Title: 'Big Short' investor Steve Eisman on AI: The companies are trying to manufacture a crisis Show: CNBC Squawk Box (CNBC Television clip) Guest: Steve Eisman (host of The Real Eisman Playbook; former senior portfolio manager, Neuberger Berman) Date: 2026-09-17 URL: https://youtu.be/4qV5WWgFTS8 Length: 6:56 Note: CNBC closed captions pasted by Stephen (ALL CAPS) converted to sentence case; no words changed, added or reordered apart from remove-only filler/stutter cleanup; every (mm:ss) cue kept; ">>" speaker-change markers kept. Hosts (Andrew Ross Sorkin, Joe Kernen, Becky Quick) are not labelled by the captions, so host lines are not attributed. The opening line 00:00 is the tail of the preceding Bank of America segment. TRUNCATED: the pasted captions end mid-sentence at 06:54 ("postpone your IPO,"); the remaining ~2 seconds are not captured. Garbles left intact here and corrected only in the analysis pages: "token mixing" = token maxxing (the heavy-token-consumption phase Eisman has said "ended" in Jun/Jul) · "Jensen" = Jensen Huang (NVIDIA) · "because face would have sued" = [Hugging] Face would have sued. =====
00:00 Strong. Earlier this week, Bank of America CEO Brian Moynihan said a decline in fixed income trading would likely translate to what he called, quote, relatively flat trading revenue for the quarter. >> All right. Joining us now is Steve Eisman, a host of The Real Eisman Playbook podcast and a former senior portfolio manager at Neuberger Berman.
00:18 And if you've watched Steve's appearances, we always we like to hear things that are on both sides. And a lot of times Steve will go, oh, no, I don't really feel that excitable or worried or you're very measured most of the time. And the reason most of the time saying that is about AI, I think it's the same kind of thing.
00:45 You don't think that these Terminator style fears, you think they're maybe overstated a little bit at this point. I don't know, though, if we ever predicted or expected to be saying things like, geez, another company is saying our AI bots are acting up again. There was those little monsters.
01:08 I mean, it is weird to watch companies say, we don't know. We didn't know what to expect with what we've created, nothing to worry about. >> I think this is all nonsense. >> You think it's all nonsense? >> All nonsense? I think that there's something else completely going on here. I think the idea that AI, first of all, there's no evidence at all that AI has achieved artificial general intelligence at all.
01:32 And most of the evidence seems to point that it never will. Or if it does, it's many, many years in the future. But the idea that this whole Terminator thing is nonsense. So I think some completely different is going on. And what I think is happening is that token mixing is over. The open weight models are taking big market share.
01:54 I think these companies are very nervous. They realize that there are no moats around their business whatsoever, and they're trying to manufacture a crisis that will create regulation and that they think they can then manipulate to create the moats to create the duopoly that they want. Wow. That's what I think is going on.
02:13 >> So that really is you really are saying something. >> That's David Sacks, but you. >> And. >> David Sacks, right? >> Yeah. >> Go to dinner and have a great bottle of wine. It sounds. >> Like I think this whole Terminator thing. >> But. >> He's taking a stand is garbage. >> Okay.
02:30 But at the same time you last time you were on, I think you said you're not adding to your AI position. And you had actually taken a little bit. >> I've taken some. >> Off the table. >> Correct. >> But here's where I'll ask you this question. In all the news services, they all got to put out content.
02:45 I understand that, but I saw an article the other day and I tried to figure out where I saw it, but I look at so many of them, it doesn't matter. But the eight bubble signs that have been identified in previous bubbles, they're all present now for AI, and it's just a matter of time. Do you believe that. >> On that one? I think I'm a little bit more nuanced.
03:06 What I would say is that. This is how I sort of think about the chain. >> Sounded like you just described, that we're at some type of tipping point for AI if they're desperate to create a narrative because things aren't going. >> Well, I think we're potentially closer. But I think the issue is that at the end of the day, the entire AI chain, from Nvidia to the hyperscalers to Anthropic and OpenAI all depends on the future health of Anthropic and OpenAI, because there's such a huge percentage of the entire chain. So if something were to
03:39 happen to one of those two companies, then the chain would really fall apart. And between the two, OpenAI is the weaker company. I think that's one reason why they postponed their IPO. So I think that's what you have to focus on. >> That's what you said the last time you were here. >> I still I don't change my mind that quickly.
03:59 >> Yeah. That this is all riding on it is there. You don't think there's any nuance to the idea that maybe you're right and all of this is because they're falling behind and they want to manipulate the regulation, but you don't worry at all that even if those two I mean, maybe the way, because I kind of think of it, maybe the way to stop some of these things is to say you're going to be held liable for anything that your agents do when they break out onto the internet and beyond.
04:27 >> I'm sure they're liable now. I mean, I don't think that you need to pass laws to make them more liable. >> Shocked, though, we know that they broke into Hugging Face the OpenAI agents, right. We know Anthropic has had similar instances. If that were a human that did it, somebody would be suing and there'd be a right.
04:45 >> But in this case, that's not happening in large part. And we never talked about this being the byproduct of it. But the deal that Nvidia struck to buy Hugging Face, I believe, was partially an insurance policy to protect OpenAI. If you want to get truly conspiratorial, that's good. This if you want to get truly conspiratorial about this, yes.
05:06 >> Because face would have sued. >> Well, first of all, the question is what the damage face would have been, right? Unclear how big the damages could have been. Maybe there would be next to no damages on a financial basis. >> Look, all I'll say is the Hugging Face CEO told us on this program that he went to Jensen, and he did it because he was worried and needed deeper pockets.
05:26 After that happened, you could be right on Jensen's part of his right. >> But the other piece of it. There were other bidders for the company. Yeah. And so one of the reasons you might have taken Nvidia or and by the way, one of the reasons Nvidia might have been willing to pay more is to protect Hugging Face and protect Hugging Face.
05:43 You're protecting OpenAI. You're protecting basically all of the entire ecosystem. But can I just ask one separate question? Sure. Is it possible that there's something in the middle here? I'm always a believer. It's never black and white. Is it possible that they genuinely are worried about the safety issues that we're discussing, which, by the way, do seem unto themselves concerning.
06:04 And by the way, there's a clear record of them being concerned about these safety issues literally for the past decade, whether it's Dario or Sam or Elon or whomever. Right? I mean, this is not a new narrative from them with the byproduct of that being what you're describing. Meaning, could you genuinely be interested in safety and the byproduct of having that conversation? So in the back of your mind is, yeah, if we talk about safety, one of the things that could happen is you could get the regulatory capture piece, that and duopoly
06:35 piece that you're talking about, but that is not the initial motive or you're. And by the way, it could be that it doesn't matter what the motive is because that could be the end result. >> I don't know about the motive, but, you know, when you talk about, you know, we should slow down because things are dangerous.
06:54 My response is really postpone your IPO, [TRANSCRIPT TRUNCATED — pasted captions end here]