白宫组建AI风险评估工作组，Jay Clayton任负责人

New this morning, a White House task force will produce a report assessing the risks posed by AI, what responsibility the US government should have over the technology. That's according to the Wall Street Journal, which says the director of national intelligence, one Jay Clayton, will lead the effort effectively making him the administration's first AI czar. Alright. This comes after several AI and tech CEOs gathered at the White House last week and signed a commitment promising to better self police their AI models.

Here's what treasury secretary Bessent told Axios about AI. I believe that we have to be prepared for every occasion. But on the on the other side, you know, I think that kinda, this alarmism without solutions by some some of the AI community, that's not leadership. Well, and it's not just some in the community. It's their CEOs are basically screaming for regulation, warning of losing control of AI. Well, then they should slow down.

But it... It's, you know, it... I mean, it's kinda kinda like Hannibal Lecter. Stop me before I kill again. Joining us now is Hannibal Lecter was not was not expecting that at the end. Alright. Sorry. We're start there. Joining us now is Bloomberg opinion columnist Gautam Mukunda. Gautam, to David's point, what do you make of the Hannibal Lecter quote? I mean, if Hannibal Lecter came to me and said, stop me before I kill again, I would imagine I would like the police to do something about it and not just expect him to police himself.

Well, there's so many problems with that quote. I mean, first, as a big fan of the movie Silence of the Lambs, Hannibal Lecter would never say stop me before I kill him. Not killing people. Yeah. So so the Trump administration's tendency to get pop culture wrong is one of those mysteries that I I don't quite... We don't don't wanna solve. We won't solve it. We're not gonna solve it here. Yeah. I'm confident in that. So the broader point is just that the...

Like, companies caught in a competitive dynamic, even if they want to do the right thing, force... Face enormous market pressures to do the wrong thing. Right? Like, if you were if you were a company that was doing some... To do... Offering gambling to, you know, like, online gambling, and you didn't wanna serve serve addicted gamblers, you would still face pressures from Wall Street to do it because that's the easiest way to goose your numbers.

And so, of course, you're going to end up doing it. You're gonna end up pushing in that direction. The problem... The regulations of these kinds are actually gifts, not burdens on ethical business people because they give them the leverage to do the things that they know that they should not be doing that short term pressures push them into doing. The argument from Scott Bessent is... I mean, it's on his face when you think about it.

It's kind of absurd. Right? Like, we've seen in the financial markets what happens when you allow the lack of regulation in, you know, 2008. We all lived through what happened there. To say that, oh, companies are just going to do the right thing and stop and pull back when every form of commercial pressure is pushing them the opposite direction. It's just an argument that makes no sense. How do you think about the potential efficacy of this blue ribbon commission, this superintelligence force as it's being described, by the White House?

We had this voluntary framework, this moral agreement as the White House called it between the president and these executives. I noted earlier in the week they spelled United States wrong on it indicating maybe how hastily that document came together. My dyslexic self did not notice that, but David definitely did. Where where does that get us sort of having that kind of an agreement? And and do you have any optimism here that this force...

This task force is going to, you know, come together to give us anything tangible? I mean, I guess they could have picked someone someone worse than Jay Clayton to run it. So there is... That that did give me a little bit of a of a good feeling. And, David, I'm glad that your superintelligence catches typos all the time. That that that that... That's worthwhile for the White House stuff. I mean, look at the people on it.

Right? The... They put they put Jeff's... They put David Sachs on it, and Sachs has been very, very open about the fact that he just wants to push forward. He just is, you know, on the extreme deregulation side on the a... On the industry. It's... He's a sort of the avatar of the people in Silicon Valley who are basically saying, you know, there are no risks. There's money to be made. Let's go forward as fast as possible, which is, of course, the view that president Trump has expressed on many occasions when he's interviewed about this.

You... They may be right or they may be wrong, but this is an administration that has committed itself pretty strongly to the idea that there are no risks and, you know, to quote the president, the only safeguard you need is a high IQ president, and we've got that. So we're in great shape. I hope to be positively surprised by what it comes out with in a hundred and twenty days, but I'm not optimistic. Gautam, I like you so much.

But if you call David superintelligent one more time, I swear I will cut your mic. I I will walk off this set. No. I'm just kidding. Noted. But I actually have a question about the nomenclature there because we have this push in addition to everything else going on with the president where he wants to call AI SI, superintelligence. But this also causes a problem because in the intelligence community, SI is already a thing.

It refers to, military, army intelligence, security, communication security. It's already a thing that exists in the spaces where you're gonna be trying to have these national security discussions. Do we think he can force this adoption of SI, or is the president just trying to make Fetch happen? Mean, I I mainly use the term when I'm making fun of my friends at Anthropic. So at this point, I think... I also note that I'm sure this is purely coincidental that that the .si domain name that when he made this announcement, it turned out that a lot of people had sort of bought those in advance and the value of them went up suddenly.

I'm... I mean, I'm sure that's an accident, but it's funny how it's funny how these things happen. Like, it's it's just the the the extent of the... To which the administration at sometimes goes from, you know, policy to farce. Does anybody care whether we call it artificial intelligence or superintelligence, or is this just something the president has sort of grabbed onto and got a Twitter... You know, and did a Truth Social poll on.

I I I I don't care what we call it. I care a lot about the fact that these companies are doing things where they don't seem to have control over their own instruments. Yeah. And we have not chosen to make any meaningful reforms to try and stop that from happening. That seems like the the issue that that the president should be spending his time on. Let's let's move from, Washington to Silicon Valley. So we're on the precipice of of Anthropic going public, and we saw its perspectives.

And indeed, in there was this warning that this technology could propose a catastrophic or existential risk to to humanity. That was there in the the document itself. Makes me wonder, sir, how you're thinking about the way in which this broader conversation is coloring the way that investors might be looking at this company and others poised to go public here in the months ahead? Well, you know, I guess if if Claude does decide to go rogue and exterminate the human race, our investment returns won't be the first matter.

Yeah. Right. Okay. It's nice that they to close. Let's close that. Yeah. So, clearly, there's some marketing component here. Right? Like, let... Let's not pretend that there isn't. But, you know, saying, hey. I'm building someth