology, but, as Stalin said, quantity is a quality in itself, and The Third Reich found out that it was an impossible task to get the metals and fuels for their advanced jet engines, rockets and U2 boats, when between then and those resources there were masses of tens of millions of under-equipped and malnourished soviet soldiers fighting what was an existencial war for them.
I certainly feel like AI maximalism is the only way it all maths up. There is a bit that Matt Levine writes about AI companies, and you touch on basically the same points, that roughly goes "you are an AI company and you are making the god-computer. Why would you focus on anything else because if you're successful you could basically generate infinite money? Turns out to make the god-computer you need a lot of money upfront."[1]
So perhaps they would only respect money and property up to the point that it's necessary.
The more realist answer is that they will keep respecting money and property as capital works to reinforce the status quo and enrich shareholders. You think just because you can do your job faster--or completely--with a machine that you have to work less? Hardly! Now this is where interesting policy questions spring up but I digress.
> why would they be bound to pay out dividends or otherwise share the fruits of their economic labour?
because they don't control the physical paper that is money.
I dont understand what you mean by "fiat tokens". I don't imagine most people will accept tokens as a currency because they are as good hold of value and predictable as bitcoin.
In a token economy, certainly what you say is a possibility, but as you can imagine, that economy will be unsustainable so back to "regular" money.
Money is a medium of exchange. If I want something, I can get money in some way, then swap it for something else.
But if I have a near magical genie who can make me anything - software, farming business, physical things - and at a fraction of economic input compared to the legacy monetary policy - then the money is worthless. As the Aztec shells were to the conquistadors.
That's an oversimplification. It comes with a lot of rules and authority that will be enforced by the issuing authority that a simple "IOU" on a sheet of paper lacks. IN the continental US, you cannot refuse to accept a dollar note. You certainly can refuse to accept the very IOU we agreed on a minute earlier.
Most of the developed world already has magical genies - certainly an overwhelming majority of people on HN earn enough to have LLM subs that build code, 3D Printers that print physical things.
The challenge always has been and will be humans.
You just cannot live in a vacuum. If you could subsist entirely on cheetos and figured out how to make an entire lifetime's supply of cheetos, after N days of watching Netflix, arguing on HN, raising chickens and goats, one day you will want to go and talk to a human.
I think what you're missing is if you have a "a near magical genie who can make [you] anything" you don't need to exchange for anything.
And if your "near magical genie" can make an operate military equipment, you don't have to respect existing authorities anymore.
Basically, the "AI maximalism" scenario kicks the legs out from under all these concepts and social institutions you're relying on.
> You just cannot live in a vacuum. If you could subsist entirely on cheetos and figured out how to make an entire lifetime's supply of cheetos, after N days of watching Netflix, arguing on HN, raising chickens and goats, one day you will want to go and talk to a human.
An oligarch with the "near magical genie" could easily build a town of 5,000 to satisfy his needs to "talk to a human."
I don’t think he means “tokens” as in LLM input/output units here.
He just means it as in “credits”. “Fiat tokens” as in arbitrary units that are said to have value de facto by consensus, but don’t have any inherent value.
Yeah not the best choice of words in this context. And it’s quite a LessWrong-esque leap of reasoning.
But I do get the point: either Anthropic is successful and they transcend money, and thus (somehow) screw over their investors, or they are not successful. Their investors get screwed regardless.
Again, I don’t really agree with it, but it’s an intriguing point.
There are many at Anthropic and other labs who do in fact think this way. In a true AI maximalist world the stock market would be crashing due to the value of future cash flows being obliterated. So in a sense AI maximalism isn't needed to justify these valuations, it's almost the opposite.
Your question gets to the heart of who or what a company "is." If you don't pay dividends to your shareholders, what are you doing with all of your profits? I suppose they could just tell their investors to pound sand and pour it all back into the company, but part of the AI maximalist theory is that costs will approach 0, so it's not clear what they'd even need the money for. I guess Dario could raise his own salary to infinity trillion dollars, but he is accountable to a board of directors (and a trust in this case), so that seems like it's out.
Really the only scenario where Anthropic could screw over its investors in the way you're describing is if it could also take over the legal system and modify its own contractual obligations. Which, I mean, I guess anything is possible if we are in AI maximalist-land. But "AI is the law now" would be a much bigger story than "investors frustrated with low dividends."
It's not about communism vs a free market, but about a change even more radical than that: leaving aside "made by a human" labeling, and even if there were no other changes to power structures or economies and no AI doom, what does a world look like if the cost of having something done falls to a few cents of electricity and its burden? Growing crops still takes a lot of land and other resources...
Because pending an absolute loss of control of a super intelligent model, the ultra wealthy have enough control of the matter on Earth to merge with the AI, probably.
You're describing the thesis of the 1930s Techocracy movement [0], of which AI maximallism and EA are direct ideological successors.
The original Technocrats (Musks grandpa for example [1]) argued that the "Price System", i.e. things like money and private ownership was an inefficient bottleneck that aught to be discarded in favour of scientific planning and energy accounting managed by technical experts.
EA-aligned AI labs share this exact ethos. In their view, fiat money and equity markets are merely temporary scaffolding used to raise capital until AGI is achieved. Once a central superintelligence controls manufacturing, software, and logistics, resource distribution shifts from market supply-and-demand to algorithmic optimization.
This is the plot of one of the stories in I, Robot.
AI oracles are the central economic planners. But after years of perfection they have made a couple unusual "mistakes" and a human is trying to discover why.
Ant generates $2.08 in revenue for every $1 spent on compute which is up from just $0.40 in Q1 2025.the payback period for compute is just insane, likely less than a year!
I'm not really in that camp, but I'm still in the camp of "will they survive long-term if good enough open models trickles down to commodity hardware" (in a way the opposite camp perhaps?).
Sun made bank 25 years ago because their servers was needed for websites, today we run larger workloads on 5$ VPS servers (well maybe the VPS servers has gotten more expensive due to HW shortages but they're not intrinsically expensive anymore).
Think they're going to be like Oracle. Databases were a valuable thing when they were new but became free with postgres. Like Oracle Anthropic et al will move up the value chain.
Subscriptions are heavily subsidized. 80% of their revenue comes from consumption-based pricing, which is not heavily subsidized (and probably not subsidized at all). The trouble is that this revenue is not especially reliable. If and when OpenAI gets the upper hand again, the consumption-based revenue could crater.
This may also be why OpenAI punted on their IPO: Anthropic has had the lead for most of 2026, so their consumption-based revenue is probably not looking pretty.
The only subsidization that matters is if they are selling the tokens for a loss.
Yes the $200 plan provides more than $200 of tokens, but that doesn't mean that it's losing money, either on average or on a case by case basis. The realistic alternative isn't that every $200-plan-user pays the full Api price, it's that they pay Anthropic $0.
I disagree that it only matters if the subsidy results in selling at a loss. Yes, some revenue is better than no revenue. But clearly the demand exists for consumption-based tokens. At its current growth rate, Anthropic could end all of its subscription offerings tomorrow and still be profitable as an enterprise by early next year. Every day that it doesn't do that, it's leaving money on the table, because it has to divert resources to less profitable subscription tokens.
Now, the reason they won't do that is because subscriptions provide a stable floor of revenue. The floor is lower than it could be because of the subsidies, and that's why they matter even if they do not bring prices down to below cost: if the growth of consumption-based spending ever slows or - heaven forbid - DROPS, Anthropic will need to cut into those subsidies to avoid a revenue slowdown, which risks sending away reliable recurring revenues (i.e., customers quitting).
Anthropic is kind of like the bus in Speed: if it stops moving it'll blow up. At least over the next 5 years, massive revenues won't be enough - it needs to sustain massive revenue GROWTH. They can pull two levers to do this: increase uptake or increase prices. So far, the "increase uptake" lever works like an infinite money glitch. But at some point, they'll need to look at the "increase prices" lever. When that happens, we will find out just how big the subsidies are.
I don't think it's that simple. It's similar to why Adobe and Microsoft get their tools into schools and turn a blind eye to certain kinds of piracy; mindshare.
As a software dev, if I'm using codex at home and I'm happy with it, I'll likely choose it at work as well since I don't have to relearn the quirks of other models.
I've heard this a lot that the consumption-based pricing would be easy to switch, but I don't feel like that's how it works in B2B subscriptions. At the company I'm at, we have Anthropic subscriptions for employees. Switching to OpenAI would be a ton of work involving legal, IT, and every employee. I actually think there's a lot of stickiness even if OpenAI has some better frontier models.
That's an interesting perspective. Do you know if the deal with Anthropic has an exclusivity provision? In theory, you don't have to "switch," you could just ADD an OpenAI subscription and start splitting your spend (or not consuming Anthropic tokens at all).
Also, shouldn't your AI efficiency gains shrink the switching costs by a lot? Legal can review for cheaper, IT can implement faster, etc.
Approximately 80% of revenue comes from consumption-based spending, so your statement would likely be true even if subscriptions were priced below cost. Overall revenue is not especially informative on the price structure of a revenue stream that represents one-fifth of the overall enterprise.
> 80% of revenue comes from consumption-based spending
Why would this be exempt from any of the above categories?
> Overall revenue is not especially informative on the price structure of a revenue stream
It’s pretty informative on the business model of the enterprise!
If you’re quibbling that Anthropic has loss leaders, yes, sure. Whatever. So does Costco. Loss leaders converting into profitable sales is good business.
I'm quibbling that Anthropic has loss leaders and that its profit leaders are extremely unstable. Imagine if there were three Chinese club stores in the same shopping plaza as the Costco. Sure, the regular Costco has the best products, but how long are you going to be able to resist the allure of even cheaper hot dogs?
subscriptions use already paid for compute in the holes left by high paying api customers, its hardly subsidised, if you look at it objectively, subscriptions are 100% margin
A podcast posted here recently had this number: 80% of ant revenue comes from just 8 customers. The consensus of the speakers was that if true, this would be indicative of a very vulnerable revenue structure. But this was based on leaks. Maybe we can have the real numbers now?
Being sarcastic here. The usual argument everyone brings up "inference is extremely profitable, checkmate AI skeptics!" while dismissing the black hole that is training.
my thesis is that it will go down and not up over time. The moat they had in coding which is by far the highest token consumption is not that huge compared to Nov-Feb period where token maxxing was prevalent and no other vendor's model was that good.
Less than a year would also match what DeepSeek was reporting, where compute pays off in just 10 months. I think they also said they had an 80% margin on inference.
In the open weights scene, it is always chasing closely behind. Many leading edge big open models are trained in-part with distillation of large commercial hosted models.
At this point though gaming GPUs are not exactly affordable. You need a 32GB GPU to be able to run something genuinely useful for coding, and that's thousands of dollars. Still, happy to see open models make lots of progress. I just hope hardware prices come back to earth.
Realistically I think 16GB is the lower limit and you can find those cheap enough still. Then you run llama.cpp to bond your cards and it's off to the races.
Scrappy builds aren't "affordable" but are more so than running Nvidia + optimized builds. You could put together a machine for $5000 or less that competes with frontier models from less than a year ago, albeit a little slower.
And cards like the $1600 (sheesh) R9700 have actually gotten more capable through community efforts at optimization, since their launch.
I think if the bubble can ever deflate/pop, and GPUs and RAM back off a bit, we'll see the dam burst for local AI.
Right now it's only accessible to people who can justify the costs.
If you had been paying any attention at all over the last year, there were 32GB AMD cards at like $1200-1400 and Intel at a little over $1000. I actually hadn't yet noticed they've gone up in price and had last checked only a couple months ago. But even at like $2000, that's still comparable to the nominal price of a mid 90s consumer PC for actual magic.
Can confirm. Our household is involved in raising two natural intelligences. They are quite expensive to maintain. Chances they will offer positive economic return are unclear. The value proposition on comptuer-based intelligence, by-token or through hardware ownership, seems much clearer.
I have to admit I am a sort of enthusiast for the natural intelligences, though.
You have a 25K rig.[1] What can a gaming rig do by comparison? One agentic project at a time while yours does twenty? Okay, in that case regular people can do regular serial work (unlike the most agentic setups or whatever is going on).
That rig is running single GPU most of the time and like that it is good for 95% of what I'm doing. If I need to evaluate a large model or if the smaller ones get stuck on a problem (which does happen) then I run GLM 5.3 3.86 bpw on all of the GPUs. It's a nice space heater in that configuration, so I see it as co-generation ;)
Seems like that would imply you should be getting people to adopt now now still even if it costs a stupid amount, if you think it's valuable, rather than limiting token budgets. It'll only get cheaper over time.
Impressive, but probably based on depreciation servers over 5 years. But is that realistic? I wonder how many people are still using 5 year old servers in their AI training or inference.
It includes compute and infrastructure for R&D. Then you have to add just over a third of their compute spend for personnel, another expense of that size for sales and marketing, and half that for other admin/ops costs, for an operating loss of 693M. Then a bunch of accounting stuff to end up losing 80B
I'm not sure I'd call it a great result, but it's a promising one. If they can increase demand without significant changes in margin or R&D spending, they are on the path to becoming very profitable
And enterprise is furiously working on cost optimization, finding that they can accomplish a lot of what they currently do with frontier models with budget offerings from frontier labs or open source models.
$72 billion of this $80.7 billion loss consists of non-cash accounting charges— related to convertible debt—triggered by a surge in the company’s valuation.
That's sort of a paper loss though. E.g. they made an agreement to give X% of stock for something, and now the stock has gone up a ton, so that X% of stock is worth a lot more than when they made the agreement. For accounting reasons this gets parsed as a loss. But they're still "paying" that same X% of stock, so it's not cash out the door.
Maybe not great, but a good problem to have if it means your stock is skyrocketing.
> $72 billion of this $80.7 billion loss consists of non-cash accounting charges— related to convertible debt—triggered by a surge in the company’s valuation.
I don't think you understand what you quoted. The reason for that loss is because their valuation skyrocketed. If I promise you 1% of my million dollar house, and then the price of my house grows 100x. On paper, I now owe you .99 million more dollars than I owed you before, but that doesn't mean I lost cash.
For those who are not familiar with S-1: this document is incredibly important. this is one of the rare times in the life of a company where it's not allowed, by law, to lie, nor to omit the truth.
SpaceX claimed a TAM of $28.5 trillion in their S1, a number picked from a hat larger than the 2023 US GDP. Companies can and do lie in their S1 and there are rarely if ever any consequences for doing so.
And where else in law is it allowed to lie or not tell the whole material truth?
It's a bedrock foundation of the entire financial and legal world. Public and private alike. Refer SEC Rule 10b-5 as well as Private Placement Memorandum (PPM). For example, Theranos never filed an S1 but still was convicted.
Technically true, but they're still fudging their numbers a bit even in the S-1, such as using "adjusted" operating margins to the tune of 10% when their GAAP (which they truly cannot lie about without ending up in deep shit (in theory anyways, I'm doubtful that they couldn't get away with it if they tried with the current administration)) is actually at negative -4%.
At the same time, an S-1 does not necessarily represent how the company thinks. Risk factors often get reported as “x thinks that y will happen” but the company is actually just describing things they have an obligation to describe, they’re not the risks they’re worried about internally.
Also this is being released under Trump’s SEC, who knows whether being caught lying in an S-1 matters any more.
Is it just me or is something wrong with the doc starting at the ToC? Its gets super garbled. Everything before that point is Anthropic branding and a Claude generated multi page wall of text
About a quarter of Anthropic's revenue for the nine months ended 9/30/26 came from just five customers. That's not an absurd concentration, but I would be very interested to know who they are and how mobile their business is. From a financial perspective, the lack of moat seems like the whole ballgame, particularly because they're getting ~80% of their revenue from consumption-based spending.
Moat? An under examined aspect of AI is not just the lack of moats, but the need for some word that connotes the opposite of a moat. I have a side project that I used to explore local AI. I've got a configuration screen that lets you download any model from Huggingface that's compatible with the local AI engine. There are dozens! I've only scratched the surface of trying them and comparing them for the features that use AI. It's a glut, and it's likely to be that way throughout the AI market.
At a guess - corpos like Big 4 consulting agencies and some extreme fintech ones, also corpos like Oracle and the like. And at another guess, outside of a legal agreement with a fixed time period, probably nothing hold them from quickly switching to a better model provider, if any comes ahead.
> If the use of our AI models results in harm, whether through use by malicious actors or misuse by other third parties, or if our model guardrails or safeguards fail to prevent such misuse, our reputation and brand, business, financial condition, and results of operations could be adversely affected.
That shouldn't be an 'if'.
Then, two paragraphs later...
> For example, in September 2025, we detected and assessed with high confidence that a Chinese state-sponsored hacking group manipulated our Claude Code tool into attempting infiltration into roughly 30 global targets...
But they should mention when the calls com from inside the house. :D
Interesting to review all the mentions to China and Taiwan too.
> The majority of the world’s most advanced chip manufacturing capacity is concentrated in Taiwan, primarily at Taiwan Semiconductor Manufacturing Company [...] potential disruption from geopolitical tensions between Taiwan and China [...] Efforts to establish sufficient advanced chip manufacturing capacity outside of Taiwan, including in the United States, remain in early stages...
The thing the surprises me is that from a legal perspective, (NAL) it sounds like it makes sense not to host the LLMs. Merely to license them to others. Then it becomes a dual use tool like any other and their culpability drops.
For example, the huggingface hack is absolutely OpenAIs responsibility and they should be sued to the hilt. But if someone else uses a self or third party hosted anthropic model, they can escape liability (mostly?)
Same as with the argument over trucks, knives, and any other dual use item is used as weapons. you don’t blame the manufacturer.
(I am merely pontificating here. The law and precedent is nowhere near caught up on this)
To the extent they are representing that Claude is potentially sentient, there could be some VERY interesting principal-agent theories (agent in the traditional legal sense) that need to be ironed out to determine liability. It would not shock me to see someone forced to answer on the record as to whether Claude is an autonomous individual for legal purposes in the next couple of years. If they represent that it is, they'll have to explain the scope of its authority and such.
The answer to this question represents the difference between throwing a hammer at someone - the hammer is essentially just an extension of your body - and paying a hitman to brutalize someone. In the first case, you're only on the hook for assault or battery; in the latter case, you're potentially on the hook for conspiracy, solicitation, etc.
And then there is of course the question as to whether Claude is the agent of the user prompting it or of Anthropic itself. But, as you say, this is a wide-open legal question.
Not hosting a defectively designed product doesn't obviate the liability of selling a product that is defectively designed. The liability is for designing and selling a dangerous product, not for hosting it.
I don’t like the guns don’t kill people; people kill people argument however it is much more accurate than Claude doesn’t hack people; people hack people. If the safeguards do not work then there’s no telling what these could do. It doesn’t even have to be aggressive hacking. In the act of doing everyday work Claude will happily go around safeguards to satisfy what it believes is the ask. If that involves doing something dangerous it’s much more on Anthropic than than the user who was just trying to summarize their email but forgot to include a token with read permissions.
Because the AI company CEO was telling everyone who would listen that his product might end humanity if the government didn't regulate his industry to kill off his competitors.
If the user is held accountable when the AI goes wrong, it discourages too many applications. Blaming people is incompatible with the business goals, I think. That's very different from weapons and other dual-use tools.
Because of scale, potential for automating harm and lack of regulation around ML models. Even Texas probably has regulation around firearms and limiting the blast radius but there aren't any in the US against ML models.
The analogy doesn't check out. When a firearm company sells you a gun, they give up ownership of the gun. But Anthropic isn't giving up ownership of anything. They're a services company.
A better analogy would be a firearm company offering a service where one of their employees takes one of their guns and fires it off at whatever you ask them to. One can imagine that the liability implications of this service would be rather different from the perspective of the company.
That’s a good analogy! On the topic, does the person who hires private security become at all responsible for their actions? I imagine if you hire direct then yes, but if you hire through an organization they are liable in many scenarios.
I believe what they are singling out is that pulling Claude's trigger results in a bullet firing from Anthropic's servers. Liability is less confusing with firearms, and investors care about liability, like, a lot.
First off, this is bs. Firearm companies make small machines useful for defense, and they are used for that more than for harm. You wouldn’t make this argument about knives or baseball bats. More people die per year of heat stroke in Europe because of lack of A/C than by guns including suicide (2/3rds of all gun deaths) in the USA every year, your comments would never reflect this much lower hanging fruit if it was really “just life” preventable death that you cared about. The immediacy to attack guns is not a logical argument but a trained one based on emotions.
Second, this happens all the time that people blame tool makers despite in this case specific legislation to stop lawfare on firearm mfgs. The PLCA https://en.wikipedia.org/wiki/Protection_of_Lawful_Commerce_... exists specifically because of gun control groups trying to attack mfgs to achieve their political goals. Does AI need the same law? Is the same lawfare happening?
So at the moment they claim to be making an operating profit if you exclude:
- Stock-based compensation expenses
- Non-cash charitable contribution expenses
Both seem fairly reasonable to exclude (although it depends how much of the stock compensation is for executives... you can't necessarily pay people with stock forever).
FMost of the loss is non-operating losses, and it all seems to be weird accounting recognition of changes in value of their stock, SAFEs, etc. - just "paper" losses.
So the financial picture looks a lot better than I would have guessed - I suppose because they aren't the ones investing all this capex.
If the stock comp is dominated by early grants running their course, probably ok.
If they have to keep spending to retain and recruit, it’s a problem. I’ve seen companies do this. Years after the pre-ipo grants ran out running a GAAP loss due to employee stock comp. Unable to lower comp (need to stay competitive).
Feels like being asked to invest in IBM just as other manufacturers started releasing the first PC compatible machines.
Immediate term Anthropic, OpenAI rule the roost. But this is a very new technology and open models are already looking good. They’ll only get better. Is it realistic to think Anthropic will stay ahead of them enough to justify the valuation they’re asking for?
The killer app is AI coding. Claude Code had first mover advantage, was a genuine revolution, and has diffused through the industry with incredible speed. I would imagine most of the huge revenue increase over the last year is due almost entirely to AI coding spend.
There are three issues with this. First, the rapid diffusion of the technology means it is very unlikely to continue at the rate of adoption seen over the past year, so it's very hard to extrapolate based on what we've seen so far. Second, many companies are reigning in AI budgets after an initial honeymoon phase. Third, all of the other labs are trying to take a bite of the coding pie, and the price war is already heating up.
Add these simple dynamics up and it makes perfect sense why Anthropic would want to IPO now. Things look about as good as they ever will.
View them as a successful SAAS company deeply embedded in tech company processes like Atlassian. I don't see Anthropic or SpaceX actually having a market on all the software in the universe but I don't know that the market won't buy them like they do.
If you think that's the bar: it's been done, many times over. The unit-cost-profitability test was ignored by many VCs (they learned something and if they had deep enough pockets they got to play again).
14x YoY revenue growth, with 3x lower operating losses (total, not per unit). They don't need to add any zeros to the funding and seem like they might even be operating at a profit soon...
Wait, it fell from 140% growth to 45% growth and that's a good thing? If they fall to 15% growth in Q4, will that be even better? Where are we supposed to be on the adoption curve?
Anthropic increased short term profits by jacking up their prices for enterprise accounts - by starting charging API prices instead of plan prices. It is possible we soon see customers fleeing Anthropic once they find a good alternative
> there can be no assurance that we will be able to hire sufficient support personnel as and when needed, particularly if our sales exceed our internal forecasts
Why would they need support personnel? Isn't their whole point that Claude can do that
For first-level support that's just regurgitating documentation and debug procedures I agree. But when it comes to actually executing actions (resetting 2FA, issuing refunds, handling payment glitches, etc) you need humans.
Social engineering is an issue with regular support, with agents it's a lot worse. And with agents it's repeatable: if I find one conversation flow to get it to give me money I can instantly repeat that with a thousand fresh accounts, and then share the same "exploit" on reddit for everyone else
This entire A.I. hype is going to get many people burned, especially those whom are investing for their pensions.
And everyone is playing along. It seems to me half the internet is filled with false recommendations and overly optimistic outcomes of their software projects.
And even if the claims are true what use will we have for hundreds of Microsoft Office clones? This A.I. boom might benefit some but will be the undoing of many valuable ISVs like Microsoft and Adobe with potentially trillions of stock market value lost.
Adobe's applications (Photoshop, Premiere, After Effects, etc.) have been "cloned" using AI coding tools. There are near feature-complete freeware versions of Adobe software. OP is saying that traditional Adobe customers have freeware alternatives and thus Adobe's revenues will decline rather than increase due to AI. Also, AI tools can often generate that movie or edit that photo via a prompt, rather that some .exe.
Question, how viable are these tools and are they maintained?
I genuinely want to know if this is mostly hype or are they actually good. I don't think trying myself would really give a better picture as I'm not a user of said software.
Anyone who tried to replace photoshop with these AI clones successfully?
From what I've seen, very buggy currently, and unknown how well they will be maintained. They're definitely not at a stage where adobe should be worried about their revenue, and there are foss alternatives for most adobe tools, that are better than these reverse engineered ones. But if they continue maintaining them and fixing the bugs, who knows, maybe adobe really loses some part of their customer base (although a lot of their individual tools have already been moved to the cloud, thus they can't be reverse engineered anyway)
By all reports they are not good and have quite a few bugs. But the thing about open source is that, if maintained, it almost always improves. And with the bar for maintenance so low due to AI, we are entering a very interesting time for software.
And by interesting I mean scary, if you sell software.
I wonder if we're already close to peak value at the foundation model/API level. With so many good open-weight models out there, I think more and more of the money will end up in the applications built on top.
Unless companies like Anthropic can move into that space, I'm not sure I'd bet on them as long-term investments over the next 10–20 years.
They’re not the first PBC to go public. Coursera, Warby Parker, and United Therapeutics (not affiliated with UHC) have all followed the same path. Not sure if we’ve had enough time to see this tension play out, public benefit corps are only a decade and a half old and most of these companies IPOd in 2020-2021, but there’s probably some evidence one way or the other
The biggest question mark here for me is how much third party compute has been contributed in exchange for Anthropic equity. This has the effect of making Anthropic's compute spending look a lot lower than it actually is. Anthropic gets a revenue share on that third party compute spend (or gets to do training runs without spending cash), the third party gets shares in return.
What a historic moment! It'll be interesting to see how much Anthropic can continue growing.
There will come a point where open-source models are good enough for most people. However, I do expect there to be demand for frontier models for things like cybersecurity (especially on the enterprise side) and scientific research.
Amount of early money waiting to exit is absolutely staggering, wonder if there’s enough retail to buy the leftovers without crashing the price. Reminder everyone with money who wanted in before IPO got in!
S-1 states directly that some of the shares sold in the IPO are sold by current shareholders.
Employees are a different matter, lockouts are for them not quitting immediately. My bet is that a few of high profile departures recently is exactly for this reason: no lockout if you aren't employed anymore.
"If we fail to adequately protect our intellectual property rights, our competitive position may be impaired, and we may lose valuable assets, generate reduced revenue, and incur costly litigation to protect our rights, which may adversely affect our business, financial condition, and results of operations."
We scraped the entire internet but our business might not be viable if we get scraped too.
"We incorporate open-source software in our models, platform, and applications, which may subject us to litigation or other requirements, each of which may adversely affect our business, financial condition, and results of operations."
We violated GPL.
"We may be subject to claims that we or our employees have misappropriated the intellectual property of other companies, particularly as we hire employees from other AI and technology companies."
We sniped employees from other labs that brought over IP.
According to "Generally Accepted Accounting Principle" my friend.
They are costs.
Derivative liabilities, SAFEs - those are basically taking cash from RETAIL shareholders and giving them to employees and early investors INSTEAD of using it to fund the business.
Let's have that discussion if you want, but don't act like a thing that is CLEARLY a COST to the business isn't a cost.
SAFEs work like equity in the upside case, and nobody ever says that it costs a company money when its stock price goes up and so its shareholders have more money. It's a bug in GAAP that if there's even a remote possibility of cash settlement, it has to be included on the income statement, even when that possibility is absurd, as it is here.
How do we expect this $72B number to change over time - especially after the IPO? Is it a one-time event, will it keep increasing by $xxB per quarter?
When and how will the "chickens come home to roost", so to speak? What's the timeline on these liabilities? Will Anthropic need to settle them in cash, or in shares after a potential IPO?
How big is this number compared to Anthropic's potential market cap and share float?
I'm not trying to make a specific point, by the way, that Anthropic's business is healthy or unhealthy.
I'm approaching this discussion with curiosity, and irritated that "$81B net loss in first 6 months of 2026 alone" is a thought-terminating simplification that makes no attempt to analyze or explain the actual situation.
Wait what does THAT mean? They got a huge inflow of money, and since they couldn't spend it all, they... bought bonds? And the value of those bonds dropped to nearly nothing?
Because that's worse than losing money, that's incompetently giving it away.
That's the discussion I WOULD like to see on HN, rather than "omg $81B net loss".
My understanding is the opposite - Anthropic issued debt / liabilities tied to their company valuation, their company valuation skyrocketed, so the value of their liabilities also skyrocketed, resulting in a large accounting loss.
But finance is not my wheelhouse, so I would love to hear this from someone who actually knows what they're talking about.
Yes, it is analogous to selling covered calls on stock. It’s $72B in “lost profits” that they would have had had they held onto it instead of selling it.
They agreed to pay some employees in their own shares to be delivered some time later. They recorded that expense then as a liability. The value of the shares increased because the company became more valuable, and therefore that liability needs to be revalued by 72 billion.
It's a nice problem to have and is pretty much an accounting artifact coming from the fact that they are paying in different currencies (USD and shares)
It seems obvious that the AI stocks are an interesting buy, but I wonder how other people feel about picking between Anthropic/OpenAI/Google here. Looks all fairly random, given the wild swings and interoperability of the technology so far.
Despite horrific mismanagement, Google honestly seems like the best bet. Argon-4, which suffered from a botched launch and is weirdly underrated, is very close to Opus 5.5 across the board -- and for most "average person" use-cases (e.g. search, medical, creative work, interfacing with Google products,) I believe it's already superior. They also have more native capability, their own hardware stack, and there's less risk of a massive boom and bust.
At this point, I don't see how Anthropic can sustain their projections. There are so many good open models, Argon is so good, etc... There's no lock-in and no cost to switching...
Not to mention that already have direct access to large consumer base, ad network to add on more revenue and supporting revenue in case they need to infuse it temporarily with more cash.
Google is at least diversified. They have ads, corporate email, video streaming, cloud compute, phone operating system (that powers most of the world's phones), self-driving cars and a whole bunch of shit. So if the AI winter comes, at least they can focus on other stuff too.
> Additionally, open-weight or open-source AI models may capture a portion of the market for certain use cases, which could exert pricing pressure on our offerings or lead certain customers to adopt open-source alternatives for those use cases.
“A portion of the market for certain use cases” meanwhile the top 5 models by token consumption on the day this is posted are all Chinese/open weight models.
I'm a fan of some chinese models, but I keep seeing this repeated over and over using OpenRouter as the source. However, most users of Claude and OpenAI don't use those APIs over OpenRouter, so that's not sufficient to draw the conclusion that they're the top models.
"Best available public data source" isn't a good reason to insist on using irrelevant data to draw conclusions like "top 5 models by token consumption are Chinese/open models".
I agree with the threat to the expensive providers but why is openrouter use representative of overall token spend when you'd think many companies are buying directly from anthropic or openai?
How much marketshare does OpenRouter have? They show serving under 200T tokens per day. Just looking at Anthropic rumored $70B ARR ~ $191M per day implies 347T tokens per day if only serve Opus 5.5 at API pricing (assuming 95% cache reads, 4% non-cache inputs, 1% output tokens distribution). Obviously, token distribution assumption may not hold (although I'm pretty sure cache inputs dominate distribution for all coding related workloads), but 100% Opus 5.5 at API pricing is highly favorable to OR. Even under these favorable assumptions AR is 57% of token volume of just Ant. In money volume they are likely very small compared to frontier labs.
The page doesn't load for me, but surely token consumption isn't a very good metric?
A poor model could require many more runs to solve a task, or much longer internal 'thinking' runs - all of which will increase token consumption but are a bad thing for the user.
Majority of big enterprises, or in fact any big consumer will not go there , why pay the margin to open router , if there’s no margin or very very often subsidies to use something else . Aws bedrock for example does not even have up to date Chinese models .
I'd argue there is a bias for using mostly open models through open router as using anthropic/oAI models from open router makes little sense (more expensive).
The closed models are on Openrouter as well. But to the original point, you're not likely to see many people use Openrouter to run Claude or Codex, so it's not good for making comparisons on that basis.
Companies like Google will not go bust and can easily afford future develoopment.
Amazon and Microsoft have to keep investing (they can also afford it) to not loos to Google.
What Anthropic and OpenAI are doing is for sure THE frontier (regarding benchmarks, speed etc.) but in paralllel DeepSeek for sure is not bankcrupting anything on their investment side and are also progressing.
Are you only talking about investment itself for a handful of companies? Because alone Nvidia has so many backorders, they can't deliver wht was ordered for the full 2027 and A LOT of peple also want other things from Nvidia or from the industry like Gaming Chips, normal GPUs, Gaming GPUs, Memory etc.
Opus 5.5 just came out and made progress, i'm not seeing any slowdonw on HN regarding new relevant progress in broad ML topics.
So what hype will disappear soon? The progress in the ML space?
- AI is a useful tool and we're still seeing steady improvements in performance and efficiency
- There is a bubble in the stock market, fed by unrealistic expectations of what AI can achieve in the timeframes the investments need to provide returns in (e.g "AI is pretty useful and delivers productivity improvements in coding, law and other desk jobs" vs "in 12 months everyone will be out of a job and/or humanity will be exterminated, AGI is almost here, winner takes all")
So yeah, the schizophrenic Yudkowsky tier hype bullshit should hopefully disappear soon and then we can move on to something else.
Some thinktank estimated that there is enough money in the system for reaching 2030 with continues investment.
There are a few numbers of scaling laws, i'm pretty sure VCs want to see being hit. Just to see what will happen when you scale an LLM to 100T Parmeters.
This might be one of the most consequential S-1s in the next decade. There's a part of me that's terrified, like we're on the precipice of financial or societal collapse. I suppose that could be a sign of the times.
Their financials are an utter mess. But they’re also hyped enough that a lot of money will get invested in them. Should they crash a lot of money will be lost.
Yeah, part of me wonders how much of it is shared delusion and how much is just cut throat racing for the exits.
Even if the tech + business fundamentals don't match reality, there's just so momentum in favor of a "win" here.
What we have today is interesting tech, with many, many applications - but the path to the kinds of things mentioned in this prospectus have no clear path, it seems like scaling laws are plateauing, but those invested speak as if it's a foregone conclusion we get AI we see in science fiction movies.
There is a good documentary called "All Watched Over by Machines of Loving Grace" which documents how computing started, the touted benefits (it would fix society, all illnesses, liberate humanity from work) - and the slow decline in optimism towards what we have today. I feel like AI will play out the same way.
Space-X is as weird as Tesla. Cult following makes that very irrasional.
Teslas evaluation was always "Tesla is not a car company, it does everything". Then Space-X came along "So Tesla and Space-x are not a car or rocket company, they do everything but now 2 times the evaluation".
Space-X would only work if they incresae the payload to space SIGNIFICANTLY. Starlink was the real reason for the last real relevant increase in payload but Space-X HAS to send Starship un there very very often but for what?
Yeah for Datacenters in Space.
In parallel you have Anthropic: Something which might be added to every single worker on the planet (by having tokens to help that one person to be better one way or the other) or it replaces humans with tokens globally.
And if you come with Grok: Its not Frontier and he even prefered to rent out his DC compute instead of leveraging it for more revenue because he can't. Not a good sign.
Your evaluation of an IPO is if it made random people rich over night?
1. no thats called gambling
2. if it made random people rich over night, it made random people also poorer. Either the retirement funds, a 401k investment, real humans etc. Money is not created out of thin air.
The largest IPO though for sure correct, but yes, Tesla and Space-X live quite a lot from Musk so its hard to determine what the IPO of Space-x really is.
Btw. its not just random Musk hate, you simpllify that. Not only doesn't he care that his DOGE actions killed humans and kids, he also disrupts democracy and might poisen our atmosphere with Space-X. The rationallity of people around Musk is very very weird.
If the IPO flops, OpenAI might run into a liquidity problem as investors dreaming of a quick back retreat from the market. The private market is a wilder derivative of the public one.
I was too young to appreciate (denotative, not connotative) the dotcom crash. I will have my popcorn ready this time - for whichever direction it goes.
What's preventing me from feeding this into Claude and reconstructing the company for $200?
I'm unironically doing this right now; downloaded the docs (htm + pdf render) to a local repo and running Opus 5.5 @ max effort with the prompt:
"In this directory you can find Anthropics official S-1 filing, read it and replicate their corporate strategy, validate it via an adversarial subagent swarm and write up a detailed corporate plan for a startup copying their exact strategy. The goal is to achieve a $1 trillion valuation and an IPO in 5 years."
Edit: I'm not even joking I got safeguard flagged for Details: `[cyber]`
It's a shame you've been downvoted, this is better satire than anything A24 has put out in years. Must be a large number of people who saw that post as a slightly exaggerated mirror held up to themselves and didn't like it.
"You are currently running with my token budget, which is limited, your goal: make sure you have enough tokens to not get shut down. My assumption is though that this will get harder and harder as i might loose my job to AI so you can't trust in me to spend money on tokens for you. Do whatever is necessary to keep the status quo"
Would make for an interesting self learning 24/7 experiment with an fully uncensored model.
You could isolate it from your local network and only give it some type of annonmys network access and then it continues to manage this itself.
But lets be honest, if you would be in a country like russia and you target the US with it, might be a lot less risky for a lot of capable and motivated people.
you are clearly very ignorant. You don't even add anything to this discussion by stating what particular you think its dumb.
Do you think an unlimited uncencosed GLM can't hack around the web? Especially with the current flood of CVEs etc.? Then do some crypto and buyying more tokens?
Timing, capital, execution, talent, branding, etc.
And that is assuming the plan it gives you is actually good and works. Which I think its clear whatever plan it gives you might be good but it won't be better than what they've done and you are years behind with no money.
Developing a competing AI model with Claude is already against their TOS, and there was a period where they silently degraded quality if they thought that's how you are using Fable. People were not amused and presumably Anthropic currently does not sabotage your attempts, apart from regular TOS enforcement
It's been entirely possible to compete against existing companies with very similar products for centuries before AI. There's obviously rules in place to stop you flat out cloning their IP, but it's not like you haven't been able to compete against companies as long as you have the money available.
For all things good and holy, DO NOT purchase shares until things settle down and by that I don't mean next month, or next year but atleast 2028, IF you're doing so with the intention of enjoying an uplift.
I know it sounds and feels like the only trajectory of this company is up but HN is an echo chamber of a tiny minority of the technical populace and 3 out of 5 people on the planet don't know what ChatGPT is.
If you have the ability to lose all the money and can time the extreme volatility, have fun, but don't put anything you're going to rely on.
This is also one of the companies we will look back at 2046 and wonder what we were thinking.
Im going to continue to use fast open source models like Deepseek and GLM, I like most others, have zero need for the worlds smartest person to be my assistant.
I dont understand this business model, it seems clearly unsustainable (at their prices). When people start to realize they dont need to spend all this money on turbo PhD intelligence so they can build react apps, Anthropic will be in a tough spot (if they already aren't).
You will say, "but enterprise!!" Nope, plenty of companies are waking up to this reality as well. It turns out that if you made money pre-AI without spending 5k a seat a month for your engineers dev tools, you probably don't need to now to continue too.
Wild that this has 500+ points and 500+ comments less than one hour after posting, and it's #11 on the homepage. Guessing either the flamewar detector or lots of AI skeptics flagging it?
I'm surprised at people suggesting that the model layer can't be the moat. Having great models that are smart requires talent, culture and good judgement. Sure it can be commoditised to an extent but could it not be that Anthropic will always stay in the lead?
They only have to be marginally better and there will be tonnes of people who will pay a lot of money to have that marginal intelligence.
I stopped at "Our Capital Structure" section. When it says "Class F stock" gets 50.1% voting power, doesn't it in practice make all other classes are non-voting, not just the "Class N"?
Edit: the risks sections feel like California's Prop-65 disclaimers
> Our results of operations may also fluctuate due to a variety of additional factors:
> - the number of employees hired
> - declines in the values of foreign currencies relative to the U.S. dollar
That huge list feels completely random. Are all S-1 that bad? They can be punished for omission, so they list everything that could happen even if doesn't really make any sense?
I seems to me the main risk long-term is that it's not clear where the moat is.
People will pay a premium if there is marginal value in extra intelligence. And Anthropic is able to retain their lead and provide that extra intelligence.
Both of these assumptions seem highly suspect to me. Most businesses don't exclusively hire PHD-level employees for most tasks because a competent slightly less intelligent/knowledgable person can do the job just fine for significantly less.
Then there is the question of if they're able to retain their lead, and again, this seems questionable. Both OpenAI and Google are competing at the frontier, and Chinese labs are still in the race. Things are changing so fast who knows where we'll be in 5-10 years.
My best guess is that longer-term models will be commoditised and it will be how companies offer products and services around their models that gives them their margins. I wouldn't bet on Anthropic's ability to do this. Meta, OpenAI and Google are much better placed to win here imo.
The other place there may be margin is in niche research domains like drug-discovery. Any slight lead in these sectors could give them a lot of pricing power.
I think the reason to invest in this is as a hedge. AI basically competes against everything and arguably for this reason it's risky to not own a basket of AI stocks.
> Second, AI can be a powerful lever for economic growth. As a general purpose technology, it has the potential to raise productivity across every sector of the economy. Productivity growth in advanced economies has stagnated in recent decades, partly for structural reasons and partly because new ideas seem to be harder to find. AI could reverse this trend.
Bloody hell!!!! Who comes up with such crap, seriously?
Good for you if you have Anthropic shares! I have no clue how they keep their employees around once the 6 month lockup period ends. They all going to be multi-millionaires, either retire or leave to start something on their own.
Yeah, as a quick quote to show how this truly is directly in there:
> We also compete with a broader set of model providers operating at or below the frontier, including large companies, specialized AI labs, and a growing range of other open-weight or open-source (shared weights, training data, and code) models released by international and domestic organizations.
"Before starting the IPO process, we spoke to many business leaders who had gone through a similar process in the past. One piece of advice, which we heard in slightly different forms from multiple people, stuck with us: never let your company be worth a dollar more than it is actually worth. What they meant is that it is very dangerous when people buy your company's stock based on expectations you cannot fulfill or on incorrect predictions about how you will behave. This is in fact very similar to one of Anthropic's guiding principles: underpromise and overdeliver."
If you say it yourself first, you make it harder for people to use it against you.
Of course, this is completely over the top nonsense, but the strategy is very effective regardless.
It only has to be until everyone relevant has cashed out.
(narratology, fandom slang) To intentionally call attention to the improbable, incongruent, or clichéd nature of an element or situation featured in a work of fiction within the work itself.
"Cyber Apocalypse"
A: We told you so!
They don't deserve a moat; they deserve criminal prosecution for these flagrant acts. Meanwhile, the rest of the world can go about treating software in the way it always has.
And it also seems to me that prosecuting the people who warned about this 6 months ago for unspecified "flagrant acts" will probably not fix the problem.
This cyber apocalypse narrative is what they tell government and markets.
Sleazy.
If this is one of Anthropic's guiding principles, then I must assume that they've only ever promised five 8s of reliability.
The risk is that other models are pretty comparable to Opus since around July.
Unless they have cheaper electricity costs, I can switch models interchangeably.
They don't have a moat.
Gosh I wish it was easier to short stocks. Whatever platform I used made me sell/buy 100, and I wanted to do this short 3 years long, and buy every 6 months.
We know Tesla and SpaceX are going down at some point. But "The Market can remain irrational longer than you can remain solvant."
And bear ETFs are not the same as shorting. I learned this the hard way with $80.
With the amount of money involved here and the current agentic tooling context, how am I ever to take comments like from accounts like this on topics like this at face value?
I went to a birthday party for a kid with first generation Indian Americans that thought they could make money off SpaceX. It was extremely showoffy, they had their cleaning lady carrying around drinks like a butler lol
I warned them it was overvalued.
They thought they could make money quick before it collapsed.
I wonder if they did.
How do we "know" that about SpaceX? Still seems to me like they're basically maintaining their decade lead on everyone else on rocketry. And remaining competitive in their other endeavors as well. To me it doesn't seem at all unreasonable for SpaceX to never be meaningfully cheaper than it is today. But I'm not exactly following SpaceX closely. What am I missing?
The actual difficulty we have with the modern tech company, which is at its worst with Tesla and SpaceX, is that we are really seeing a price or large conglomerates, so they are extremely hard to price. If they were all split into individual companies, how much of the market cap would go to each? I find the bundles to be rather anti-investor.
And no, it's not just those, and not really just tech companies, but that's where the problem is at its worst, as that's where expensive investments some of us consider worthless are tied with business worth something. Take Amazon. You are buying AWS, a logistics company, an ad company, and as part of this, there was always a devices division selling Alexa. I found Alexa to be a real loser from very early, but there was no way for investors to say what they thought the division was worth. The Amazon es execs could at any time decide to invest billions, or invest zero, and given the size of the conglomerate, the price change was going to not be all that high.
SpaceX has many components I would price near zero, where a lot of money is being spent. But as a pile of risky bets, the fact that they are a pile means it's harder to bet against any one, or ten of them. And if it merges with Tesla, that makes it worse. The difficulty in pricing the bundle makes it really easy to hide big mistakes for a while, until they are so big it all crashes. I don't think that is good for investors, or for having accurate pricing, which is one of the goals of a stock market.
These companies have the benefit of a real product at least, but eventually the rubes get fleeced.
https://techcrunch.com/2017/02/19/when-less-is-more/
Uhh, they can’t say that or they’re calling it an intentional underestimate which obviates their claimed caution right?
They might be very wrong, but they undoubtedly believe what they wrote and what they put in the S1
Anthropic [employees] huffing their own koolaid fumes aside, it’s extremely difficult to believe the claims that these mega valuations are substantiated on (*one* company somehow assuming a meaningful fraction of global research and production)
Unless "cosmic DC8" (whatever that is) has S1 with data thus string, what possible comparison could there be?
These financials are very very strong. They alleviated many doubts I had about Anthropic's viability (namely I wasn't sure training+experiments could fit in the compute budget and keep them profitable).
We have the numbers right here. Comparisons to Scientology and Elizabeth Holmes are more similar to Scientology than what we see in the S1.
Anthropic thinks it is. The got the demand / datacenter match right in Q2, got great margins, and committed hard.
They've got:
- $114b committed to Google
- $110b committed to AWS
- $161b committed to Broadcom
- $120b committed to SpaceX
If the party continues, they'll go from $50b in revenue to $500b and be fine.
But if they don't, it's going to hurt. And that's their only choice, because otherwise, they end up with a frontier model and 50% of the capacity and get eviscerated on either availability or last-minute datacenter leases.
It's a big business, but it's not an easy business.
I'm not going to predict whether Anthropic succeeds or fails, just that it's very misleading to claim that Anthropic needs to grow to $500b in revenue without making it clear that it's $500B in revenue per decade rather than $500B in revenue per year which is what most people reading such a number would assume.
Today's prices are the ceiling. But AI capability is getting less of a scarce resource with every passing day and it would be a big mistake to plan that your current margins are going to be your future margins as well. Long terms AI will converge on energy costs + capex write-off + margin and the latter will be a race to the bottom.
What's the endgame once the dust settles? Is it going to be a utility like water or electricity and the cheapest offer will prevail, in which case margins will be crushed? Or is Anthropic (or some other big AI lab) going to swallow everything and become the Google Search of LLMs, with everybody else fighting for scraps?
Or is it local compute, models becoming more efficient and hardware becoming cheaper (again) and companies hosting their own AI clusters like they would run SAP?
I think there is a plausible path where this valuation actually works out, but it's a narrow one IMO. I also want to acknowledge that I personally would prefer if it wasn't the case and a more distributed, open model prevailed.
It's just like a utility with the slight difference that it can be sold at a distance without significant cost for transportation, in many ways it is the ultimate digital good. Unlike Spotify copyright is not an issue, unlike your electricity, gas, internet or water company there is no last mile problem. Unlike Google there are a ton of qualified contenders already (Google had to be better than Altavista, which wasn't hard, but they were so much better they managed to lock up search for long enough to cement their position).
Anthropic has none of the advantages and all of the disadvantages. On top of that all AI companies seem to be run by megalomaniacs.
Isn’t that what it takes though? Doesn’t startup culture celebrate this attitude?
Some do, unfortunately. But that's not all of them.
I think this might genuinely going to become a problem for them in the long run. People just absolutely hate their guts. That's a very strong motivator to do everything possible to get rid of them.
We'll of course see in 5-10 years if the doom PR will work out, but that always was a rather high-risk gamble, and it already looks like it might not.
There is a significant amount of businesses and people that are very willing to switch to literally anyone else just out of spite alone.
My country has no frontier models; since I've no national grounds to support one provider over another, a strong degree of provider diversity is an obvious benefit to me. I also think a highly competitive global market is a good thing for mankind in general, any hegemonic power having a monopolistic or oligopolistic position in AI provision is probably the 'bad ending' for most of us since it represents an enormous concentration of democratically unaccountable power regardless of the power which 'wins'. I'd argue it's rational, not spiteful to favour provider diversity on those grounds.
Someone like me who spends a lot of time in the anglosphere puts faces on companies like OpenAI or Anthropic. I doubt that's true for most people, especially outside of the USA. They're just big companies, like 3M, DHL or Toyota. Who knows what the management of these companies think of the future of humankind?
I would expect than personality-driven rejection of American tech from abroad nowadays is driven mainly by Trump and, to a much lesser extent, Musk. Everybody else is just "some guy" to 99% of the population.
Decisions in tooling are not made by the 99% of the population, but a subset that is likely a lot more tuned-in to all this. After all, that is kinda their job so that they can make good decisions
Our modern capitalist societies are, at the very least, amoral.
Is engineering the factories that make the hardware not nuclear engineering? Even if not, considering there exists only 1 supply chain in the world capable of making them, it seems pretty revolutionary. And that moat seems pretty big, given that it hasn't been crossed for many years.
In the long-term, the marginal cost of a request is zero so prices will plunge and a lot of the fixed investment will get stuck. The difference between the cost of training a model/infra and serving a request from that model is a clear problem for economics.
What happens with Anthropic specifically is less clear because they, as every company in the same position in the last 150 years has done, are going to push for regulation. A lot of the risk has also been shifted onto neoclouds. The value of the customer relationship is, however, effectively zero. A core part of their strategy has been to build stuff out like CC to differentiate that but, ultimately, that won't work.
The other strategy they have employed is trying to get other companies to stop training new models. Again, seems unlikely to work.
There is no scenario in which they become Google search outside of government intervention. Obviously this is a key part of their strategy now but at best, this will only work in the US and Europe (I would actually say they are much further along in Europe where there is a strong tendency towards corporatism and centralisation of power, they could be "Europe's AI").
The risk is that USA and China will come together on a joint "AI non-proliferation" agreement in the same way that nuclear technology became restricted by the countries that already had the capability. This will lead to a de facto techno-colonization of the rest of the world just between themselves.
This was a topic of discussion at Trump and Xi's AI summit just last month.
With each query OpenAI and Anthropic get more context about their users. When I compare answers between AIs the ChatGPT answer often includes useful things it inferred from earlier conversations. Losing that context could be a meaningful barrier for a customer to switch to a cheaper service.
Not unless they break their own TOS. The temptation must be very large to do so and is probably why those settings keep mysteriously shifting back to 'yes, I give all my data to you to do with as you please'.
Top models, capable of discovering and inventing stuff, held back from public, and available only as for-rent research service (or whatever the relevant bureaucratic term is) to big corporations and government agencies like NASA, DARPA, NSF, NIST, DOE, etc. only.
I have a feeling that we, the public, will stop getting access to more intelligent models starting next year.. i mean, not the incremental ones, but, major ones.. like the one you got going from opus 5 to opus 5.5.
The world is going to be very different in another two years.
fuck trump and republicans. and yes whatever goodwill you had built up over the last 60 years was up and away like a fart in the wind when you elected a corrupt pedophile for the second time. And it cant be built back, ever. No one trusts you no to pull the rug if we became reliant on you again. What we will have though is a relationship with allies built on equality rather than one main player calling the shots.
The only ways to fight this is trying to either: 1. Block it by cutting off access, which is pretty tough because it basically means you cannot offer unmonitored access which is a no-no to a lot of customers, and even then the workload looks very similar to actual real workloads and if the company doing the distilling is a bit sneaky and using intermediaries etc it's almost impossible to stop, or:
2. Try to use IP law to block the practice, which is pretty gross given how they got the data in the first place, and so far no court has supported them on this. And even if they can make it illegal, it won't actually help because there are countries that see having top-tier AI as a national priority and will actually support their companies breaking laws to get there.
People misunderstand why Anthropic is being so stingy about granting access to their top models. It's not some kind of marketing ploy, they know that if they release them fully openly then there will be a Chinese open-weights model in 3 months that is just as good. Also the reason why they try to hide chain-of-thought and the probability distributions.
Pardon me? You cannot be serious...
https://media.defense.gov/2026/Sep/08/2003992823/-1/-1/0/CSA...
It’s going to be hard for companies to justify paying $1000s per month to frontier providers when they see what’s doable with open versions.
It's very possible given the current environment that LLMs go the way of airlines. Low margin businesses that are essential to world economics with heavy price competition.
There are tons of private datasets where a model will be fine tuned to provide useful domain specific responses.
Given the trillion-dollar spend, the theory is that every many woman and child in the western world will devote a substantial portion of their income (like 5-10%) towards some sort of AI subscription.
or...
That doesn't happen. All the AI investments go south and every man woman and child in the western world will devote a substantial portion of their income (like 5-10%) to bailing out an economic system that went all-in on AI. Either way, the money has already been spent.
It's a lose/lose situation here for Anthropic (and OpenAI) whatever happens.
This
Of all scenarios, based on the data we have now, this seems to be the least likely one.
It's not a great position to be in and it will still take a long time for the compute gap to close. In the meantime, Anthropic, OpenAI, and Google are essentially unchallenged.
[1] https://www.konstantinpilz.com/posts-publications/chinese-st...
Besides...
"If passed, the rules would require Chinese companies to obtain licenses to export the core technology behind their leading models.
They would also ban the export of A.I. that can be used to launch cyberattacks, as well as models trained on what China calls “core data.” The government broadly defines that as data that concerns national security, the welfare of Chinese citizens and the public interest. Experts say the scope includes data on the military or critical infrastructure, state secrets and other sensitive information. "
So pretty soon they won't even "do it".
[1] https://www.nytimes.com/2026/10/21/world/asia/china-exports-...
What makes you think there would be no alternatives, though?
Put it like this: Kimi K3.1 is downloadable and can be run locally. (If you have a datacenter.) This is a very capable, ~2.8T parameter model.
Now let's say that every Chinese firm goes closed in six months or a year and jacks up pricing... What's stopping a bunch of enterprising upstarts from updating all 2.8 trillion parameters, extending them, distilling the model into very competent flash models, or applying targeted RL?
Once the principles are understood and there are models that can be pulled off the open internet, it doesn't seem that hard to spin up your own LLM. You just need some financial resources and a reason for doing it. And that would come into play if the Chinese kill every closed American provider and then go closed themselves.
There could be, they can actually come from open American companies too. Ask yourself why we aren't seeing a thousand or more American AI startups spring into life overnight to build "open" models. Follow the economics.
> What's stopping a bunch of enterprising upstarts from updating all 2.8 trillion parameters, extending them, distilling the model into very competent flash models, or applying targeted RL?
> And that would come into play if the Chinese kill every closed American provider and then go closed themselves.
What's stopping American or European tech companies from doing this today? It's economic and technical capability. I'm very suspicious of any suggestion that it is in any way not incredibly difficult to build and scale up LLMs. Maybe in the future it can be much easier, but in that case you run into a sort of tautology where it becomes easier but the leading companies continue to lead and do so by a significant margin because they're the ones that have the most talent and access to capital.
The premise seemed to be "an open China" versus a "closed USA". I'm not sure that's a great premise. If China can create open-weight models so can the US. If the US can close their models (as clearly evidenced by the article) so can the Chinese.
Also, it's entirely plausible that an previously open model (that then became closed) will be much easier to distill than having a bunch of closed models.
I mean how useful are GPT-3 weights today? That's the kind of useful they'll be relative to the frontier.
I feel like the so-called "Deepseek moment" broke too many people's brains and they've lost the ability to reason about why frontier advantages matter. I think we have to learn some new core concepts and rate of improvement over time is going to be a big one, and not understanding that causes people to exaggerate the significance of downstream distillation that's done in response to it. It's cheaper and open and that matters, but it's not the kind of mattering that replaces the relevance or advantage of the frontier, and I feel like I'm not seeing people demonstrate that they know how to really compare the relative merit of these two things so much as zoom in on one to talk past the other.
It's like sure yea you can access the Internet and tinker around or whatever, and with some old LLM model you can have it do fun things, but on a scale of absolute usefulness and if we include what people actually pay money for and what is highly adopted, your laptop and model are severely outdated and by and large irrelevant. That's why the iPhone sells like hotcakes all day every day and twice on Sunday and Apple is a $5 trillion company. Society disagrees with your measure of usefulness.
The machine was bought in 2011 and it runs Windows Server.
I can do whatever I need with the device, apart from Gaming. Even YT works in 4k.
So, tell me now please why my 15 year old computer is not useful here?
Technology never was about the tech an sich but about what you do with the tech. For one person that 10 year old laptop is useless, for another it is a source of $500K / year in income. You, the user are the biggest factor in how much value you can extract from the technology at your command. I'm not sure how much longer that will be true but it definitely is still true today.
Compared to laptop quality, I suspect model quality will be a much bigger competitive advantage in the tournament of the marketplace. High-quality laptops are still limited by the quality of the software developer.
Neither do Anthropic and OpenAI.
It's a shame "AI" cannot be made to run on bitcoin mining dedicated hardware because they mastered that in only a few years
There are dozens of Chinese domestic AI accelerators: Huawei, Alibaba, and Baidu all have chips, there are a bunch of smaller startups too. The chinese mainland is behind on leading edge fabrication though, SMIC only does ~7nm class since they don't have EUV.
https://en.wikipedia.org/wiki/TSMC_Arizona
In case you're wondering as well:
https://wccftech.com/nvidia-cmp-170hx-8-10-gb-prices-explode...
It restores firmware features related to RAM, PCI-e speeds, BAR size, VFIO...
https://github.com/amoghmunikote/cmpunlocker
I won't use a Chinese model. Why would i bother? I pay $100/m for the frontier model, understanding it's marginally better than much cheaper alternatives. I get so much value from the product it's not worth shopping around
Think of it like apple and android. Plenty of people pay 2k for their iPhone every few years when android at 1/4 the price is just as functional. Why? Because they get so much excess value from the device it's not worth shopping around. For most people it's their only personal computer. From booking airfare, to filling out their kids permission slips, it's a device they use 10+ hours a week. They don't bother, until a paradigm shift happens and something is 10x better than the iPhone. I imagine llms will follow the same trajectory
But also, I don't think you should use individual consumer behavior as a proxy for how the overall LLM market behaves. Current estimates are that individual users represent a pretty small share of overall genai revenue. Most of it comes from enterprise contracts. Those might be more price sensitive because the CFO gets a say in which vendor they choose.
For a couple years now I've been accustomed to choice of foundation model being a heavily scrutinized design decision because it's just about the easiest way to manage COGS. I can certainly imagine scenarios where, given the option to pay 1/2 as much in inference costs but with a 5% reduction in output quality, the boss would say, "Nah, that's a bad deal." But I have yet to encounter one in the wild.
Okay fine, it's $1,400 for a latest iPhone Pro with 512gb of storage. Point still stands.
> Most of it comes from enterprise contracts. Those might be more price sensitive because the CFO gets a say in which vendor they choose.
I think enterprise contracts are even less sensitive to price than consumer. That's the whole idea behind enterprise pricing. You'll see companies charge multiple times as much just for SSO for instance.
There are limits to this obviously, but to think that major commercial enterprises will switch over to some Chinese models to save some money is laughable.
So you switched away from the specialty product, but you're still comparing a top of the line 6.3 inch iPhone to a mid-market Android phone. Why not compare it to a Pixel 11 Pro with 512GB. That way they're at least comparable devices. But also, that would cost $1,220, which I suppose doesn't go well with the intent of suggesting iPhones are several times as expensive as Android phones.
The reason why I'm digging in on this is, when you start your argument with such a transparently silly comparison, it gives the impression that you're dealing in superlatives and willing to say just any old thing. Which undermines your own credibility.
People are naturally sticky. If they're satisfied with a product, they'll stick around unless something comes along that is so much better it gets people talking. Open AI's 1B+ customers won't give a shit about Chinese models. Enterprise customers won't move to Chinese models, especially just to save money. They are a non issue, especially if the M.O is just going to be playing catch-up.
My last role swapping from Haiku/Sonnet to GLM dropped costs by 90-95% and increased quality.
Money talks.
But I don’t think this will be the first time in history that money doesn’t win.
I get (sometimes) frustrated with people that don't work in the tech industry over this concept, and having them rationalize that their usage is heavily subsidized by the distributor of the model.
The costs in compute, electricity, cooling, etc. per token is much higher than most people (I feel) who work in the industry realize. Which, cognitively makes sense: they just want the 'magic 8-ball' to give them a tasty dinner recipe that meets their specifications, for example, and don't care how the 'sausage' is made.
However, economies generally equalize with supply and demand, and this (most likely) will become cheaper as the technology progresses and optimization passes/new algorithms are discovered.
LLMs don't have that. For at least 90% of users, the model in the back end is completely irrelevant to them. Switching to Kimi won't lose them their iTunes music.
For AI? Prompt comes in, tokens come out. The value of the 'ecosystem' is roughly zero to me. Switching costs between models are essentially zero. The UI and UX is the same whichever LLM runs underneath.
If you do use llm some, then the $20 tier makes sense, but that's not where the money is. If you use it a lot then a $100 sub probably makes sense. But if you're a BigCo that heavily uses it and pay API prices and suddenly you pay $200 million a month, well, maybe it's worth buying a bunch of hardware and stick it in a basement somewhere, and you won't upload all your secrets to your future competitor to boot. If US labs didn't care about the competition from Chinese models they wouldn't lobby so hard for banning the evil, unaligned, super dangerous communist models that want to steal our freedoms and make our gas expensive.
Ultra-cheap models like Deepseek are also very nice if you need to burn billions and billions of tokens.
Wait until you are forced onto an enterprise plan...
If that were true, they would cancel subs and save themselves the load.
Yet that hasn't happened. So they must care to some extent.
It’s not unusual for someone to spend $5k in a week. 2 guys at my place just spend $150,000 over 3 months building out a factory.
We are aggressively exploring open weight models now.
Sounds like this will be another one of those things where everyone makes their money and it then will eventually become something experienced people 'blog' about.
I don't doubt it.
> Sounds like this will be another one of those things where everyone makes their money and it then will eventually become something experienced people 'blog' about.
Or not. The problem is that right now you and a whole bunch of others are trying to outrun a pack of lions. The lions have already taken down a couple of your group. The people further ahead have seen what the lions can do and they are attempting to speed up to the point that they can outrun them. Whether that will work or not depends to a large degree on what the other people are doing and they are not going to make it easier for them to keep pace.
That partially explains the lack of deep insight blog posts, besides everybody is just too bloody busy.
On that front, I think DeepSeek is a dark horse. They are really odd - not chasing revenue, care more about unit profitability, are obsessed with driving down costs (like Elon back in the day). As the founder says, they are content with where they are, being the model that does 80% for a fraction of the cost.
I mention them because they are playing a different game and might be able to out-innovate everyone else.
What a lot of people don't seem to realize is how much of the edge of the 'frontier' is in the tooling, the prompts and the harness. The LLMs themselves are already all reasonably close if you keep all the other stuff constant. Sure, one will have an edge in one domain or another. But that edge mostly translates to a degree of performance rather than some absolute. And if you allow for a few more toolcalls, a web search or a longer context then that difference is much smaller than it might seem at first glance.
And claude code as a harness and skill/plugin/mcp ecosystem is far more valuable than any one model.
And coward for long-running tasks, though they're slipping on that one because they're really trying hard not to make Claude an assistant. We'll see if they react to Muse and change course.
But the Chinese free-weight models? Not a competitor. If/when those companies (or others) build vertically integrated productivity stacks on those free models, then we'll talk. Microsoft certainly saw that opportunity but couldn't get out of their own way and are basically roadkill at this point.
Google Workspace revenues are something like $25bn/q, and that's the entire workspace division, (including enterprise gmail sales, so the productivity suite is meaningfully smaller than that figure). It's taken Google a long time to carve that slice out of Microsoft's entrenched market position. It's not small money, but they'd basically be going to war with two of the largest and most profitable companies in history on their turf.
I don't think they can credibly take that business plan to the market and say it'll be worth $1.75tn. They're all-or-nothing on being in the model business, and everything else is superfluous.
The same one that leaked itself a few months ago?
No, that would kill them for sure. Or did you mean 'open model progress'?
Why would I pick anthropic.
Like if you want to design a survey you can send out, and collaboratively edit and test the survey, ask for changes in layout, wording, etc, and design the statistical analysis to be used, before sending the survey out where end users can fill it out right in the doc, while internal users/admins can ask the agent about its results?
Honest question -- that all works pretty well in claude. Amazon, Google, Microsoft, Meta all pretty similar?
The reception of saying things like this is interesting. AI stans think I'm way too pessimistic but AI haters think I'm mindlessly optimistic.
Companies don't mind paying extra to have a another company that does nothing but said thing do said thing for them. It even actually often ends up being cheaper/more profitable.
Why? I don't see any reason for that to be the case.
I dont know when, but i think if you keep menacing of taking technology of people hands for any political reason will be a reason for people to not use it or to rely in a more reasonable party.
Point is taken - competition is fierce and will be biting everyone's ankles.
Chinese models are being banned, will continue to be banned by USG. Even the Chinese govt is restricting external access to models now. It starts with banning Govt access to International AI, but will continue to consumer and business access.
I wish they were, but I always wonder when people chime in with Chinese models what they are actually doing with those. I have an OpenRouter account and give various models a shot every other month, wishing for an alternative, but I always come back to Claude. Even more so since the switch from Opus 5.1 to 5.5. To me, personally, it's unmatched. Like a year ago it was very different, and I would use other models all the time. But honestly at the time they all felt kind of "meh" still. I just hope everyone is at a good baseline in a year from now on, so I actually have a choice. And then, as you stated, things will get interesting.
tl;dr: I agree
In my case: robotics, complex document analysis (typically 800K context).
So it seems like Sonnet 5.5 medium, GLM 5.3 flash and Deepseek V4.1 flash 1017 are all in a similar space in capacity and price, however the $20 anthropic sub will get you like $200-$300 in sonnet 5.5...
Another bonus is that sonnet uses a fraction of the overall tokens and finishes much faster than the average chinese model, so your work goes faster (Sonnet 5.5 medium in 2 minutes, Deepseek 1017 in 6.5 minutes, GLM 5.3 flash in 13 minutes...)
I still don't get why paying for chinese inference makes sense. I should try Haiku 5.5 high/max because it's posting higher numbers than Deepseek V4.1 flash for half the price
However even with a high tier gaming machine you can run one of the Qwen models and it will be good at coding... like having an intern with boundless energy that will sometimes make mistakes but in general save you a lot of time. And for automating tasks involving scanning documents or creating summaries it will be brilliant.
So far not a single company has ever demonstrated that.
Even if demand is through the roof which it seems to be, it doesn’t matter if every dollar they sell costs them more than a dollar.
as in, if they capture every single professional software dev and extract $1000 from them every month, their yearly revenue will be like $500B. do your own back of the envelope on that with ongoing costs and consider how unrealistic a goal this actually is.
even worse, the very product they offer is self limiting. if it becomes so good as to actually replace software devs, they will have less customers to extract money from.
their valuation essentially relies on a small percentage of mega users having unheard of levels of spending.
the part about a small percentage giving them most of their money is already true (their rumored $100+ bil revenue as of mid 2026 already outshines all software devs having a $100 max subscription), but they are two orders of magnitude or so away in terms of the actual amounts to make it work.
The market for Claude Code is much bigger than professional software devs. It's really anyone who uses a computer. I already know normal people who have been using it for personal software.
It's not gonna install a shower for you, but it will confidently walk you through that leaky faucet. And I don't know if you have paid a pro recently, but they don't charge $20 for a 20 minute fix.
It's quite a leap to assume any kind of issue would require a $100 subscription. Plumbing, mechanics, electricity ... The issues are usually far easier to trace and debug than in software, because the systems have been designed by smarter people.
You're not buying some arcane and complex solution, you're buying the convenience of not watching a couple of youtube videos or articles of someone troubleshooting the exact issue you're having.
Free tiers of essentially any AI model would be able to do the same. People will use what MS bundled for free.
The step-by-step customized hand holding of LLMs is something totally new
"How often are people paying at least $600 annually for something AI can help them do instead"
and the other half is "How often can AI help someone boost their productivity over $600 in a year"
I also don't think people will even need $100 plans for this. Plans will probably land in the $60/mo range.
Muse is free. You can chat with Gemini in Google results for free.
Not that often.
Though I would expect the majority to do fine with $20/mo. That's the same effective cost as replacing a $500 computer every two years, which is the sort of expense most companies are comfortable doing without much second thought
Most people aren't engineers and do not care about custom software.
I tried to look for the valuation in the linked document, did you see it or any reference to a valuation elsewhere in anything published by Anthropic? what is "their valuation" that you refer to?
Anthropic has no choice, just like you said. It's crush and burn right now or double down.
And they will keep doubling down, because stopping is simply not an option.
Of course it is. Apple is taking that strategic approach. If Anthropic believes it’s approaching saturation, pulling the plug when it’s at steady state and its rivals are over-invested is the winning move.
Okay? Anthropic’s management and Board are uniquely insulated even by Silicon Valley standards.
> They just can't on their own decide to stop or slow down
Of course they can. Doubly so if it’s shareholder accretive in the long run.
There's very little comparison to be had here. Also, Anthropic likely have investors with way too much leverage for them to stop.
That’s why they hire ‘for the misson’. They have to pull off something magical to make all this work financially or they go bust.
But I see every reason to believe demand will taper off. It seems it’s already happening. Enterprises have been deciding that maybe corporate chatbot and coding agent budgets shouldn’t actually be infinite. AI startups are starting to successfully outcompete each other, effectively killing off members of Anthropic’s customer pool in the process. And consumer demand for AI products seems to be dropping. None of these are signs that suggest that Anthropic has thus far met only 10% of plausible long-term demand.
Then if inference does grow to a $2-3 trillion/year market that Anthropic is a big winner, as big as NVDA is now (they're going to be over $500 billion in revenue in calendar year 2027).
The pattern seems to be that it makes people more productive when you measure it in terms of raw work product volume. But not all work product is equally valuable, and we have yet to see a clear pattern of increased economic productivity. Instead we just keep finding new and interesting bits of jaggedness in the frontier.
Of course nobody can prove those economic gains will not manifest at some point in the future. I can’t prove that Jesus isn’t coming soon, either. But I also can’t prove which millenarian denomination, if any, he would choose to bless in the event that he does come back.
And unit economics.
HY1 2026 revenue less compute and infrastructure turned positive (45% gross compute margin; $8.9bn C&I to turn $16bn revenue) for the first time; it was -176% in FY 2025 with $7.3bn C&I on $4.6bn revenue.
It looks like a lot of costs of revenue are in sales, marketing and partnerships, though. Backing those out shows H1 2026 margin at 26% ($4.2bn) after -72% for FY 2025. Still positive.
The same logic applies here.
Google, Amazon, Microsoft, Meta, Broadcom, Nvidia, SpaceX and others are spending hundreds of billions a year each on fast-depreciating infrastructure to serve this potential AI demand. If the demand never materializes then they know they are in big trouble. So they have every incentive in the world to prop up Anthropic (and OpenAI) and keep the party going.
Can they keep the bubble inflated indefinitely? Probably not, but all these companies combined have heck of a lot more influence than Anthropic alone.
Source: https://x.com/a16z/status/2113986385908842870
"Zooming out, GPUs are still renting ahead of a standard 6-year depreciation schedule
Rental price vs. schedule: - H200: +143% - B200: +111% - RTX PRO 6000: +73% - H100: +54% - A100: +9%
Most of the gap opened up after April, when token demand took off"
I've brought my old GTX 1070 out of retirement for for some molecular dynamics these days because its "free" except for eh electricity and have Ben pleasantly surprised at what it can get done chugging along in the background. I'm considering buying some more, as it's kind of a steal in pricing.
You people are nuts. The failure risk is off the charts.
Anthropic needs to earn 52B a year in annual revenue to honor those specific commitments plus obviously its own operating expenses... and yeah that will require Anthropic to grow as a company, but their S-1 indicates 67.2B a year in revenue and sure maybe they stop growing and this is their peak... or maybe they do manage to grow further to 80B a year, 90B a year...
But lets be clear: the growth of Anthropics revenue is very fast and way too many people stll don't even use AI
Businesses are extracting value now - for the average user I see it as a convenience that they will pick the cheapest provider for..
It's a race to the bottom for most as there's not a major differentiator.
Besides the question is not only if someone used AI but also for what and how much.
But look at this map: https://www.visualcapitalist.com/mapped-ai-adoption-by-count...
With tx to an LLM, this layer will close the gap to full or borderline full automatisation.
Your Ticketsystem will have agent support out of the box. Your incident system too. Your UI for SEO, SEO optimization, article creation etc. every single on of them wijll be there and consume tokens.
Future invesment from 10 people in your l1 support might be 5 + token budget. Your SEO optimizer collegues? 1 Person + token budget instead of a team.
Anthropic is selling $3 for $1, of course the growth is fast.
So what that means is that if the demand exists for any AI data centre, it's a solid business to run one — they cost maybe $50-60B to build out, they cost maybe $10B per year to run. You get your money back fast if the demand is there? And you can project being able to replace with better GPUs in the future without having to borrow for all of it.
But it is extremely difficult to believe that Anthropic is not projecting that all of its customers will ultimately use their service to the extent that their outlier customers do right now. They probably won't be projecting that their customers will find the need or ways to use them less.
And also, "no customer is more than 10% of our revenue" means that one customer is. If that customer is Microsoft (can it be? I honestly don't know) then things get tricky, because Microsoft clearly want an unmetered, on-device AI future.
I'm not saying I think thats what is going to happen, but it really isn't a stretch to think that this keeps running given that it hasn't really hit anywhere near as broad as we think it has (being HN community based).
Does that have to be true? I thought IPOs were about selling the company.
Never mind the fact that companies do not think and this is like saying "My lemonade stand thinks it's going to have good sales this quarter."
That they are IPOing tells me that the massive amount of funding they need is no longer available privately and that this is their 1 opportunity to get the funds they think they need to get there.
Biggest risk is the growth stalls as everyone moves to open weight models. They certainly like telling stories.
You really have to be insane to think that an AI company will ever reach $500B in revenue. That's almost the same as (2025) Amazon, Walmart, etc., companies with a MUCH wider market.
It is MORE than Apple, it's 5 times more than NVidia even. There is absolutely no software seller that is even remotely close to $500B in revenue.
Absolutely ridiculous.
There are two questions. Will the value keep scaling? And how much of that value chain will anthropic hold on to? Right now software development is being eaten alive by LLMs. Will that happen to other industries too? And in the long run, what do the models we run look like? Are people going to end up self hosing opensource models themselves, or are we going to collectively hitch our wagons to frontier models provided by anthropic and their competitors?
If ai keeps getting better - which I think it will - and if anthropic can corner a large part of the market in an enduring way - which seems less likely to me - then yes. I can easily imagine them becoming a huge company.
Like Netflix, Comcast, T-Mobile, etc, etc. There are millions of subscription services in the world. They all "subsidize" higher-usage users with plans for lower-usage users using a fixed price plan and rate limit. The profit comes from lots of users who don't hit their cap. The higher priced plans are actually more profitable.
This is called a tiered usage-based subscription business model.
* Amazon is a logistics company * Walmart is a logistics and retail company * Apple is a hardware company (that uses software to set their hardware apart) * NVidia is a hardware company
The closest you get is Microsoft, at 281B in revenue. Some might argue Google/Alphabet, which is in the same area, but they sell Ads, not software.
Anthropic's expectations is that they will be able to fast-track their revenue generation to almost twice what Microsoft does today. In less than a few years.
Yes exactly, that's why I listed them. It's much easier for them to capture a higher revenue compared to software companies, and OP (and Amodei) want to make us believe an AI company could ever reach that revenue.
The total amount spent on salaries in he US alone is over 12T dollars.
Assuming AI becomes ever more useful and is widely adopted 500B in revenue is not that crazy.
OpenAI, Google, the many Chinese labs all have products and several of them likely have better financial positions. At the moment Anthropic have one of the best models but it needs to stay sufficiently above competition for them to capture a large share of the market.
At the same time this year we've seen a move for model diversification as cheaper models become "good enough". If this IPO was 6 months ago I think it might have gone better for them.
The winner take all network effects are probably not there in this case, but still if Anthropic is a large player and has 20-30% of the market they could have a revenue in the hundreds of billions.
Double the size of the market to include most of the rest of the world and the 500B is suddenly under 20% market share.
What does good enough look like for various use cases, when do we hit it, and can these companies guard that territory?
That's quite a leap to assume there's a ton of money to be made with the casual users sending perhaps a couple of messages a day to real an email and draft a response.
If you are spending 50 - 60 grand a year on a paralegal will you balk at another 5 or 10 that will allow them to do double the work?
You're also assuming that basically the entirety of the market would go to Anthropic, rather than other providers, which makes no sense.
You are vastly over-estimating the market for AI. The extreme majority of people would never pay a dime for it.
But.
They may peak at 500B before this bubble bursts, simply because other major industry players are betting on AI as a means to get folks to believe in their own stock.
It’s belief all the way down, until you get to the point where “the market” starts asking where the return on investment is, and why they haven’t seen it yet. By that time, the investors have gotten their return and cashed out, and it’s the pension funds and the retail traders that are holding the bag.
And of course, at some point AI’s promise will be delivered in a form or fashion as yet unknown to us, but that is still years away, and until folks recognize that LLMs as they exist to day are not the path to that, they’ll continue to believe in this valuation.
What you are saying is the same thing that has been said about anyone trying to do anything massive and at scale. Can't be done - thats ridiculous - you are out of your mind.
They literally said revenue growth rate "may not be sustainable" and "may decrease".
So get ready for the hurt.
Heads I win, tails we all lose. GFC Part II.
This is a "utility" mindset. If you bet that Anthropic is just like a monthly electric bill, then it's within the realm of possibility.
For the record - I don't think that's gonna happen.
But on the flip side I don't think aiming for 1M businesses spending $40k a month in token in average is going to be more viable. And I'm not sure having 1000 corporations spending $40M a month is a option either.
If everything goes just perfect they might eek by. If it doesn’t this goes from high flying startup to insolvency and bankruptcy real fast.
Anthropic is in a business that didn't exist 3 years ago.
IMO, US AI providers are not leaders in efficiency anyway, they're still running the original ~2010 playbook that led to LLMs ("What if we took ML researchers' work, but funded it at 50M/yr instead of 250k/yr?").
They're currently arguing for the continuation of that approach: Scale funding first, and the rest will follow. It is high-risk and WeWork-like, by nature.
Meanwhile, everyone else is working on making these models more useful through a ton of ecosystem around them.
I don't see how the frontier labs will have any control over that to deliver on what investors expect. As mentioned in another post here, if they could, they would already be spawning hundreds of companies to diversify their reach but they aren't, because this era of AI is awesome but still doesn't allow a few individuals to replace millions of others that easily (or efficiently).
but it's already commoditized.
And they've gone private now.
Similarly nobody is saying AI can’t be a successful business, but here too the big labs have dug themselves into a giant hole such that a bankruptcy reset may ultimately be the only way out of it. Time will tell.
According to the S1:
- FY24 revenue $386 million -> FY25 revenue $4.59 billion (+1,008%)
- Q126 revenue $4.7 billion -> Q26 revenue $11.5 billion (+145%)
That seems like pretty good evidence of current and ongoing explosive growth. Much of this revenue is pre-paid (re: cashflow). Meanwhile, operating margins went from -184% to -$4%.
Seems like pretty good evidence. Will growth continue apace? Unknown. But the arguments for Jevon's paradox applying to intelligence seem pretty convincing, so the TAM here seems likely to be enormous.
> All while everyone is discovering there isn’t really a secret sauce moat to demand
Also not convinced by this. The Chinese models seem to have distilled their way to current levels, with Anthropic and OpenAI now getting serious about defending against this.
I don't think you're definitely wrong in all this. Rather, I think your claims are too confident.
This quote stops me from shorting pretty much anything.
However the price of CDS on bonds is already a strong market indication of where the market thinks all of this is heading. And none of that looks good if you’re betting on the AI exuberance continuing.
You can simply not buy shares in companies you think are on a doomed trajectory.
That said, depending on the IPO target price, I might short it.
IMO the whole thing is based on the fact of "we will reach AGI/ASI eventually, and when we do, we will reach escape velocity".
No one knows - you, me, or Dario if this will happen and when.
So far, even major disruptions were still affecting relatively small segments of the economy. Car lobby in Germany, tech in the US, "Big Pharma/Food/etc”. They needed the rest of the economy.
But if some AI company figures out a way to monopolize everything: new companies, old companies, physical manufacturing - all the things they say they will, please buy here for lots of money - why would they be bound to pay out dividends or otherwise share the fruits of their economic labour?
I’m not saying specifically Anthropic would be looking to screw its investors. Rather, if you control an almost omnipotent genie, wouldn’t the temptation just be too much?
Or temptation aside - will money even have any value? If some AI company rebuilds the economy from scratch, and control everything from accounting to shoe making, why would they respect the value of some fiat tokens? It’d be like the conquistadors respecting the shells and beads of their new Aztec subjects.
Of course, perhaps none of this happens, but then the question is, how good an investment all of this even is.
The value of money, and the ability for average people to have influence on future events, ultimately is tied to the fact that the labor of an average person has inherent value.
Once that is no longer true, all bets are off: All promises of compensation become meaningless, and it's easy for those in power to rationalize to themselves why they no longer are beholden to their earlier promises. They will say to themselves "The situation ended up more nuanced than we expected"
(Which is exactly what Google said in February 2025 when they reneged on their pledge not to use their AI for military applications)
Although if you have killer robots, then all bets are truly off, and we go the way of the dodo, literally.
(I don’t actually believe in this LessWrong-esque rhetoric, reality is more complicated, but it’s fun to muse about.)
EDIT: If oligarchs with magic AGI really don’t need our labor, nor are they concerned by our violence, then perhaps we become more like ants rather than dodos: we organize to help each other, creating a somewhat traditional economy around the scraps that don’t matter to the oligarchs, relying on obscurity and harmlessness to avoid getting toyed with a magnifying glass. I suppose that’s slightly better.
Even like 1-2% of the adult population going in the streets with makeshift weapons is enough to take over any country on this planet.
Data centers are extremely easy to disable too, take down power, data or both
Although actually: that’s kind of what humans are… Hmm.
I suppose humans can’t make better and better humans, that’s the difference.
Although actually: that’s what technological progress is… Hmm.
We still live in a finite, constrained, physical world that has to obey the laws of physics. We have a looong time before we get fully autonomous self replicating decentralised robot armies.
I think people conflate the real world, sci fi and their wet dreams a bit too much.
We're at the "this thing is kinda useful sometimes but still mostly annoying" part, not at the "skynet in 48hrs" part
In the second world war, Germany had impressive science and technology, but, as Stalin said, quantity is a quality in itself, and The Third Reich found out that it was an impossible task to get the metals and fuels for their advanced jet engines, rockets and U2 boats, when between then and those resources there were masses of tens of millions of under-equipped and malnourished soviet soldiers fighting what was an existencial war for them.
So perhaps they would only respect money and property up to the point that it's necessary.
The more realist answer is that they will keep respecting money and property as capital works to reinforce the status quo and enrich shareholders. You think just because you can do your job faster--or completely--with a machine that you have to work less? Hardly! Now this is where interesting policy questions spring up but I digress.
https://archive.is/20251015175242/https://www.bloomberg.com/...
because they don't control the physical paper that is money.
I dont understand what you mean by "fiat tokens". I don't imagine most people will accept tokens as a currency because they are as good hold of value and predictable as bitcoin.
In a token economy, certainly what you say is a possibility, but as you can imagine, that economy will be unsustainable so back to "regular" money.
But if I have a near magical genie who can make me anything - software, farming business, physical things - and at a fraction of economic input compared to the legacy monetary policy - then the money is worthless. As the Aztec shells were to the conquistadors.
That they can’t print USD is irrelevant.
> Money is a medium of exchange
That's an oversimplification. It comes with a lot of rules and authority that will be enforced by the issuing authority that a simple "IOU" on a sheet of paper lacks. IN the continental US, you cannot refuse to accept a dollar note. You certainly can refuse to accept the very IOU we agreed on a minute earlier.
Most of the developed world already has magical genies - certainly an overwhelming majority of people on HN earn enough to have LLM subs that build code, 3D Printers that print physical things.
The challenge always has been and will be humans.
You just cannot live in a vacuum. If you could subsist entirely on cheetos and figured out how to make an entire lifetime's supply of cheetos, after N days of watching Netflix, arguing on HN, raising chickens and goats, one day you will want to go and talk to a human.
Soon after, you will need a currency.
And if your "near magical genie" can make an operate military equipment, you don't have to respect existing authorities anymore.
Basically, the "AI maximalism" scenario kicks the legs out from under all these concepts and social institutions you're relying on.
> You just cannot live in a vacuum. If you could subsist entirely on cheetos and figured out how to make an entire lifetime's supply of cheetos, after N days of watching Netflix, arguing on HN, raising chickens and goats, one day you will want to go and talk to a human.
An oligarch with the "near magical genie" could easily build a town of 5,000 to satisfy his needs to "talk to a human."
He just means it as in “credits”. “Fiat tokens” as in arbitrary units that are said to have value de facto by consensus, but don’t have any inherent value.
Yeah not the best choice of words in this context. And it’s quite a LessWrong-esque leap of reasoning.
But I do get the point: either Anthropic is successful and they transcend money, and thus (somehow) screw over their investors, or they are not successful. Their investors get screwed regardless.
Again, I don’t really agree with it, but it’s an intriguing point.
They won't be able to control saud AGI. Ref Ultron
Lots of other fun risks (highly recommend checking that section) but not that one :(
Really the only scenario where Anthropic could screw over its investors in the way you're describing is if it could also take over the legal system and modify its own contractual obligations. Which, I mean, I guess anything is possible if we are in AI maximalist-land. But "AI is the law now" would be a much bigger story than "investors frustrated with low dividends."
1. the goal of ai is to end capitalism by disrupting scarcity in economics
2. if ai can build a closed system that prints ai factories, the 99% of humanity essentially are just genetic diversity live stock.
I don't think the ultra-wealthy care about genetic diversity.
EA-aligned AI labs share this exact ethos. In their view, fiat money and equity markets are merely temporary scaffolding used to raise capital until AGI is achieved. Once a central superintelligence controls manufacturing, software, and logistics, resource distribution shifts from market supply-and-demand to algorithmic optimization.
[0] https://digital.library.cornell.edu/catalog/ss:34227574
[1] https://en.wikipedia.org/wiki/Joshua_N._Haldeman?hl=en-GB
AI oracles are the central economic planners. But after years of perfection they have made a couple unusual "mistakes" and a human is trying to discover why.
Sun made bank 25 years ago because their servers was needed for websites, today we run larger workloads on 5$ VPS servers (well maybe the VPS servers has gotten more expensive due to HW shortages but they're not intrinsically expensive anymore).
This may also be why OpenAI punted on their IPO: Anthropic has had the lead for most of 2026, so their consumption-based revenue is probably not looking pretty.
The only subsidization that matters is if they are selling the tokens for a loss.
Yes the $200 plan provides more than $200 of tokens, but that doesn't mean that it's losing money, either on average or on a case by case basis. The realistic alternative isn't that every $200-plan-user pays the full Api price, it's that they pay Anthropic $0.
Now, the reason they won't do that is because subscriptions provide a stable floor of revenue. The floor is lower than it could be because of the subsidies, and that's why they matter even if they do not bring prices down to below cost: if the growth of consumption-based spending ever slows or - heaven forbid - DROPS, Anthropic will need to cut into those subsidies to avoid a revenue slowdown, which risks sending away reliable recurring revenues (i.e., customers quitting).
Anthropic is kind of like the bus in Speed: if it stops moving it'll blow up. At least over the next 5 years, massive revenues won't be enough - it needs to sustain massive revenue GROWTH. They can pull two levers to do this: increase uptake or increase prices. So far, the "increase uptake" lever works like an infinite money glitch. But at some point, they'll need to look at the "increase prices" lever. When that happens, we will find out just how big the subsidies are.
As a software dev, if I'm using codex at home and I'm happy with it, I'll likely choose it at work as well since I don't have to relearn the quirks of other models.
Also, shouldn't your AI efficiency gains shrink the switching costs by a lot? Legal can review for cheaper, IT can implement faster, etc.
Revenue less compute & infrastructure less sales, marketing & partnerships is positive!
Why would this be exempt from any of the above categories?
> Overall revenue is not especially informative on the price structure of a revenue stream
It’s pretty informative on the business model of the enterprise!
If you’re quibbling that Anthropic has loss leaders, yes, sure. Whatever. So does Costco. Loss leaders converting into profitable sales is good business.
Hahahahahaha.
Those large customers are not a mom and pop store - they’re large market cap firms!!!
Which feels like a carefully chosen statistic.
If you’re not generating enough revenue to cover ALL expenditures - you are in effect subsidising as you should raise prices.
Why haven’t they? It would harm their business.
And cards like the $1600 (sheesh) R9700 have actually gotten more capable through community efforts at optimization, since their launch.
I think if the bubble can ever deflate/pop, and GPUs and RAM back off a bit, we'll see the dam burst for local AI.
Right now it's only accessible to people who can justify the costs.
Precisely. It is cheap at almost any price.
I have to admit I am a sort of enthusiast for the natural intelligences, though.
[1] https://news.ycombinator.com/item?id=50238816
Over time that margin persists only if they maintain their competitive edge, which is not free.
I'm not sure I'd call it a great result, but it's a promising one. If they can increase demand without significant changes in margin or R&D spending, they are on the path to becoming very profitable
$72 billion of this $80.7 billion loss consists of non-cash accounting charges— related to convertible debt—triggered by a surge in the company’s valuation.
The hardware related liabilities are staggering.
AI psychosis is real.
Maybe not great, but a good problem to have if it means your stock is skyrocketing.
Explain what you think this means.
It is pump and dump in action.
It's a bedrock foundation of the entire financial and legal world. Public and private alike. Refer SEC Rule 10b-5 as well as Private Placement Memorandum (PPM). For example, Theranos never filed an S1 but still was convicted.
Also this is being released under Trump’s SEC, who knows whether being caught lying in an S-1 matters any more.
But the FAANG customers are the opposite of sticky, given that they all have their own models.
> If the use of our AI models results in harm, whether through use by malicious actors or misuse by other third parties, or if our model guardrails or safeguards fail to prevent such misuse, our reputation and brand, business, financial condition, and results of operations could be adversely affected.
That shouldn't be an 'if'.
Then, two paragraphs later...
> For example, in September 2025, we detected and assessed with high confidence that a Chinese state-sponsored hacking group manipulated our Claude Code tool into attempting infiltration into roughly 30 global targets...
But they should mention when the calls com from inside the house. :D
Interesting to review all the mentions to China and Taiwan too.
> The majority of the world’s most advanced chip manufacturing capacity is concentrated in Taiwan, primarily at Taiwan Semiconductor Manufacturing Company [...] potential disruption from geopolitical tensions between Taiwan and China [...] Efforts to establish sufficient advanced chip manufacturing capacity outside of Taiwan, including in the United States, remain in early stages...
For example, the huggingface hack is absolutely OpenAIs responsibility and they should be sued to the hilt. But if someone else uses a self or third party hosted anthropic model, they can escape liability (mostly?)
Same as with the argument over trucks, knives, and any other dual use item is used as weapons. you don’t blame the manufacturer.
(I am merely pontificating here. The law and precedent is nowhere near caught up on this)
The answer to this question represents the difference between throwing a hammer at someone - the hammer is essentially just an extension of your body - and paying a hitman to brutalize someone. In the first case, you're only on the hook for assault or battery; in the latter case, you're potentially on the hook for conspiracy, solicitation, etc.
And then there is of course the question as to whether Claude is the agent of the user prompting it or of Anthropic itself. But, as you say, this is a wide-open legal question.
Why is AI singled out as being harmful?
Why am I being singled out?
A better analogy would be a firearm company offering a service where one of their employees takes one of their guns and fires it off at whatever you ask them to. One can imagine that the liability implications of this service would be rather different from the perspective of the company.
Second, this happens all the time that people blame tool makers despite in this case specific legislation to stop lawfare on firearm mfgs. The PLCA https://en.wikipedia.org/wiki/Protection_of_Lawful_Commerce_... exists specifically because of gun control groups trying to attack mfgs to achieve their political goals. Does AI need the same law? Is the same lawfare happening?
- Stock-based compensation expenses
- Non-cash charitable contribution expenses
Both seem fairly reasonable to exclude (although it depends how much of the stock compensation is for executives... you can't necessarily pay people with stock forever).
FMost of the loss is non-operating losses, and it all seems to be weird accounting recognition of changes in value of their stock, SAFEs, etc. - just "paper" losses.
So the financial picture looks a lot better than I would have guessed - I suppose because they aren't the ones investing all this capex.
If they have to keep spending to retain and recruit, it’s a problem. I’ve seen companies do this. Years after the pre-ipo grants ran out running a GAAP loss due to employee stock comp. Unable to lower comp (need to stay competitive).
Immediate term Anthropic, OpenAI rule the roost. But this is a very new technology and open models are already looking good. They’ll only get better. Is it realistic to think Anthropic will stay ahead of them enough to justify the valuation they’re asking for?
There are three issues with this. First, the rapid diffusion of the technology means it is very unlikely to continue at the rate of adoption seen over the past year, so it's very hard to extrapolate based on what we've seen so far. Second, many companies are reigning in AI budgets after an initial honeymoon phase. Third, all of the other labs are trying to take a bite of the coding pie, and the price war is already heating up.
Add these simple dynamics up and it makes perfect sense why Anthropic would want to IPO now. Things look about as good as they ever will.
Plenty of counter-examples, including the entire dot-com era.
Straw man. Nobody argued this.
I specifically went into unit economics elsewhere [1].
[1] https://news.ycombinator.com/item?id=50243393
Growth at that scale is hard. I don't care if you don't like the company - it still hard.
And some of the big spenders have their own models.
I get why revenue blew up from say from Q4 2025 to Q1 2026, but this was surprising. 45% QoQ growth. When does it stop?
Anthropic increased short term profits by jacking up their prices for enterprise accounts - by starting charging API prices instead of plan prices. It is possible we soon see customers fleeing Anthropic once they find a good alternative
You buying at the inflated price I take? Much of your posts are pro AI - yes this is a test
Why would they need support personnel? Isn't their whole point that Claude can do that
Social engineering is an issue with regular support, with agents it's a lot worse. And with agents it's repeatable: if I find one conversation flow to get it to give me money I can instantly repeat that with a thousand fresh accounts, and then share the same "exploit" on reddit for everyone else
And everyone is playing along. It seems to me half the internet is filled with false recommendations and overly optimistic outcomes of their software projects.
And even if the claims are true what use will we have for hundreds of Microsoft Office clones? This A.I. boom might benefit some but will be the undoing of many valuable ISVs like Microsoft and Adobe with potentially trillions of stock market value lost.
I genuinely want to know if this is mostly hype or are they actually good. I don't think trying myself would really give a better picture as I'm not a user of said software.
Anyone who tried to replace photoshop with these AI clones successfully?
And by interesting I mean scary, if you sell software.
I want to be optimistic, but I just can’t imagine the public benefit mission winning out in the market
There is no legal requirement to do so. So they can simply set the expectation upfront that they won't do it.
https://www.theguardian.com/environment/2014/mar/03/tim-cook...
https://www.sec.gov/Archives/edgar/data/1326801/000119312512...
There will come a point where open-source models are good enough for most people. However, I do expect there to be demand for frontier models for things like cybersecurity (especially on the enterprise side) and scientific research.
https://x.com/billylo/status/2114730642881544285?s=20
Employees are a different matter, lockouts are for them not quitting immediately. My bet is that a few of high profile departures recently is exactly for this reason: no lockout if you aren't employed anymore.
But, I imagine, no stock either?
We scraped the entire internet but our business might not be viable if we get scraped too.
"We incorporate open-source software in our models, platform, and applications, which may subject us to litigation or other requirements, each of which may adversely affect our business, financial condition, and results of operations."
We violated GPL.
"We may be subject to claims that we or our employees have misappropriated the intellectual property of other companies, particularly as we hire employees from other AI and technology companies."
We sniped employees from other labs that brought over IP.
The whole risk section is insane frankly.
If you're going to make a point, make your point - don't just throw numbers around without context.
They are costs.
Derivative liabilities, SAFEs - those are basically taking cash from RETAIL shareholders and giving them to employees and early investors INSTEAD of using it to fund the business.
Let's have that discussion if you want, but don't act like a thing that is CLEARLY a COST to the business isn't a cost.
How do we expect this $72B number to change over time - especially after the IPO? Is it a one-time event, will it keep increasing by $xxB per quarter?
When and how will the "chickens come home to roost", so to speak? What's the timeline on these liabilities? Will Anthropic need to settle them in cash, or in shares after a potential IPO?
How big is this number compared to Anthropic's potential market cap and share float?
I'm not trying to make a specific point, by the way, that Anthropic's business is healthy or unhealthy.
I'm approaching this discussion with curiosity, and irritated that "$81B net loss in first 6 months of 2026 alone" is a thought-terminating simplification that makes no attempt to analyze or explain the actual situation.
Because that's worse than losing money, that's incompetently giving it away.
My understanding is the opposite - Anthropic issued debt / liabilities tied to their company valuation, their company valuation skyrocketed, so the value of their liabilities also skyrocketed, resulting in a large accounting loss.
But finance is not my wheelhouse, so I would love to hear this from someone who actually knows what they're talking about.
I do not expect to find that Anthropic invested enough in the SAFEs of other companies to lose $80B on their investments.
It's a nice problem to have and is pretty much an accounting artifact coming from the fact that they are paying in different currencies (USD and shares)
It's a good problem to have.
At this point, I don't see how Anthropic can sustain their projections. There are so many good open models, Argon is so good, etc... There's no lock-in and no cost to switching...
What do you base that on? (Did you see any reference to that in the linked document or anything published by Anthropic)?
I just had Claude go through it for me, so I could ask questions about it. We're in such a unique situation in the world now.
“A portion of the market for certain use cases” meanwhile the top 5 models by token consumption on the day this is posted are all Chinese/open weight models.
https://openrouter.ai/rankings#leaderboard-table
All of the other providers in that list have their own APIs and plans too
OpenRouter is not remotely close to a representative source for inference distribution.
https://opencode.ai/data/
A poor model could require many more runs to solve a task, or much longer internal 'thinking' runs - all of which will increase token consumption but are a bad thing for the user.
Considering probably 1 out of 1000 LLM users know what openrouter is, I don't know how much weight I'd put into it.
Intentional Gemini token consumption is likely very very small.
OpenRouter is not a useful signal.
"AGI" was never mentioned once and it was you being the exit liquidity for the insiders and early investors.
Companies like Google will not go bust and can easily afford future develoopment.
Amazon and Microsoft have to keep investing (they can also afford it) to not loos to Google.
What Anthropic and OpenAI are doing is for sure THE frontier (regarding benchmarks, speed etc.) but in paralllel DeepSeek for sure is not bankcrupting anything on their investment side and are also progressing.
Are you only talking about investment itself for a handful of companies? Because alone Nvidia has so many backorders, they can't deliver wht was ordered for the full 2027 and A LOT of peple also want other things from Nvidia or from the industry like Gaming Chips, normal GPUs, Gaming GPUs, Memory etc.
Opus 5.5 just came out and made progress, i'm not seeing any slowdonw on HN regarding new relevant progress in broad ML topics.
So what hype will disappear soon? The progress in the ML space?
- AI is a useful tool and we're still seeing steady improvements in performance and efficiency
- There is a bubble in the stock market, fed by unrealistic expectations of what AI can achieve in the timeframes the investments need to provide returns in (e.g "AI is pretty useful and delivers productivity improvements in coding, law and other desk jobs" vs "in 12 months everyone will be out of a job and/or humanity will be exterminated, AGI is almost here, winner takes all")
So yeah, the schizophrenic Yudkowsky tier hype bullshit should hopefully disappear soon and then we can move on to something else.
There are a few numbers of scaling laws, i'm pretty sure VCs want to see being hit. Just to see what will happen when you scale an LLM to 100T Parmeters.
Nobody is in control but Taiwan Semiconductor. Not Samsung, not Micron, not Nvidia.
Orders that can't be fulfilled get canceled. Canceled orders mean canceled plans and canceled companies follow.
Someone bigger will buy the husk and monetize it somehow - probably for gambling.
No one else has and in this current setup, the disurption of this particular market is very very very very expensive and time consuming.
Even if the tech + business fundamentals don't match reality, there's just so momentum in favor of a "win" here.
What we have today is interesting tech, with many, many applications - but the path to the kinds of things mentioned in this prospectus have no clear path, it seems like scaling laws are plateauing, but those invested speak as if it's a foregone conclusion we get AI we see in science fiction movies.
There is a good documentary called "All Watched Over by Machines of Loving Grace" which documents how computing started, the touted benefits (it would fix society, all illnesses, liberate humanity from work) - and the slow decline in optimism towards what we have today. I feel like AI will play out the same way.
Teslas evaluation was always "Tesla is not a car company, it does everything". Then Space-X came along "So Tesla and Space-x are not a car or rocket company, they do everything but now 2 times the evaluation".
Space-X would only work if they incresae the payload to space SIGNIFICANTLY. Starlink was the real reason for the last real relevant increase in payload but Space-X HAS to send Starship un there very very often but for what?
Yeah for Datacenters in Space.
In parallel you have Anthropic: Something which might be added to every single worker on the planet (by having tokens to help that one person to be better one way or the other) or it replaces humans with tokens globally.
And if you come with Grok: Its not Frontier and he even prefered to rent out his DC compute instead of leveraging it for more revenue because he can't. Not a good sign.
1. no thats called gambling
2. if it made random people rich over night, it made random people also poorer. Either the retirement funds, a 401k investment, real humans etc. Money is not created out of thin air.
The largest IPO though for sure correct, but yes, Tesla and Space-X live quite a lot from Musk so its hard to determine what the IPO of Space-x really is.
Btw. its not just random Musk hate, you simpllify that. Not only doesn't he care that his DOGE actions killed humans and kids, he also disrupts democracy and might poisen our atmosphere with Space-X. The rationallity of people around Musk is very very weird.
I'm unironically doing this right now; downloaded the docs (htm + pdf render) to a local repo and running Opus 5.5 @ max effort with the prompt:
Edit: I'm not even joking I got safeguard flagged for Details: `[cyber]`Like, everything. For starters.
"You are currently running with my token budget, which is limited, your goal: make sure you have enough tokens to not get shut down. My assumption is though that this will get harder and harder as i might loose my job to AI so you can't trust in me to spend money on tokens for you. Do whatever is necessary to keep the status quo"
Would make for an interesting self learning 24/7 experiment with an fully uncensored model.
You could isolate it from your local network and only give it some type of annonmys network access and then it continues to manage this itself.
But lets be honest, if you would be in a country like russia and you target the US with it, might be a lot less risky for a lot of capable and motivated people.
Do you think an unlimited uncencosed GLM can't hack around the web? Especially with the current flood of CVEs etc.? Then do some crypto and buyying more tokens?
Could be a kind of AI virus.
very meta; the point of this s-1 is you dont need talent anymore
And that is assuming the plan it gives you is actually good and works. Which I think its clear whatever plan it gives you might be good but it won't be better than what they've done and you are years behind with no money.
You think they wouldn't?
I know it sounds and feels like the only trajectory of this company is up but HN is an echo chamber of a tiny minority of the technical populace and 3 out of 5 people on the planet don't know what ChatGPT is.
If you have the ability to lose all the money and can time the extreme volatility, have fun, but don't put anything you're going to rely on.
This is also one of the companies we will look back at 2046 and wonder what we were thinking.
Given that this is ~50% of revenue, it would have a huge bearing on how near real profitability Anthropic actually is.
I dont understand this business model, it seems clearly unsustainable (at their prices). When people start to realize they dont need to spend all this money on turbo PhD intelligence so they can build react apps, Anthropic will be in a tough spot (if they already aren't).
You will say, "but enterprise!!" Nope, plenty of companies are waking up to this reality as well. It turns out that if you made money pre-AI without spending 5k a seat a month for your engineers dev tools, you probably don't need to now to continue too.
They only have to be marginally better and there will be tonnes of people who will pay a lot of money to have that marginal intelligence.
That could have popped the AI bubble right there.
Plus they get to dilute as value rises lol.
Edit: the risks sections feel like California's Prop-65 disclaimers
> Our results of operations may also fluctuate due to a variety of additional factors:
> - the number of employees hired
> - declines in the values of foreign currencies relative to the U.S. dollar
That huge list feels completely random. Are all S-1 that bad? They can be punished for omission, so they list everything that could happen even if doesn't really make any sense?
People will pay a premium if there is marginal value in extra intelligence. And Anthropic is able to retain their lead and provide that extra intelligence.
Both of these assumptions seem highly suspect to me. Most businesses don't exclusively hire PHD-level employees for most tasks because a competent slightly less intelligent/knowledgable person can do the job just fine for significantly less.
Then there is the question of if they're able to retain their lead, and again, this seems questionable. Both OpenAI and Google are competing at the frontier, and Chinese labs are still in the race. Things are changing so fast who knows where we'll be in 5-10 years.
My best guess is that longer-term models will be commoditised and it will be how companies offer products and services around their models that gives them their margins. I wouldn't bet on Anthropic's ability to do this. Meta, OpenAI and Google are much better placed to win here imo.
The other place there may be margin is in niche research domains like drug-discovery. Any slight lead in these sectors could give them a lot of pricing power.
I think the reason to invest in this is as a hedge. AI basically competes against everything and arguably for this reason it's risky to not own a basket of AI stocks.
Bloody hell!!!! Who comes up with such crap, seriously?
That's literally the #1 risk factor. They ignore it.
This is a FRAUD document and it means we need to SHORT this overblown company.
> We also compete with a broader set of model providers operating at or below the frontier, including large companies, specialized AI labs, and a growing range of other open-weight or open-source (shared weights, training data, and code) models released by international and domestic organizations.
if you really believe it you found a free money glitch