This author is out of touch with how much people are spending on Claude tokens. People are spending tens of thousands of dollars per month on Claude tokens.
For example, according to a leaked internal dashboard from Meta, employees used 60 trillion tokens in one month. The top user consumed 281 billion tokens in one month tokens, which cost approximately $1.4 million. With an enterprise discount, Meta probably paid $500 million for these 60 trillion tokens.
I can completely understand how Anthropic is projecting $3.5 billion/month in revenue.
True, but irrelevant. Assuming Facebook is paying its bills that employees are incurring, how much money other people aren't making isn't relevant to the question of is someone/something paying Anthropic in real actual today dollars, which then Anthropic will (well, in Q2 when they start paying SpaceX) pay their bill with. Which SpaceX can then turn around and pay their employees and suppliers and contractors with in order to build Starship N+1 and Colossus N+1, and to send Falcons into space with Starlinks on them.
Meta also took on an additional $100B in debt recently to push AI and fired 4000 employees. The economics just don't add up at Meta, so they're a bad example.
So what? Meta is still paying Anthropic hundreds of millions of dollars per month. The author of the submitted article is accusing Anthropic of fabricating its revenue numbers. The fact that Meta may make questionable financial choices has no bearing on the validity of Anthropic's revenue.
The author is doing everyone a disservice by not mentioning that Anthropic has power users spending 5 figures per month on their APIs, and other corporations spending 10s of millions per month on Claude code. Anthropic is swimming in revenue right now.
> Dario Amodei and Elon Musk worked out a sweetheart deal, which they - framed as a “ramp-up,” - that allowed Anthropic to artificially depress its costs. I also question how much of a ramp-up there really was, or what Anthropic’s actual compute constraints were, because it immediately loosened rate limits for Claude subscribers on announcing the deal, meaning that it immediately started having higher inference costs, which…somehow led to it making a higher profit? Or did Musk — as literally described in its S-1 — have SpaceX charge Anthropic less for two specific months to make the numbers look better?
Given xAI is struggling and seemingly that Musk has sold compute capacity to Anthropic, what's the play here? To screw with OpenAI's IPO?
> Let me speak directly and with more empathy than usual: if you want Anthropic to win, you should be just as skeptical of these numbers as I am. You should want to smash my face in the tarmac with the most crystal-clear, impossible-to-argue with numbers, bereft of asterisks or discounts from suppliers or obfuscated accounting metrics.
that feels like the opposite of "empathy" to me. why would i, a reader of better offline, care about a corporation "winning" something, whatever it is? he really is becoming more like idk bill o'reilly or rush limbaugh with every successive "successful" post to me. maybe they should hire him at the new infowars. if nothing else, he knows how to do pr
There’s a pretty clear example of Zitron being dishonest here:
> I also severely doubt that Anthropic managed to make the cost of running its services profitable in the space of six months.
> [Per The Information in January], Anthropic missed on its gross margin projections, saying that its inference costs were 23% higher than the company had anticipated.
> How did Anthropic, which faced a massive influx of new business to the point that Anthropic was forced to buy more compute from Elon Musk, magically become profitable? Other than that discount, of course.
If you follow the link to The Information, you’ll see that it’s a paid article with the headline “Anthropic Lowers Gross Margin Projection as Revenue Skyrockets”. But what happens when you actually read the article?
> Anthropic last month projected it would generate a 40% gross profit margin from selling AI to businesses and application developers in 2025, according to two people with knowledge of its financials. That margin was 10 percentage points lower than its earlier optimistic expectations, though it’s still a big improvement from the year before.
So, according to Zitron’s own source, Anthropic are actually earning 40% gross profit margin on inference, and that is a dramatic jump upwards! This totally contradicts his position that it’s an implausible “swindle” for Anthropic to claim profitability. He’s counting on the fact that most of his readers don’t subscribe to The Information and will only see the headline, or that they will just see a citation and trust that it backs him up without checking.
This author is out of touch with how much people are spending on Claude tokens. People are spending tens of thousands of dollars per month on Claude tokens.
For example, according to a leaked internal dashboard from Meta, employees used 60 trillion tokens in one month. The top user consumed 281 billion tokens in one month tokens, which cost approximately $1.4 million. With an enterprise discount, Meta probably paid $500 million for these 60 trillion tokens.
I can completely understand how Anthropic is projecting $3.5 billion/month in revenue.
I think you are out of touch with people in general.
Most people don't even make tens of thousands of dollars per month.
True, but irrelevant. Assuming Facebook is paying its bills that employees are incurring, how much money other people aren't making isn't relevant to the question of is someone/something paying Anthropic in real actual today dollars, which then Anthropic will (well, in Q2 when they start paying SpaceX) pay their bill with. Which SpaceX can then turn around and pay their employees and suppliers and contractors with in order to build Starship N+1 and Colossus N+1, and to send Falcons into space with Starlinks on them.
Meta also took on an additional $100B in debt recently to push AI and fired 4000 employees. The economics just don't add up at Meta, so they're a bad example.
So what? Meta is still paying Anthropic hundreds of millions of dollars per month. The author of the submitted article is accusing Anthropic of fabricating its revenue numbers. The fact that Meta may make questionable financial choices has no bearing on the validity of Anthropic's revenue.
The author is doing everyone a disservice by not mentioning that Anthropic has power users spending 5 figures per month on their APIs, and other corporations spending 10s of millions per month on Claude code. Anthropic is swimming in revenue right now.
> Dario Amodei and Elon Musk worked out a sweetheart deal, which they - framed as a “ramp-up,” - that allowed Anthropic to artificially depress its costs. I also question how much of a ramp-up there really was, or what Anthropic’s actual compute constraints were, because it immediately loosened rate limits for Claude subscribers on announcing the deal, meaning that it immediately started having higher inference costs, which…somehow led to it making a higher profit? Or did Musk — as literally described in its S-1 — have SpaceX charge Anthropic less for two specific months to make the numbers look better?
Given xAI is struggling and seemingly that Musk has sold compute capacity to Anthropic, what's the play here? To screw with OpenAI's IPO?
> Let me speak directly and with more empathy than usual: if you want Anthropic to win, you should be just as skeptical of these numbers as I am. You should want to smash my face in the tarmac with the most crystal-clear, impossible-to-argue with numbers, bereft of asterisks or discounts from suppliers or obfuscated accounting metrics.
Love Ed or Hate Ed, he has a point.
> To screw with OpenAI's IPO?
That's actually a very interesting idea. At least it could be a cherry on a cake given he lost the lawsuit with OpenAI.
that feels like the opposite of "empathy" to me. why would i, a reader of better offline, care about a corporation "winning" something, whatever it is? he really is becoming more like idk bill o'reilly or rush limbaugh with every successive "successful" post to me. maybe they should hire him at the new infowars. if nothing else, he knows how to do pr
This seems related: https://news.ycombinator.com/item?id=48240156
I bet that in 10 years they will name the next financial (and perhaps economic) crisis "the subsidized bubble"
Until they find someone else paying the costs it will not pop.
Nothing wrong with billionaires and VC burning their money.
There’s a pretty clear example of Zitron being dishonest here:
> I also severely doubt that Anthropic managed to make the cost of running its services profitable in the space of six months.
> [Per The Information in January], Anthropic missed on its gross margin projections, saying that its inference costs were 23% higher than the company had anticipated.
> How did Anthropic, which faced a massive influx of new business to the point that Anthropic was forced to buy more compute from Elon Musk, magically become profitable? Other than that discount, of course.
If you follow the link to The Information, you’ll see that it’s a paid article with the headline “Anthropic Lowers Gross Margin Projection as Revenue Skyrockets”. But what happens when you actually read the article?
> Anthropic last month projected it would generate a 40% gross profit margin from selling AI to businesses and application developers in 2025, according to two people with knowledge of its financials. That margin was 10 percentage points lower than its earlier optimistic expectations, though it’s still a big improvement from the year before.
— https://archive.is/aKFYZ
So, according to Zitron’s own source, Anthropic are actually earning 40% gross profit margin on inference, and that is a dramatic jump upwards! This totally contradicts his position that it’s an implausible “swindle” for Anthropic to claim profitability. He’s counting on the fact that most of his readers don’t subscribe to The Information and will only see the headline, or that they will just see a citation and trust that it backs him up without checking.
It really appears to be a naked pump and dump scheme.