(COI note: I work at OpenAI. These are my personal views, though.) My quick take on the "AI pause debate", framed in terms of two scenarios for how the AI safety community might evolve over the coming years: 1. AI safety becomes the single community that's the most knowledgeable about cutting-edge ML systems. The smartest up-and-coming ML researchers find themselves constantly coming to AI safety spaces, because that's the place to go if you want to nerd out about the models. It feels like the early days of hacker culture. There's a constant flow of ideas and brainstorming in those spaces; the core alignment ideas are standard background knowledge for everyone there. There are hackathons where people build fun demos, and people figuring out ways of using AI to augment their research. Constant interactions with the models allows people to gain really good hands-on intuitions about how they work, which they leverage into doing great research that helps us actually understand them better. When the public ends up demanding regulation, there's a large pool of competent people who are broadly reasonable about the risks, and can slot into the relevant institutions and make them work well. 2. AI safety becomes much more similar to the environmentalist movement. It has broader reach, but alienates a lot of the most competent people in the relevant fields. ML researchers who find themselves in AI safety spaces are told they're "worse than Hitler" (which happened to a friend of mine, actually). People get deontological about AI progress; some hesitate to pay for ChatGPT because it feels like they're contributing to the problem (another true story); others overemphasize the risks of existing models in order to whip up popular support. People are sucked into psychological doom spirals similar to how many environmentalists think about climate change: if you're not depressed then you obviously don't take it seriously enough. Just like environmentalists often block some of the most valuable work on fixing climate change (e.g. nuclear energy, geoengineering, land use reform), safety advocates block some of the most valuable work on alignment (e.g. scalable oversight, interpretability, adversarial training) due to acceleration or misuse concerns. Of course, nobody will say they want to dramatically slow down alignment research, but there will be such high barriers to researchers getting and studying the relevant models that it has similar effects. The regulations that end up being implemented are messy and full of holes, because the movement is more focused on making a big statement than figuring out the details. Obviously I've exaggerated and caricatured these scenarios, but I think there's an important point here. One really good thing about the AI safety movement, until recently, is that the focus on the problem of technical alignment has nudged it away from the second scenario (although it wasn't particularly close to the first scenario either, because the "nerding out" was typically more about decision theory or agent foundations than ML itself). That's changed a bit lately, in part because a bunch of people seem to think that making technical progress on alignment is hopeless. I think this is just not an epistemically reasonable position to take: history is full of cases where people dramatically underestimated the growth of scientific knowledge, and its ability to solve big problems. Either way, I do think public advocacy for strong governance measures can be valuable, but I also think that "pause AI" advocacy runs the risk of pushing us towards scenario 2. Even if you think that's a cost worth paying, I'd urge you to think about ways to get the benefits of the advocacy while reducing that cost and keeping the door open for scenario 1.

(Clarification about my views in the context of the AI pause debate) I'm finding it hard to communicate my views on AI risk. I feel like some people are responding to the general vibe they think I'm giving off rather than the actual content. Other times, it seems like people will focus on a narrow snippet of my comments/post and respond to it without recognizing the context. For example, one person interpreted me as saying that I'm against literally any AI safety regulation. I'm not. For a full disclosure, my views on AI risk can be loosely summarized as follows: * I think AI will probably be very beneficial for humanity. * Nonetheless, I think that there are credible, foreseeable risks from AI that could do vast harm, and we should invest heavily to ensure these outcomes don't happen. * I also don't think technology is uniformly harmless. Plenty of technologies have caused net harm. Factory farming is a giant net harm that might have even made our entire industrial civilization a mistake! * I'm not blindly against regulation. I think all laws can and should be viewed as forms of regulations, and I don't think it's feasible for society to exist without laws. * That said, I'm also not blindly in favor of regulation, even for AI risk. You have to show me that the benefits outweigh the harm * I am generally in favor of thoughtful, targeted AI regulations that align incentives well, and reduce downside risks without completely stifling innovation. * I'm open to extreme regulations and policies if or when an AI catastrophe seems imminent, but I don't think we're in such a world right now. I'm not persuaded by the arguments that people have given for this thesis, such as Eliezer Yudkowsky's AGI ruin post.

One thing the AI Pause Debate Week has made salient to me: there appears to be a mismatch between the kind of slowing that on-the-ground AI policy folks talk about, versus the type that AI policy researchers and technical alignment people talk about. My impression from talking to policy folks who are in or close to government—admittedly a sample of only five or so—is that the main[1] coordination problem for reducing AI x-risk is about ensuring the so-called alignment tax gets paid (i.e., ensuring that all the big labs put some time/money/effort into safety, and that none “defect” by skimping on safety to jump ahead on capabilities). This seems to rest on the assumption that the alignment tax is a coherent notion and that technical alignment people are somewhat on track to pay this tax. On the other hand, my impression is that technical alignment people, and AI policy researchers at EA-oriented orgs,[2] are not at all confident in there being a viable level of time/money/effort that will produce safe AGI on the default trajectory. The type of policy action that’s needed, so they seem to say, is much more drastic. For example, something in the vein of global coordination to slow, limit, or outright stop development and deployment of AI capabilities (see, e.g., Larsen’s,[3] Bensinger’s, and Stein-Perlman’s debate week posts), whilst alignment researchers scramble to figure out how on earth to align frontier systems. I’m concerned by this mismatch. It would appear that the game plans of two adjacent clusters of people working to reduce AI x-risk are at odds. (Clearly, this is an oversimplification and there are a range of takes from within both clusters, but my current epistemic status is that this oversimplification gestures at a true and important pattern.) Am I simply mistaken about there being a mismatch here? If not, is anyone working to remedy the situation? Or does anyone have thoughts on how this arose, how it could be rectified, or how to prevent similar mismatches from arising in the future? 1. ^ In the USA, this main is served with a hearty side order of “Let’s make sure China in particular never gets ahead on capabilities.” 2. ^ e.g., Rethink Priorities, AI Impacts 3. ^ I’m aware that Larsen recently crossed over into writing policy bills, but I’m counting him as a technical person on account of his technical background and his time spent in the Berkeley sphere of technical alignment people. Nonetheless, perhaps crossovers like this are a good omen for policy and technical people getting onto the same page.

I'm fairly disappointed with how much discussion I've seen recently that either doesn't bother to engage with ways in which the poster might be wrong, or only engages with weak versions. It's possible that the "debate" format of the last week has made this worse, though not all of the things I've seen were directly part of that. I think that not engaging at all, and merely presenting one side while saying that's what you're doing, seems better than presenting and responding to counterarguments (but only the weak ones), which still seems better than strawmanning arguments that someone else has presented.

Who tends to be clean? With all the scandals in the last year or two, has anyone looked at which recruitment sources are least likely to produce someone extremely net negative in direct impact or to the community (i.e. a justified scandal)? Maybe this should inform outreach efforts.