Episode summary
‘Big Short’ trader Steve Eisman: It's very clear Anthropic's revenue growth has slowed dramatically
Summary
Steve Eisman argued that the loud warnings from AI leaders about existential risk are not credible and are really a push to shape regulation. He said he thinks the executives are trying to create hysteria so the industry gets regulated in a way that keeps Chinese open-weight models out, and he described current AI as “not sentient” and more like a sophisticated word-retrieval system than anything close to human intelligence.
He pushed back on the idea that the major AI founders are suddenly being opportunistic, saying their public warnings go back years. Still, he said he does not believe their rhetoric and thinks they have all “read too much science fiction.” In his view, the bigger market issue is that the leading AI companies have no moat and could end up in a race to the bottom.
On Anthropic, Eisman said the prospectus appears more negative than many investors realize because it lacks third-quarter numbers. He said the available data make it “very, very clear” that Anthropic’s revenue growth slowed sharply after the end of token maxing, and he cited a drop in sequential revenue growth from 147% in the second quarter to 45% in the third quarter, alongside a decline in efficiency.
Key takeaways
- Steve Eisman said AI safety warnings from leaders like Dario and Sam are aimed at creating hysteria and shaping regulation, not reflecting genuine concern.
- He argued current AI is not sentient and is better understood as a complicated word-retrieval model.
- Eisman said Anthropic and OpenAI have no real moat, raising the risk of a race to the bottom across the AI stack.
- He said Anthropic’s prospectus is more negative than investors think because third-quarter numbers appear to show a sharp slowdown in revenue growth.
- Eisman pointed to Anthropic’s sequential revenue growth falling from 147% in the second quarter to 45% in the third quarter.
Notable quotes
“I THINK THE WORD I ACTUALLY USED WAS STRONGER THAN LIE, BUT CAN'T SAY IT ON TELEVISION”
“I JUST THINK THAT WHAT THEY'RE TRYING TO DO IS CREATE HYSTERIA SO THAT THE INDUSTRY WILL BECOME REGULATED”
“IT'S NOT SENTIENT”
“IT'S ACTUALLY THE WAY I WOULD PUT IT IS IT'S A IT'S A VERY COMPLICATED WORD RETRIEVAL MODEL”
“I THINK THESE GUYS HAVE ALL READ TOO MUCH SCIENCE FICTION”
How the episode opens
Well you don't need a percentage of it. You just got to do, do the do the math. — I have to do the math first, but then I have to split.
You do. It's still hundreds of points taken off off the dow and the dow still up. Joining us now Steve Eisman, host of the real eisman playbook podcast and former senior portfolio manager at neuberger berman.
And I don't want you to sugarcoat that, sugarcoat what your thoughts are here. Tell us the truth. You're saying that what dario and sam say about about concerns for humanity, plain and simple.
It's just a lie. And it's all self-serving. And it has to do with with the emergence of open models.
And there's no moat for them.
I think the word I actually used was stronger than lie, but can't say it on television. Yeah, but. Look, I just think that what they're trying to do is create hysteria so that the industry will become regulated. And they think then that they can manipulate the regulation to keep the chinese open weight models out of the country.
And you think that that ai, at this point in its evolution, is that like the stage of a firecracker and not a nuclear, not a hydrogen bomb? In other words, you think it's not sentient at this point. — It's not sentient.
At all. Glorified search.
It's actually the way I would put it is it's a it's a very complicated word retrieval model, meaning you ask it a question like, I'll make up a question, why did the roman empire fail? And so it goes through the entire chain and then predicts what the next word would be. And then it goes back to the beginning of the chain, works all its way back again, takes the word that it just created and predicts the next word that that in a way, is what ai is.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai boom
- ai safety concerns
- anthropic ipo
- anthropic revenue growth slowdown
- market trends
Companies and tickers mentioned
- Anthropic
Watch the episode
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from CNBC Squawk Box
-
Rep. John Moolenaar on investigating Nvidia's Jensen Huang: CCP influence raises multiple concerns -
Jensen Huang's affair has gone from a personal scandal to a corporate governance issue: Paul Argenti -
Experts react to the October CPI report -
Fairlead's Katie Stockton on market selloff: Don't sell weakness, sell the strength -
Ron Baron on his Musk investments: Can make 'hundreds of billions of dollars' over the next 10 years -
Tech selloff looks like a normal correction, says JPMorgan's Jack Caffrey
CNBC Squawk Box, this episode and its artwork belong to their owners. Octagon is not affiliated with the show; this page is an independent summary.
