Dario Amodei: belief

Transcript context

…Speaking of governments getting involved, on December 26, the Tennessee legislature introduced a bill which said, “It would be an offense for a person to knowingly train artificial intelligence to provide emotional support, including through open-ended conversations with a user.” Of course, one of the things that Claude attempts to do is be a thoughtful, knowledgeable friend. In general, it seems like we’re going to have this patchwork of state laws. A lot of the benefits that normal people could experience as a result of AI are going to be curtailed, especially when we get into the kinds of things you discuss in “Machines of Loving Grace”: biological freedom, mental health improvements, et cetera. It seems easy to imagine worlds in which these get Whac-A-Moled away by different laws, whereas bills like this don’t seem to address the actual existential threats that you’re concerned about. I’m curious to understand, in the context of things like this, Anthropic’s position against the federal moratorium on state AI laws. There are many different things going on at once. I think that particular law is dumb. It was clearly made by legislators who just probably had little idea what AI models could do and not do. They’re like, “AI models serving us, that just sounds scary. I don’t want that to happen.” So we’re not in favor of that. But that wasn’t the thing that was being voted on. The thing that was being voted on is: we’re going to ban all state regulation of AI for 10 years with no apparent plan to do any federal regulation of AI, which would take Congress to pass, which is a very high bar. So the idea that we’d ban states from doing anything for 10 years… People said they had a plan for the federal government, but there was no actual proposal on the table. There was no actual attempt. Given the serious dangers that I lay out in “Adolescence of Technology” around things like biological weapons and bioterrorism autonomy risk, and the timelines we’ve been talking about—10 years is an eternity—I think that’s a crazy thing to do. So if that’s the choice, if that’s what you force us to choose, then we’re going to choose not to have that moratorium. I think the benefits of that position exceed the costs, but it’s not a perfect position if that’s the choice. Now, I think the thing that we should do, the thing that I would support, is the federal government should step in, not saying “states you can’t regulate”, but “Here’s what we’re going to do, and states you can’t differ from this.” I think preemption is fine in the sense of saying that the federal government says, “Here is our standard. This applies to everyone. States can’t do something different.” That would be something I would support if it would be done in the right way. But this idea of states, “You can’t do anything and we’re not doing anything either,” that struck us as very much not making sense. I think it will not age well, it is already starting to not age well with all the backlash that you’ve seen. Now, in terms of what we would want, the things we’ve talked about are starting with transparency standards in order to monitor some of these autonomy risks and bioterrorism risks. As the risks become more serious, as we get more evidence for them, then I think we could be more aggressive in some targeted ways and say, “Hey, AI bioterrorism is really a threat. Let’s pass a law that forces people to have classifiers.” I could even imagine… It depends. It depends how serious the threat it ends up being. We don’t know for sure. We need to pursue this in an intellectually honest way where we say that ahead of time, the risk has not emerged yet. But I could certainly imagine, with the pace that things are going at, a world where later this year we say, “Hey, this AI bioterrorism stuff is really serious. We should do something about it. We should put it in a federal standard. If the federal government won’t act, we should put it in a state standard.” I could totally see that. I’m concerned about a world where if you just consider the pace of progress you’re expecting, the life cycle of legislation... The benefits are, as you say because of diffusion lag, slow enough that I really do think this patchwork of state laws, on the current trajectory, would prohibit. I mean if having an emotional chatbot friend is something that freaks people out, then just imagine the kinds of actual benefits from AI we want normal people to be able to experience. From improvements in health and healthspan and improvements in mental health and so forth. Whereas at the same time, it seems like you think the dangers are already on the horizon and I just don’t see that much… It seems like it would be especially injurious to the benefits of AI as compared to the dangers of AI. So that’s maybe where the cost benefit makes less sense to me.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.