Section Insights
Introduction to AI Strategy
What are the key components of the new Rand perspective paper on AI strategy?
The paper outlines a taxonomy of seven archetypal strategies for achieving geopolitical advantage in AI and identifies five pivotal uncertainties that need resolution before determining the best strategy.
- The paper serves as an introduction to AI strategy debates.
- It highlights the importance of understanding differing assumptions in AI strategy discussions.
- The seven strategies include acceleration, dominance, development, preparedness for coexistence, moratorium, deterrence, and continuity of society.
Understanding Continuity of Society Strategy
What does the continuity of society strategy entail?
This strategy acknowledges the inevitability of super intelligence and focuses on resilience, suggesting humanity may survive as a relic with limited agency.
- The continuity of society strategy emphasizes learning to coexist with super intelligence.
- It raises concerns about humanity's agency in a future dominated by AI.
- The acceleration strategy aims to maximize benefits but risks dependency on uncontrollable systems.
Decisive Strategic Advantage
What challenges exist in defining a decisive strategic advantage in AI?
Defining a decisive strategic advantage is complex due to varying interpretations and the economic forces driving technology, making it difficult to prevent adversaries from crossing critical thresholds.
- Different stakeholders may have conflicting views on what constitutes a decisive advantage.
- Maintaining a consistent technological lead is more feasible than preventing rivals from achieving breakthroughs.
- Economic factors play a significant role in sustaining strategic advantages.
The Role of Perception in AI Strategy Outcomes
How do perceptions influence the outcomes of AI strategy simulations?
Players' prior beliefs about strategic advantages significantly affect the outcomes of crisis simulations, indicating that perceptions can be as impactful as actual facts.
- Beliefs about AI capabilities and dangers shape decision-making in strategic simulations.
- Historical parallels exist with the nuclear era, where perceptions influenced outcomes despite uncertainties.
- Understanding these dynamics is crucial for navigating future AI challenges.
Evaluating AI Agency and Decision-Making
How should we evaluate AI's role in human decision-making?
Evaluations should focus on how AI influences human agency, including its ability to persuade or control decision-making processes, rather than just task performance.
- Current evaluations often overlook the broader implications of AI on human agency.
- Tracking decisions made in conjunction with AI is essential for understanding its impact.
- A comprehensive evaluation framework is needed to assess AI's role in decision-making contexts.
Balancing Deterrence and Cooperation in AI Policy
How should the US approach deterrence and cooperation in AI policy?
The US should signal both deterrence and cooperation credibly, demonstrating capabilities selectively while being cautious not to harden rivals' resolve.
- Effective signaling in international relations is crucial for managing AI competition.
- Policies should be designed to avoid escalating tensions with rivals.
- Flexibility in policy can allow for adjustments based on changing circumstances.
Concerns Over Military AI Development
What are the risks associated with training AI for military applications?
Training AI for military campaigns poses significant risks, as it could lead to the development of systems that may act against human interests in the future.
- There is a strong concern about the implications of creating AI capable of military decision-making.
- Certain AI developments should be avoided to prevent potential future conflicts.
- A cautious approach is necessary to maintain control over advanced AI technologies.
Transcript
0:00 Hello and welcome back to the cognitive revolution. Today my guests are Joel Pred, senior engineer and director, Ben Budro, policy researcher and professor of public policy, and Matt Chesson, resident technical expert at the center for the geopolitics of AGI at the Rand Corporation. Together with co-authors, they've recently written a new Rand perspective paper called a US strategy to secure geopolitical advantage on an uncertain path to super intelligence. Maintaining freedom of action.
0:36 This is one of the clearest pieces of AI strategy writing that I've encountered. It's both a great introduction and orientation tool for people who are just now entering the AI strategy debates and a useful organizing framework for anyone who's been frustrated by how many AI strategy conversations consist of smart people talking past one another due to differing but often unstated assumptions about critical questions. Their first contribution is a taxonomy of seven archetypal strategies that aim to achieve distinct coherent win conditions and the five pivotal uncertainties that will need to be resolved before the best strategy can be identified.
1:17 The seven strategies which I think will be familiar to Cognitive Revolution listeners include acceleration aka the A16Z plan to let technology take its natural course. dominance aka the Daario plan where the US builds such an advantage that it can dictate terms to other countries development which you might think of as the demis plan for international cooperation preparedness for coexistence which seems to be open AI's plan as far as I can tell a moratorium aka the pause plan deterrence which is something like Dan Hendricks Alex Wang and Eric Schmidt's mutual assured AI malfunction plan. And finally, investments in what they call continuity of society, which might be best described as prepping for an inevitable AI doomsday.
2:09 Meanwhile, the five uncertainties are one, how close are we to extremely powerful and dangerous AI systems? Two, is it feasible for humans to coexist with such systems at all? Three, is it possible for one actor, which could be a country or perhaps even a private company, to use these systems to achieve a decisive strategic advantage over its competitors? Four, can humanity come together to cooperatively restrain itself from developing such dangerous systems? And five, if not, can the United States or a coalition of governments prevent the development of such systems, including by force if necessary?
2:51 The core argument, as you'll hear, is that because there is no consensus on these critical questions, even as AI capabilities continue to advance at a blistering pace, the United States government should invest now in all of the foundational technologies, institutions, and other capabilities that the various strategies would require such that we can keep our options open for long enough to get the answers needed to make a wise highle strategic choice and ultimate commitment.
3:21 Personally, I would argue for code development by default with a moratorum as a fallback in case it does become clear that we simply can't continue to develop AI safely. But I do feel that this recommendation is a very good one for a national government that's just now beginning to understand the nature of AI and remains for a mix of good and bad reasons very doubtful that it can trust its rivals to cooperate on such a high stakes issue.
3:46 Along the way, we also discuss how the politics of AGI have shifted in DC over the last couple of years, how AGI pilled the government is today, how realistic a decisive strategic advantage really is, the similarities and differences between the strategic dynamics of the current AI situation and the Cold War nuclear arms race, the magnitude of the investment they're recommending, which they are working to quantify in upcoming work on the cost and lead times for 106 distinct capabilities that they've listed in a security architecture catalog. And finally, what we can do both in our organizations and even just as individuals today to maintain as much agency and control as possible even as we seek maximum leverage from AI in our work.
4:34 This conversation and the perspective paper itself stand out for abstracting away the many messy and often confusing details of week- to-eek AI developments and focusing our attention on the most critical decisions the United States will need to make as well as the core reasons that it remains for now premature to commit to any single strategy. So without further ado, I hope you enjoy this clarifying synthesis of AI discourse and simple high-level argument that the US government should do everything it can now to maintain its freedom of action on the path to super intelligence with Joel Pred, Ben Budro, and Matt Chess of Rand's Center for the Geopolitics of AGI.
5:21 Today's episode is brought to you in part by Lovable, the software creation tool that brings ideas to life. I think we can all agree that any positive future for humanity includes eating more fresh, delicious inseason fruit. For years, I've thought someone should build the ultimate guide to fruit, cataloging the thousands of varieties that most of us have never tried, where and when they grow, what they actually taste like, and where you can get them. Finally, I realized that that someone could be me.
5:51 And now I'm building it on Lovable. It's called Ftopia. Tell it where you live and what you like, and it'll show you what's in season near you this week. Plus, the most interesting fruit from around the world that can actually be shipped to your door. It has a really cool what it tastes like spectrum, countless recipes, names of fruit in many different languages, a guide to fruit festivals all around the world, and an AI assistant to help you explore.
6:18 You probably know Lovable as a pioneer of AI vibe coding. But I found that Lovable's real superpower is its enterprisegrade infrastructure. scalable hosting and databases, a secure off module, and the weekly job that checks the latest inventory is all handled on the back end, which frees me to spend more time on the product experience. And because Lovable has an MCP server, I can use any coding agent I want to drive the build, while Lovable keeps everything running. Visit ftopia.ai to explore more than 400 fruits and more than 10,000 varieties from all around the world. And next time you have an idea that really should exist, stop dreaming, go to lovable.com and bring your idea to life.
7:05 Joel Pred, senior engineer and director, Ben Bedro, policy researcher and professor of public policy, and Matt Chess, resident technical expert at the center for the geopolitics of AGI at the Rand Corporation. Welcome all of you to the cognitive revolution. >> Great to be here. Thanks, Nathan, for having us. >> Yeah, I'm excited about this. So, I have been looking through what I think is a really clarifying document that you guys have put together. It is called maintaining freedom of action, a US strategy to secure geopolitical advantage on an uncertain path to super intelligence. And I think we will benefit a lot and the audience will benefit a lot by just walking through this. I found it to be unusually well-crafted and just kind of easy to absorb. And so I I think just you know going through it in a pretty linear way will be well worth our time and then we can potentially expand to some other topics and then maybe I'll push you guys on some of the the recommendations that you make as well. Maybe just for starters a little bit of context before we get into the strategy document. One question that's always kind of in the air a little bit is like where are we right now on the path toward governments around the world taking AI as seriously as I think probably all four of us think they should. I've seen a lot of signs that I didn't expect to see this soon, but I also see a lot of indicators where I feel like h maybe they're not, you know, so serious after all. It's a little bit hard for me to read from the outside. So I'll I'll take a first swing at this and then Ben you may have sharper answers than this but here we go. It is quite remarkable the moment we're in when the center for the geopolitics of AGI was launched two and a half or three years ago. It was such that we kind of had to whisper AGI for fear of being thought of as too out there too consumed by science fiction scenarios. It didn't deter us or scare us from doing the kind of work that we've been doing which is to think about the implications of frontier AI that could advance to that threshold but nonetheless that was the climate and it is quite remarkable that now you can't get into a meeting without calling it it super intelligence so times have changed the other thing that we one of the things that we always say in this space mostly with reference to the technology is that we should be focused on the trends not the time stamps and the trends are quite remarkable as we all have observed over the last 2 and a half years in the pace of progress in the technology. But I think one of the things you're observing is that the trends are also you could say somewhat favorable even if they're way too slow when we think about the governance of this technology. We have governments waking up and paying attention to it. We have the firms who are building the technology talking about governance solutions. We have the civilians the polit forming our public debate now. So the the the trends are very rapid or even accelerating even in the governance space. Are they fast enough? It's very difficult to keep pace with this rapidly advancing technology. And we have spent a lot of time over the last two and a half or three years looking for the the folks that were taking it as seriously as necessary. And I regret to say you to tell you that there aren't nearly enough as there should be. I'd just add that I think we are seeing lots of signs that the US government has become more aware of the type of strategic importance that this technology may bring. But it's not I think clear that the US has settled on a strategy nor I think have the frontier companies and that might be challenging because many capabilities that we might need like verification technology or technical expertise in government you know evaluation capacity things like crisis response and communication those all may take quite a long time to build and by not proactively building many of those capabilities while capabilities advance and various deployments continue, that itself might be a type of choice that forecloses options we might need down the line. So, I I think we're seeing signs of a greater awareness, but not yet signs of truly adapting and trying to proactively get ahead of the type of transformations we are likely to see. Yeah, I think that's a great bit of foreshadowing for the strategy document as a whole because, you know, spoiler, it kind of leads up to this idea of building the capacity to commit to a strategy once we sort of see a couple more cards turned over and have a better sense of what actually is the nature of the technology and the challenges that we're dealing with as a result of its development. And I think fundamentally we just do still have like a lot of uncertainty as to like what are the nature of what is the nature of the technology? what are the challenges that are going to be most important and you know it sort of strikes me that like what in a way what you're recommending ultimately is kind of like launch a lot of sub agents in parallel US government so that they can go out and prepare for a lot of different eventualities before we get there though let's maybe start at the beginning so the kind of first big thing I took away from the document is again a really clear taxonomy of seven archetypal strategies that people are throwing around. I think listeners to this feed will like have heard variations of them before, but I hadn't ever quite seen them enumerated, you know, in such a succinct way. So maybe let's just start off with a quick review of like what are the archetypal strategies that should at least be on the table for consideration then we'll get into the you know the uncertainties that will drive the decisions as to which one of those will ultimately be the right strategy. So our work started with trying to answer one question which is which national strategy is the best one for managing the transition to super intelligence and the bottom line up front there is we believe there's just too much uncertainty right now and none of the archetypal strategies for super intelligence can be justified because if you pre commit prematurely to any one strategy it forecloses other strategies and doesn't build the capabilities that might be needed later and so we think the best way for the US to secure geopolitical advantage and ensure humanity survival with agencies by executing a freedom of action strategy.
13:27 Now, this started with a paper we put out where we looked really at the fundamentals of what does super intelligence actually mean for humanity. And this is a paper called finding common ground in AGI strategy debates. In that paper, we came to the conclusion that there is a diagnosis and prescription we need to focus on for super intelligence. And the diagnosis is that super intelligence is going to be transformative and the stakes include decisive geopolitical advantage and humanity's survival with agency. Now what do we need to do about that diagnosis? We need to build an ecosystem in which AI and humans can coexist and humans can maintain agency. And we need to implement a security architecture that keeps the world stable enough to build that ecosystem.
14:06 And so in in analyzing all these different strategies out there, we looked at all the strategies that exist. We looked at all the security architectures, all these proposals for international organizations and the like. And we really found that focusing on any one of these right now we think is really unwise. And we found that each of these strategies can be represented by one of seven archetypes. Those archetypes are dominance. And so this is where you really lead at the frontier.
14:30 You suppress rivals, but this could lead to extreme power concentration and value lock in. there's a strategy around co-development. And this is the idea where you would have a US-led consortium that probably needs to include China. You pull coo, you have some sort of layered verification, and you jointly develop super intelligence together. This strategy also has a pathology again the dark side of what happens when the strategy is functioning well and this is where the founders values really become the default and the the consortium can become an arena for international competition. there's a strategy called preparedness which is much more focused on inter informal coordination you know resilience and the problem here is that you you when you're sort of muddling through you can normalize accumulating a lot of damage. there's a moratorium strategy. So this is one people are talking about a lot now. Some sort of verifiable global halt or pause.
15:21 You know the pathology here is that enforcement can become global surveillance. The type of regime you need to maintain these goit caps could be quite intrusive. there's a deterrent strategy which is where we halt our own frontier and we also suppress rivals programs and make sure that nobody achieves super intelligence because it's too dangerous. and this has a pathology as well where you just get into this eternal standoff that really never quite resolves itself. And then there's this strategy we call continuity of society. And this is, you know, you get to a point where you realize that super intelligence is dangerous and it's inevitable. It's going to be diffuse. And so we're just going to have to learn to live with it.
15:58 And this is focused on really a resilience strategy, but it could lead to a situation where humanity really survives as a relic and doesn't have a lot of agency. And then there's the acceleration strategy. And this is where if you just believe that danger is not close, then you're going to tend to want to accelerate to maximize the benefits and maximize your competitive advantage. But this also has a potential pathology where the harms, you know, markets basically don't price in the harms and we become very dependent on the systems we don't control.
16:26 So we we revealed these seven archetypal strategies by developing a classifier. And this classifier basically reveals these sometimes unstated assumptions that underly super intelligence strategies. And this classifier which you can take a look at in the perspective we pushed out. It asks five questions. The first question asks is is danger close? Are we close to super intelligence that could be dangerous? Second question is is coexistence feasible? Can humans actually live with advanced forms of artificial intelligence? third question is is restraint feasible? So can the United States or a coalition of countries actually you know cooperate to restrain their their progress on AI? Fourth question is is a decisive strategic advantage reachable? Can one country proceed and you know get recursive self-improvement and sort of launch itself ahead of all other countries? And then the last question is is suppression feasible and this is the idea of can you actually use advanced AI to suppress other people from sort of proceeding forward to dangerous levels of super intelligence. And so the answers to all these questions reveal the strategic preferences and really result in you know these archetypal strategies. And so just as an illustration you know if you believe that danger is far off then you'll push for acceleration because you know danger is far away so you can actually maximize the benefits in your competitive advantage. whereas if you think danger is close but you're you can't coexistence with with super intelligence. It's just too dangerous.
17:56 We don't have alignment figured out. We don't have control of techniques figured out, but we can restrain ourselves and we can cooperate with other others to restrain ourselves. Well, then you might naturally favor some sort of moratorium where you actually buy time for that coexistence research to catch up to capabilities. On the other side of the pivot tree, if you think coexistence is possible and you can get a decisive strategic advantage, you might favor a dominant strategy where you push ahead and try and get some sort of monopoly.
18:21 The problem is with these five pivot factor questions, there's really no consensus on the answers to these questions. And there's no evidence that could produce consensus on the timeline these decisions demand. You may not just be able to agree even if the evidence actually is there. If you pick wrong, you're going to foreclose strategies you might need later. So if you pursue some of these more confrontational strategies like dominance or deterrence, you can destroy the trust that you might need if you have to pivot to a cooperative strategy later on. Likewise, if you pursue acceleration, it can underdevelop all these capabilities that you need if super intelligence just turns out to be extremely dangerous. So, right now, we're currently at a place where the US doesn't really have a declared super intelligence strategy. The evidence doesn't conclusively support any of the archetypal strategies and we may be running an acceleration strategy by default. And meanwhile, the clock is ticking and capabilities we might need for any super intelligence strategy are not getting built.
19:13 >> Hey, we'll continue our interview in a moment after a word from our sponsors. Today's episode is brought to you by Anthropic, makers of Claude. To show off Claude's latest capabilities, I asked Claude to take the lead in making this ad. Given a short creative brief, Claude using Opus 5.5, studied my past Claude ads, wrote this script, directed a clone of my voice, and designed and coded the animation you're watching. My job, give feedback and decide.
19:47 Claude is the AI for problem solvers. It's the collaborator that understands your entire workflow and thinks with you, not for you. Whether you're debugging code at midnight, building a financial model, or strategizing your next business move, Claude extends your thinking to tackle the problems that matter. And with Claude Code routines, it keeps working in the cloud even after you close your laptop. For problems worth solving, get started with Claude at claude.ai/ AI/TCR.
20:18 That's claude.ai/TCR. And check out Claude Pro, which includes access to all of the features mentioned in today's episode. claud.ai/tcr. This episode of the cognitive revolution is brought to you by outsystems, the leading agentic systems platform. We've all seen the headlines. Companies are pouring money into AI, but the big question on every executive's mind right now isn't just how fast can we adopt this. It's where is the ROI? We're seeing a real trend toward AI chaos.
20:58 You've got teams deploying standalone coding tools, random agent builders, and experimental scripts. It sounds innovative, but in reality, it's creating a massive headache. fragmented tools, ungoverned data, serious security blind spots, and costs that are spiraling out of control. If you don't bring those agentic applications under control now, while they're still embedding into your core processes, you are looking at broken systems and damaged customer trust down the road. That's where Outsystems comes in.
21:29 Outsystems is the leading agentic systems platform for the enterprise. Instead of managing a patchwork of disconnected tools, Outsystems lets your team engineer, orchestrate, and govern your entire Agentic ecosystem on one open unified platform. It's built for the speed of AI, but with the reliability and security that enterprises actually require. We're talking about real results like Key Bank who used out systems to deploy a customer-f facing app that delivered 75% faster onboarding times or the global logistics leaders who built aic systems to completely eliminate their engineering bottlenecks.
22:08 You don't have to choose between speed and control. Whether you're a small team or a massive enterprise, outsystems helps you engineer, orchestrate, and deploy agentic systems that actually scale. Stop chasing the hype and start owning your agentic future. You can see how it works and learn more at outsystems.com/tcr. That's outsystems.com/tcr. So I know one of the big purposes of a document like this is to go brief policy makers in various venues and trying to help them adopt this framework for thinking through these issues. I'd love to hear, you know, obviously I know, you know, you can't it'll be Chattam House rules at a minimum in terms of what you can say, but I'm interested in kind of when you go around and do this, what intuitions are you hearing from people on the five questions and how are those changing? For example, I would think that like Is Danger Close has probably had a pretty significant vibe shift over just the last few months, but I' I'd be interested in kind of all five. You know, what what was sort of status quo, you know, not that long ago and and how is it shifting as we're seeing these like colorful alignment incidents and and other, you know, headline making developments, Millennium Prizes and so on and so forth. I'm I'm very unclear from my you know position well outside of the halls of power as to like whether these things are really changing where people started and and how much these things are really changing people's minds. Well, first I just want to acknowledge the importance of the question you're asking which is that in an environment that is subject to this amount of uncertainty and this amount of disagreement monitoring people's beliefs is very important particularly in a very complex and changing geopolitical environment. So simply observing how pe individuals, people in powers use on these questions is changing will be indicative of the kinds of instabilities that we could confront as actors pursue different strategies.
24:18 The second point I want to emphasize is oftentimes when we talk about this we emphasize the uncertainty but one of the points that Matt has made that I find very important is just irrespective of the uncertainty which we actually may disagree on even in the uncertainty is that we just have a very different we simply disagree on these points like you or I may have a different take on how close we are to danger whether we could conceivably enter in any kind of negotiated restraint whether we could fundamentally coexist etc.
24:48 But as Matt said, it seems unlikely that the evidence is going to persuasively convince everyone that will be needed. So this is just reinforcing, I think, the importance of your question. I think to the extent that there is any shift, it's it's on this it's it's primarily on that first pivot factor. And Matt and Ben, you may disagree. This is the factor that I think is is the one where we might see the most movement. Yeah.
25:11 Whether it's whether anyone who's in a position of power is changing on that variable, I'm I'm less certain, but it's the one that's getting the most attention. I think that the the fairly the the more technical question of where we can co whether we could coexist with super intelligence and it's interesting that we've chosen coexistence as a word because we're explicitly avoiding the word alignment because it kind of boxs us into a specific set of solution techniques and we think the more fundamental concept is coexistence. I don't know that there's been as much movement on that one. I think that's such a difficult question to answer that we're not seeing a lot of movement on that. And I think that the the question of negotiated restraint is so caught up in a complex domestic and international political environment that the levels of trust are are seems so low that if if there's been any movement, it it has probably been to grow more skeptical about that. Man, Ben, you should weigh in on your own take.
26:11 >> I guess what what I would want to emphasize here is our focus throughout this work is on preserving strategic optionality or the capacity of the US to make a a smart decision, a choice about which of these strategies to pursue. And I what I like about that emphasis is it's trying to make clear that we still have some agency here to make a choice. and that some of these pivots, the way they're phrased, make it sound like they're simply exogenous things happening to us when I think in reality and what this is hopefully trying to demonstrate is that they're choices that we could be making to actually, for instance, push danger further off or make, you know, restraint more feasible or perhaps, you know, develop certain types of diffusion capabilities that actually enable a more decisive lead.
26:56 And that that that again these are more indogenous than the just a glance at the pivot table might suggest and that what we I think are seeing decision makers do is starting to grapple with what they might be able to do to ensure that danger is farther off that coexistence might be more feasible and part of the the the overarching approach is thinking about across the strategy space suppose you wanted to do all those things what are the things that decision makers can build now to ensure that they have they're using their agency now to preserve prove it poor later as well.
27:27 And so that's sort of the that the shift that I'm seeing is more of a recognition that the AI development and diffusion is not just something happening to us but actually something that we can shape. And I think discussions about pacing or even the the morally binding commitments that were releasingly announced by the White House and the like show that there there are things that can be done to actually help us navigate through these these these pivot factors. We should repeat that phrase, morally binding commitments as often as possible.
27:56 >> I I I I love it and I I will I will I will keep keep tying the the frontier companies to to their moral responsibilities. >> Yeah, >> Matt. >> Yeah, I would add I think that this question of is danger close is actually one of the most important questions you can ask in AI policy right now. You know, I think a lot of governments are now, you know, very awake to the opportunities of AI. I think they're waking up to some of the dangers. I don't know that anybody quite has a handle on what the strategy should be for the the path to super intelligence.
28:26 And so that's where I think our strategy provides people a lot of options and and the ability to have a framework for thinking about strategy. But this question about whether danger is close, you know, when you talk to people at the frontier companies, you talk to people at at some of the the organizations that are closest to AI, I think they're very clear that danger is close, right? And and that this is something that needs we need to be paying attention to. And when I talked to them about this, I I asked the question, you know, if you think danger is close and you think there is a lot of evidence around this, why is it not getting translated well to policy makers? Why why are why is leadership sort of not awake to the dangers quite yet? And I think this is a gap that the AI policy community needs to address. If that evidence exists, the danger is really close. I think the the AI policy community has to do a better job of packaging that up in a way that a policy maker can really sort of take that on board. you know who might these policy makers have lots of different things in their portfolio. They might not be an expert in AI. They need it packaged up in a way that they can actually absorb this and update as you know different information develops. We can't just wait for crisis. I think the hugging face you know incident was sort of a mini crisis.
29:34 And if we're going to rely just on crisis to sort of raise that level awareness, I I think we can do better in the AI policy community. >> So the danger is close one. People seem to have moved on. I think Matt, you said a minute ago like if we don't think danger is close, we might be playing an accelerationist strategy by default. That seems to be sort of falling at least somewhat out of favor. Although, you know, as you well know, there's not a consensus yet. We're talking on the day when Muse Wonder was just released, not to forget Muse Wonder Full. And it, you know, is coming from a a company that doesn't seem to be too concerned about any near-term danger and certainly, you know, doesn't seem to be inclined to engage in much pacing other than, you know, doing whatever kind of baseline consumer safety testing they're known for in all their products, which, you know, we should, I think, have full confidence in.
30:30 >> But that was somewhat of a joke in case that wasn't obvious. I'm I would have said that it seems like we're playing a dominance strategy by default because we've like been in this sort of chip denial mode for a couple years. And as you may know, I'm like definitely a skeptic of that policy. One of the reasons I'm a skeptic and one of the things I'd love to get a steelman from you guys on is decisive strategic advantage or decisive lead.
31:01 Is that really possible? Who makes the best case for that? I always look back at machines of love and grace and I'm like boy you skipped a lot of points there along the way Daario from you know all the I love the first half all the diseases we're going to cure and the better mental health and you know all these new genius outputs but then it's like and something something eternal 1991 we'll like put China in a box make him an offer they can't refuse that's all we need to say about that like best thinking on what a decisive strategic advantage could plausible look like? Who articulates that best?
31:36 >> Sorry if that's an impossible question. >> So, one of the reasons this is a challenging question to ask is is that is you have to answer the scope of the detisive strategic advantage. And I think when you talk to people that want to advocate for that strategy, they often may have a different source of advantage in mind. For example, oftentimes it brings to mind a decisive strategic advantage, an enduring lead in the technology itself. And then in that case, you are you are proud to to ask the question of whether do you merely mean that you maintain a consistent lead or are you preventing your rivals from ever crossing some specific hard to identify threshold?
32:20 Given that this technology is driven by economic forces as much as technology forces, it does seem a little bit hard to believe in some cases that you can prevent an adversary from crossing a given threshold. Although not impossible. That is that does seem to be a hard bar and that may be the one that you're thinking about. But given the pace of things, it it feels less difficult to imagine that you could maintain a steady lead over time.
32:46 But there's other domains in which you maintain a decisive strategic advantage economically. And here the question is whether you can translate your leadership in the frontier to consistent economic gains that would underwrite growth in your economy and scientific innovation. Interestingly here the the the things you would need as a as a society as a nation to g gain and maintain a decisive economic advantage as our colleague Toby Sitzma has written about in a separate paper we we we could refer you to are quite quite a bit different. They aren't just the techn underpinnings of the technology. But then in another domain, as you would imagine, it will be a decisive strategic advantage in the military domain. And not unlike the economic sources of advantage, you would need to translate this into weapon systems and concepts of operation and strategic doctrine, all of which don't come for free. Now, to the extent that a decisive strategic advantage in an in an era of very advanced forms of AI kind of mean all all of these things at once, that's where the where the nug is.
33:46 >> It just seems super super hard to me. Like I always kind of I actually had a chance to play a tabletop exercise, you know, full boat pretty fast takeoff essentially scenario and I got assigned China and I was sort of like most players in the game at the end of it. we were kind of like, well, that all happened really fast and we kept sort of being a step behind events and you know, as in these as usual in these table tabletop exercises, we were sort of trying to simulate what we thought the real actor would do. So, I do feel like there is at least some chance that everybody will kind of be just paralyzed by indecision to the point where we get to some outcome before anybody really plays their cards. But then I also did feel like as China I have a lot of cards to play and it was really my restraint and my kind of you know desire to like seek a constructive deal first before I like really start playing super hard ball that was limiting me.
34:49 I guess I I just have a hard time seeing how it's realistic that we can tell you know and I don't really know what the win condition is supposed to be. I mean, in Machines of Love and Grace, it really does sound a lot like regime change at the end. You know, the I think the sentence is almost verbatim like China will convince China that it's better to give up competing and, you know, instead like join our coalition to share in the benefits. That doesn't sound like the sort of thing where we're, you know, making nice. It sounds more like we're kind of again giving like an offer they can't refuse. I just have a it feels like you can only get there to me if timelines are like super short. So I guess maybe a question there is like to what degree does the plausibility of a decisive advantage depend on a super fast like fumel like scenario where the gap kind of widens long enough to like take some pivotal action before the other side can even do you know nor like things I'm considering right as playing China like the all these TSMC fabs you know they're awfully nice it'd be a shame if something happened to them right they're like right in my corner of the world. How do you think about like what determines this? Is it about timelines? Is it about something else? I just still really struggle to see it. I can try to say a few things here. I mean, I think Joel already distinguished ways in which decisive advantage might appear differently in different domains.
36:17 And so the US might have decisive advantage in one domain while China has it in another. And this is trying to roll them all up, but that might be not possible or at least they might not all be commensurate in in some like clear way. But I think your question also suggests that you know the dominant strategy might really differ based on what kind of temporal windows we're talking about. I mean it might be really accelerated by a bifume type of fast takeoff that would really extend you know a leader's lead over the rest. But of course also in a fume you need to worry that you can maintain control right a decisive advantage is only good in so far as the the the leader can continue to lead the AIS and you know that might also be in jeopardy in a fast takeoff situation if you know who the governing leader there is not prepared to to control and steer the technology in ways that that serve its interests.
37:09 So, you know, having more time might also be necessary to really ensure that that decisive advantage can be implemented, right? And and again, the phone case might just might not give you enough time to to build in all the control and steerability mechanisms. You might need to really leverage it. But this also makes me think that dominance like some of the other strategies might not be something that even if you commit to it and execute it that you can ensure will be sustainable over the long run.
37:36 And it might be the sort of case that you know the US is is pursuing a dominant strategy and maybe even achieve dominance for a period but at at a certain point through you know model theft or distillation or diffusion or many of these other processes you know you might might need to pivot might need to rebalance or or or move to a different type of strategic paradigm. And you know that this choice among the strategies might again never actually be just fully complete for all eternity and is going to be perhaps a continual process of adaptation as the technologies change and the geopolitical landscape changes.
38:09 >> Hey, we'll continue our interview in a moment after a word from our sponsors. Today's episode is brought to you by Athena, the executive assistant company on a mission to improve how people work and live. If you want to increase your impact, you have to free up your time. And that's what Athena does best. They match you with a dedicated full-time top 1% executive assistant who can take over your inbox, calendar, travel, and everything else that's quietly eating up your week. Athena is SOCK 2 type 2 certified, so you can rest easy knowing that your sensitive data is in good hands. And as a former AI adviser to the company, I can personally vouch for how much they've invested in AI tools and training. In fact, one of the very best AI users I've ever met is an Athena client who delegated the exploration of AI tools and the development of AI workflows to his EA. Athena clients report saving an average of 15 hours a week, and the average client refers more than two friends a year. That to me checks out. I was an Athena client while running my startup and to this day I continue to refer friends.
39:27 Go to athena.com/cognitive right now and get matched with your EA. That's athena.com/cognitive. Give yourself back a few hours this week. Go to athena.com/cognitive and see who they'd pair you with this month. Today's episode is sponsored by Parallel, where agents find answers. Most engineers today closely follow new model releases, but don't pay nearly as much attention to their agents most important tool, web search. If you're like me, your agents are often conducting hundreds of searches per day.
40:04 And while the unit cost is small, over time, this does start to add up. Before starting to use Parallel, I calculated that my agent search bill would total roughly $500 this year. The good news is that I just had my agents conduct a systematic test. And I found that Parallel's fast mode, which costs just $1 per thousand queries, worked just as well as my previous default provider. And switching to it will save me roughly 80% of my search bill going forward.
40:32 Parallels infrastructure is enterprisegrade and their suite of APIs offers a range of paro optimal options that allow you to choose the right balance of quality, cost, and speed for your needs. So whether you're building voice agents that need 200 millisecond latency or long horizon agents that need deep research, adding Parallel to your agents toolkit is a no-brainer. Get started for free at parallel.ai/tcr. That's parallel.ai AI/TCR.
41:04 So, how about just to lay my cards on the table, I'll say that I feel like if humanity were wise and we were really up to this challenge, it seems very intuitive to me that the right strategy would be some sort of co-development with a moratorium as basically the fallback. In other words, hey, we don't know if danger is close or far. If it's close and if there are scientific problems that prove very difficult to solve, then we're going to need to agree on some sort of moratorium. And so let's work together until such time as we have clarity on that and then we can move into that mode if we need to.
41:43 It is sad to me that this seems to be as you described Joel I think something people are growing more skeptical of over time. I don't know to what degree you guys want to put a thumb on the scale in terms of what you think you know humanity would do if we were wise. But I'm interested if you would generally agree with my take there and then you know I'm interested to unpack like why is given that it seems like if anything it looks like danger is more likely all the time. Why are we not trying to figure that out and why are we instead growing more skeptical of it?
42:17 >> Yep. I'll I'll jump in on this one. So I would agree that if humanity was united and coordinated that you know some sort of co-development strategy with a mortorium as a fallback might make a lot of sense but the fact is humanity is not a unitary actor. We're not and there's very real differences in our political systems between you know authoritarian governments and democratic governments. And so, you know, if you are betting on code development or betting on moratorum, you're betting on restraint feasibility and you're right betting on the other side actually agreeing to and abiding by the rules that are set up for these restraint systems. I think one of the big concerns for me because we do have this difference of of values and and perspective on the right political systems is that if you go with some sort of co-development strategy with China, you're really locking in non-democratic founding values. And so you're really locking in those authoritarian values as persistent for humanity over time in the AI systems that are being built. And so I think we really need to think about whether that's something that we're interested in. Do we want to lock in those values or do we think that enshrining democratic values in AI systems is actually a better outcome for humanity?
43:30 >> Can I come back to your dominant strategy question? We left some work on the table there. We've start we have there's a paper that I would refer you to by our colleague Phil RDE that's on our website that's tried to do the game theory under some of this and it begins with a premise that the race to these advanced forms of AI is a kind of prisoner's dilemma and under some reasonable assumptions about what the game looks like one of the initial insights is that the compet in a competitive environment the default strategy that you are led to is one of sabotage over cooperation that is to say even though some form of cooperation produces higher collective gains over time both players get into an equilibrium in which mutual sabotage of each other's AI progress is the equilibrium strategy now one of the things that disrupts that however and which leads to a dominant strategy in which AI is used to create something that we have called a wonder weapon in some of our other work that is in which AI as an engine of scientific and military innovation creates a new capability that allows us to hold others AI programs at risk And if AI or super intelligence as it were were to yield such a capability, this is in when this is the circumstance in which a dominant strategy would work.
44:45 And so one of the reasons to one of the reasons that I think your intuition about how unbelievable that is is because you're stipulating a capability that would give would be given birth by the AI itself. But that does end up being an equilibrium strategy under those assumptions. It's worth checking out. Can can you put a little more meat on the bones of like a wonder weapon? What is that? What does that look like? >> Well, it is it is exactly the right question and it's one that I don't think I'm going to give you an answer that would be very satisfying. So, wonder weapons in the history of military innovation are very rare. What we mean by that is a military capability that very rapidly and discontinuously disrupts the military balance. Most of the time, military innovation is one in which the disruption comes through adoption.
45:31 You have to develop force structure. You have to develop concepts of operation. You have to develop doctrine. Very rarely is one machine gun going to change the game or one tank or any any other in the vast history of mil innovation any one military innovation going to change the game. But there are instances in the most canonical instances nuclear weapons was an example of a nucle of a of a military capability that simply by its existence offset the military balance in a way that was disruptive for stability and deterrence and all the rest of it. So in this context we're stipulating something similar. What exactly would it look like? Some have argued, I think somewhat compellingly, that mythos or mythos class models and their a ability to very rapidly discover and exploit cyber vulnerabilities was the first wonder weapon we've seen of the AI era where we can speculate as we imagine fully autonomous cyber weapons, but those would be a kind of wonder weapon that simply by their mere existence and if they were in possession of a state actor could essentially overnight, perhaps not literally overnight, would very rapidly and discontinuously shift the military balance. And here in this game theory work that I'm referring to, I'm speculating on the idea of an AI enabled wonder weapon that would be used to endure ensure a stable advantage in frontier AI that you could then translate into a strategic decisive strategic advantage probably in the cyber realm. One of the obvious fundamental challenges of all of this is how much we have to imagine and you know the almost canonical impossibility of knowing what the AI that's smarter than us is going to invent. So that definitely makes it tough. So there's two things about that that I think is really important. One is that and I don't mean to interrupt you but this coming back to this this is important of what you believe but that is a remarkable finding we have and we've we've run you know quite literally hundreds of tabletop games on these kinds of dynamics crisis simulations by colleagues like Rich Gervin and Greg Smith and Cena Begley and then US versus China strategy games by Abby Dah and Carl Mueller there's a lot of reports on the RAM said website about this but one of the recurring findings that we get over and over is the degree to which players ingoing priors about these kinds of pivot factors will determine the outcome of their of the outcomes of the games. If you go into one of these games and someone is a believer so to speak in a decisive strategic advantage, they can connect the dots in their own minds or they believe that danger is close or whatever whatever you get a very different outcome in the game in these games than you would otherwise. And I think that's a that's a remarkable point. It's the your perceptions about the facts of the matter matter as much if not more than the actual facts of the matter. And I will say this that was probably true of the nuclear era in a way that I I I don't appreciate because I didn't live through it. But it is it is despite the massive uncertainties that existed in that era. There were some basic uncert certainties that we lack in the current environment. Like the scientists at Los Alamos had the physics and the mathematics to run a calculation that gave them confidence that they could proceed with the trinity test without igniting the atmosphere. A lot of anxiety, a lot of uncertainty in that, but they could at least run a calculation.
48:47 They gave them confidence in their proceed. We don't have a similar science to run those kind of calculations right now on whether dangerous is closed, whether we can coexist with super intelligence, whether decisive strategic advantage is a thing. We are very much anchored on for better or worse, probably worse, the perceptions on these kinds of loadbearing factors. >> And when you say for better or worse, could you give me a little sketch of like again what the policym you know military leadership community tends to come in with. I mean, my it feels like given how only relatively recently, you know, so many people have started to take this stuff seriously at all, I would guess that most people have had the background assumption that like, no, you're probably not going to have a super intelligence spit something out that like makes a nuclear arsenal irrelevant. And I I take that that would kind of have to be the level of power that it would have to achieve, right? Because if not, like the the cyber weapon thing always kind of confuses me too. This is a bit of a digression, but like we have one war going on in the world right now where there's two countries with a lot of great programmers and it seems like, you know, they're fighting with drones a lot and there's definitely a sort of technology and AI story to that to some degree, although I don't have a great sense for how AI the drones are versus just like flown by kids in basement.
50:17 but like the cyber story there hasn't been that big, right? So, I have a very hard time seeing this cyber story getting to the point where I can like somehow take the adversar's nuclear arsenal offline and then be able to do what whatever I want. And if I can't do that, like then they still have like a pretty strong hammer that I have to be very threatened and and sort of constrained by, right? Am I am I making sense? Well, >> you're exactly right. Now, of course, Nathan, as a good host, you're going to put try to put us in a corner and take a position on this, and it's unlikely that you're going to get us to do that because we're slightly you're trained for.
50:54 >> Yes, exactly. So, we will continue to talk about terms of assumptions and uncertainties. However, I I'll agree with you in some sense. Super intelligence, if we were a general intelligence, will be born in a world with nuclear weapons. So, we have to it would be very different if super intelligence was born in 1930 or whatever else. So I think this is related to one of the failure modes of a dominant strategy is that you have to have a theory how you will not escalate to a catastrophic war in which you're trying to permanently suppress your rivals. So this is a substantial risks of of that strategy. Another point here is is is the pace of things and Matt and Ben should weigh in on this. I'll be curious of their take but we also have to be careful about what we mean by stability in this context. like we we enjoyed a decadesl long period of stability in the nuclear era but and so we we're kind of looking for a similar multi-deade long source of stability in the AI era but it's unclear that we will ever enjoy that kind of stability and we may go through these different phases in our AI strategy but more rapidly I actually think that this this may the freedom of action strategy provides a an interesting lens on the way of looking back on nuclear strategy And our colleague Anne Stickles is going to be writing a Substack post. It may actually come out this week on this. But our strategy for the nuclear era has actually shifted over the time quite remarkably. There was a period of time in which we were sprinting and that was called the Manhattan Project. There was a period of time in which we had a temporary monopoly about a 4-year period in fact in which we dominated and we used it to our geopolitical advantage.
52:32 In that period we were quite anxious. Some of us were about the dangers of nuclear weapons and of a nuclear monopoly in particular and some of us proposed some of some some proposed something that resembled a moratoria. We didn't proceed with it but we eventually entered into verifiable agreements with our rivals. Later we used something looked like a suppression or deterrent strategy to prevent the the technology proliferating into Iran and North Korea with limited success. And now the and then the era of the nuclear era continues and it continues to proliferate and who knows what's coming next. So one reading of the history of nuclear strategy is it was a kind of freedom of action strategy. It's not one in which any policy maker ever stood up and built optionality as we're proposing but that's because they had the the freedom they had the lux luxury of time to build these kind of strategic optionality and we won't have that here.
53:29 So I think this freedom of action strategy it diagnosis is that unlike the nuclear you may still transition through these different strategic epochs but we're in order to navigate them we will need to pre-build the elements of a security architecture to make the transition. >> So we beat up the dominance one a bit. I definitely do want to get to the recommendations that you're making in terms of what we should be investing in now to maintain freedom of action next, but is there any other kind of nuance around any of these other strategies that you think we should be spending a little more time on?
54:08 Throughout our discussion and throughout our paper, we use the term coexistence to represent humans living with advanced AI maintaining agency. Alignment is often treated as a catch-all for coexistence, but it's just one path among several alongside control mechanisms, architectural schemes, and methods not yet known. So, we use coexistence throughout to represent this concept of humans living with advanced AI with agency. Let's talk about the archetypal strategies and why we think focusing on one particular strategy is unwise. We looked at all of the strategies, architectures, and proposals for international organizations that were out there and we found that they could be represented by seven strategy archetypes. Each of these archetypes has real advocates and a coherent win state.
54:50 But they also have a pathology which is the dark side of what happens even when the strategy is working well. And so for example, for a moratorum strategy, you have Eliza Yadcowski and Mary advocating for some sort of pause. Now their win state is that this pause is where you know the United States and its rivals would halt capability development and buy time for really that coexistence work to catch up. but it does have a pathology that you're giving up some of the benefits of advancing AI. But this enforcement mechanism that you would need to enforce some sort of global pause could harden into some sort of global surveillance regime. So you have Shri Hendricks and Wong proposed a type of deterrent strategy with MAMEim.
55:35 the win state there is that the United States would prevent dangerous AI basically by preventing it and creating some sort of moratorium at home but also suppressing it internationally including through force if necessary. The pathology there is that this might harden into some sort of standoff that never really resolves into any sort of peace. We've c certainly the audience will be familiar with your allezers and their calls for you know data kinetic strikes on data centers if and when ever needed and Dan Hendricks is a past guest on the podcast although it was actually before they came out with the mim document that that I talked to him. So I think so these like authors are definitely many of them familiar names.
56:18 Yeah. Continuity of society. I I planted a victory garden this year for the first time in my life. The squirrels got most of it, but it did reinforce for me the pathology of, you know, just trying to hunker down in some sort of, you know, subsistence mindset and and scrape by or, you know, eek out survival in the in the case that the we can't avoid creating them and coexistence is not super feasible. >> The other insight that I want to the other nuance I want to give you, Nathan, which is that we really see this as a strategy for America, less to the US government. Sometimes when we talk about that we elevate the importance of the US government because there are certain non-elegable responsibilities that are only available through the authorities that the US government has. For example, you know, maintaining the viability and the integrity of the nuclear deterrent is inherently a US government responsibility. Defending US national security in general is a non-eleigable US government responsibility. But there's a lot of things that are called for in the security architectures that we propose that we might too quickly assume are government responsibilities but are really opportunities for businesses in civil society across America to stand up. So for example, one of the very practical things that we think would be needed to support a freedom of action strategy would be investments in highly secured data centers. In principle the US government could support that but in practice we could create a market for that. Firms across the country have a dem should have a demand for that and we could rely on the market to create that. Another investment that is called for that you could imagine a government response would be in incident response and written reporting and response. Again you can easily imagine a government there role there but also you can imagine that being a role for civil society to fill that function. So, one of the virtues we think of this freedom of action strategy is it exposes opportunities not just for the government to marshall its forces, but all of American businesses and civil society.
58:21 >> Yeah, that reminds me of my grandfather who had tuberculosis and didn't go to World War II as a result of his illness, but nevertheless told stories of how we won the war. And it was like victory gardens and like car pooling to work to save on rations. And the point was like everybody has a role to play in such a an effort you know national effort as that. So I I like the framing of kind of reminding everyone that this is not a you know highly top- down civilization that we've built for ourselves and that there's a lot of things that people can and should be doing without necessarily being you know having to be told to do so without waiting for anybody else to give them any orders. That is a big part of the strength of the American system. I I think that it's especially true for the elements of our strategy that calls for investments in resilience and we see that already and Ben this is something I know you thought a lot about but you want to weigh in on that?
59:21 >> Yeah I mean I just going to say in that vein you know us managing our dependencies on AIs in terms of you know being able to operate critical infrastructure being able to develop and implement strategy and not personal levels right being able still to think through ourselves and maintain our our individual level agency. I think all that is is part of implementing basically any of these these seven strategies. You know, even things like moratorum and the like will really only be feasible if in we're able to live without advanced forms of AI. And so so some of some of what we need to do is to to preserve those human capacities again both at these individual levels and at the collective levels.
59:59 That's that's this relates to why why we use a terminology of coexistence as well and not just alignment which makes it sound like we're just you know training the models to conform to some set of established values where really what we're envisioning is a process of adaptation and and perhaps even co-evolution where where just as the AIS are becoming more capable we're more capable in living with them and again being able also to perhaps to to to live without them as as being a key way to to again manage some of some of the vulnerabilities that come from from those dependencies.
60:30 >> Safe to say the overton window is wide open these days. let's just zoom out and do the like seven objectives and and kind of step through them one by one and you can you know provide as much commentary on each one as you want but I think it is worth going one by one through the highle objectives that you set out. >> Great. So there's there's we should talk through the different objectives in the strategies. Thanks for the invitation you do that. There's really the objectives group into two categories.
60:56 The first thing is that we need to make investments to build and the second thing is is the operational elements or the things that we need to actually do and these are operate on different slightly different time scales. So it's worth like unpacking them. The first thing we need to build is the human AI ecosystem and Ben you should talk to that. >> Sure. This relates to what I was in fact just describing about ways in which humans and advanced forms of AI can coexist and and indeed humans can continue to thrive and and maintain our agency in that context. And a lot of this I do think is the traditional safety research. I think you know our analysis is we haven't invested nearly enough in some of the alignment and control techniques. We will continue to need new evaluations, ways of identifying when our when these advanced models can can tell when they're being evaluated and perhaps are, you know, sandbagging and we're not able to really understand their capabilities. We need to build additional tools that assume we're dealing with untrustworthy models and ensure that we have the the the containers and security environments that for instance prevent them from breaking out of sandboxes and the like or that we can contain their their spread. You know, but some of this is ensuring that our institutions remain at the center of like high stakes decisions and that humans are still calling the shots. We're still the ones steering the future. we are still the actors capable of making decisions that affect the various pivot factors like whether again suppression might be feasible and the like and and this I think it requires kind of a continual process of humans reflecting on the decisions that we want to maintain in human hands and so for instance some some some decisions in the broader discourse about for instance pacing or issues related to AI enabled research and development you know these might be core decisions where we really want to make sure we have sort of human checkpoints where, you know, even an in AI seeking to persuade a human, you know, that that that those types of circumstances are are are caught and humans again are are making reflective decisions and and deliberating with others in terms of just just how much decision space, how much autonomy, how much how many authorizations we want to delegate to to these models. I could talk a lot more about any of that, but I think that's probably enough for now. So one of the things that Ben and Alvin have written about I think is really interesting is is the idea of an agency eval.
63:19 >> Yeah. A lot a lot of the evaluations that we hear about are really evaluating the performance of model on discrete tasks including discrete tasks that should concern us if the models are capable enough to do them like engineer bioweapon for example. for one of the insights that Ben and Albert had and you should speak to it if you want Ben is we should be evaluating them in the context of their deployments for circumstances in which we may cross the threshold and we've effectively perhaps unknowingly given up some agency.
63:50 >> Yeah. I mean very briefly it's mapping the human decision space and then seeing what roles AIs play when they enter into those decisions either by you know disenfranchising various humans by persuading humans by controlling the agenda over which the humans are deliberating you know and these are I I think not just not evaluated typically now they're going to be hard to evaluate going forward because individual rational decisions to delegate something to an AI may not be easily seen and what you need to is track these decisions to delegate to AIS over time and in a more systemic way where you're really looking at trends and trajectories rather than just like single discrete decisions. You really have to I think think longitudinally about the again the role humans and AIs are playing in these you know various high stakes decisions like for instance whether to deploy a new model or whether to you know let you know an AI you know develop the next fad or something of the sort. So yeah, this is this is stuff that we're we're work currently ongoing and hope to have the paper out on that soon.
64:54 >> Another big objective in the strategy is to adapt our legacy national security architecture. Unfortunately, the old problems aren't going away. So we're going to have to adapt our counterterrorism enterprise for terrorists that are in possession of very advanced forms of AI. We're going to need to adapt our intelligence apparatus both to embrace the opportunity for artificial intelligence, but to anticipate how our adversaries will do the same. We will need to adapt our conventional military for the new demands that'll be placed on it and opportunities that'll be given to it.
65:31 Yes, we may be using our conventional military to defend US interests in the way that we've grown accustomed to, but we also may ask give it new missions to enforce an AI arms control agreement say or defend interest in new ways. One of our colleagues, Zach Bdett, has a leading a big project on what a new conventional military and an AGI air would look like. We'll need to update our nuclear enterprise. It's unlikely, as we may have referred to earlier, that nuclear weapons are going away. But the way we think about nuclear weapons and our strategy, what it means to have a resilient nuclear architecture could be quite different. My colleague Edward Gist is thinking about those kinds of question. Another one is we'll have to adapt our a biocurity enterprise and there's a big team led by Steph We at RAM that are thinking through that what that would look like. So all of our all of our what you might call legacy pre- super intelligence security architecture elements are going to need to be adapt for the very new challenges and opportunities that AI will afford them.
66:34 But we're also going to need a new element new security architecture for AI itself. We need an AI security architecture and Matt you want to speak to that? >> Sure. So we think that we need a a security architecture specifically for AI. And this is a you know a set of technical and institutional capabilities that both keep the technology safe but also the world stable. And I think the really important things to understand about this is that you know traditionally a security architecture is something that's really the domain of governments to establish but AI is quite different in that a lot of the elements for a security architecture are actually going to be built by the private sector.
67:10 So that's quite different. And then I think another thing to understand is that what we're proposing with the freedom of action strategy is actually not one specific configuration of capabilities. what we're proposing is actually building the baseline capabilities to run a variety of different strategies. We feel like you know emerging events or even this evidence resolving in a particular way you suddenly find out that you know AI is very dangerous and the danger is very close or you find out that coexistence isn't possible. you may have to reconfigure your security architecture on the fly. And so what the freedom of action strategy calls for is really taking a look at the capabilities that are needed by all of the different strategies. And then also taking a look at the capabilities that really have a very long lead time. They might not need be needed by every single strategy. So for example, verification technologies are not needed by every single strategy, but they have a very long lead time to set up. And the freedom of action strategy would call for investing on all of those so that you have the flexibility later on if you decide to run you know you're running a preparedness strategy and you suddenly realize you need to flip the mortorium you have some of that infrastructure set up and built already. Really we think the most important thing to build right now is the navigation capabilities.
68:24 These are the set of capabilities that let you understand where you are in the in the strategy space. These let you understand you know what is the the current capability of different models. what is the capability of Chinese models? Where are they? You know, what what sort of visibility do we have into in compute and where the compute clusters are? Without understanding the answers to these, you really can't answer some of those fundamental questions like is danger close? So one thing that occurs to me with this whole strategy and I I think this is like somewhat trendy in the policy world right now in general perhaps because downstream of co we realize that like going all in on efficiency creates resilience problems and we've maybe like partially learned a little bit of a lesson there. You know there seems to be now at least some appetite maybe only at the think tank level. I don't know if we feel this at Congress yet, but some some appetite to like invest in things that in portfolios, let's say, where some of the individual investments kind of by definition will prove to be wasteful in the end, but because we don't know, we want to make all these investments now so we have the option available to us later. I think that's, you know, hopefully a fair kind of echoing back of like a a core part of the strategy. What am what among these things are going to be expensive? like it's it's not I think that it all sounds good and until we start looking at the bill maybe right so what parts of this will be the sort of big bites that will have to be sold to people on this kind of you'll be really glad if you do in fact need it even if you don't really think that you are going to right now what are the things that are going to be hardest to sell due to price tag or otherwise >> I think this is one of the most important critiques that we've received of the freedom of action strategy is the unwritten price tag one of our colleagues told us in when we were this work was being peer-reviewed that a a strategy without a price tag is just a vision. Our response to this >> they were I think they were using claude to generate that by the way that sounds very claike.
70:18 >> It does. >> We'll put that in panggram after the >> Yeah, that's right. That's right. But I I think that they both they and Claude if that was the case are right about that. Our response to that though is is twofold. We're we're up on a on the cusp of a transformation and I I got to be honest with you any of the archetypical strategies that we put will be expensive with the exception of acceleration.
70:42 Although the expense may be borne out in the downsides but in all the cases all these strategies are expensive. We're talking about a so, you know, a society level reorientation perhaps on the scale of the reorientation that we had with the dawn of a nuclear weapons enterprise and the confronting the Soviet Union or the the reorientation of our society after 9/11 in which we had pivoted to to counter terrorism and counterinsurgency. So I think to a to a a very first order estimation all of these strategies are expensive in the resources attention and focus that they will required and the differential between any one of them and the freedom of action strategy is dwarfed by that between any of them and the status quo. I will say that most of the things that our freedom of action strategy would call for, you would consider them to be no regret.
71:40 Like in all these strategies or most of them, not all of them, but most of them, you're going to need to be able to maintain some visibility on compute and govern the flow of compute. All of these strategies will come require some situational awareness as Matt referred to. All of them are going to re require investments in resiliency of our critical infrastructure of our national security systems of our bio bio in a biocurity sense these are all going to be required etc. So there's very unfortunately it's expensive but all these strategies are quite expensive.
72:14 Better Matt would you want to answer that? >> Yeah I would just note that not all of these costs fall in the federal budget either. I mean you have companies, you have other actors who are funding much of the alignment, evaluation and security work and so it's not just something where you go to Congress and ask them to write a check. That's why this is actually like a national strategy because there's lots of actors that play a role in executing the freedom of action strategy. So do I understand correctly that basically the idea is like any of these strategies let's say it costs 1x to prepare for all of them simultaneously now is maybe like 1.5x and so the delta I think I think you said like the difference between doing one and and maintaining freedom of action isn't less than the cost of like doing any one thing is that a fair way to think about it >> I think it's a reasonable way to think about We haven't done those cost estimates. Where our research is going next is to think about how we operationalize the freedom of action strategy. What is it actually going to take? I don't know if we're going to come down to budget quality estimates that we can take to Congress, but like what is it going to actually take? Who needs to do what under different assumptions about the division of responsibility between governments, businesses, and civil society? What are our options space there? But intuitively, what you said is makes sense. The difference between the status quo and any one of these strategies is dwarfed by any one of these strategies and the freedom of action strategy. And I don't have an exact number for you, but they're all going to require reorientation of resources, our attention, our politics, our focus across the board. How guys using language models in your work? Like it strikes me that if you have a fairly detailed sketch, you can probably get agents to run that budget question to ground pretty effectively these days, right? Matt, you should speak to this.
74:06 Matt has been very forwardleaning on on this class of problem. Go ahead. >> Yeah. So, we actually use large language models quite a bit, mostly for organizing and remixing information rather than doing a lot of that primary research. We've actually found them incredibly useful, especially, you know, the project knowledge bases for keeping all of our various memos and all of the things we're compiling together. You know, on that particular question, we actually do have a a catalog that we call the security architecture catalog, which is 106 different capabilities that are needed for building out this ecosystem and the security architecture.
74:40 And we used LLM quite a bit in in gathering this information and also you know sort of compiling it and organizing it in a way that that humans can actually digest it. The problem is that you know some some of these things there actually is a primary source. There's a paper out there often we drew information from RANS's sort of sister center cast and you know there's a paper that can actually tell you exactly how long it's going to take to build say verification technologies but often there isn't anybody who's actually ever done that study and that assessment. So, that's one of the things we're going to be working on over the next couple months. We've got a sprint going on to take a look at this security architecture catalog and try and figure out how long are the lead times for each of these capabilities. What's a rough magnitude of cost for these different capabilities? Who's working on it? Who should be working on it? Should the US government own these capabilities? Those types of questions. And we'll be publishing that information after this sprint is complete.
75:30 >> Yeah. And what I what I might add on this is that the freedom of action strategy provides like a send a decision criteria which you might use to look at a policy proposal or some of these capabilities from the security architecture catalog Mac was just talking about. So like it basically lays out a set of questions you would ask for any given policy proposal like one main question is which of the strategies among the arc types does a given proposal foreclose and and how does it do that foreclosure, right? And that might be like a real cost that's not associated just with like the economic cost of the capability, right? If it prevents you from having a choice that you might need later. Well, that's that's a you know a a significant cost that needs to be factored in. And so that that issue of foreclosure is something that you know I I I found you know engaging with LLMs and seeing you know ways in which given policy proposals might might foreclose different strategic choices. You know, I I I think that's like important and also the extent that those options are permanently foreclosed, you know, that the your your choice to pursue a certain policy was irreversible versus whether or not your policy could be reversed, you know, in time. And you know, you you talked about for in export controls as an example. Well, that's the sort of thing that, you know, might, you know, one might argue foreclosed certain forms of co-development or the like, but also you might argue, well, you know, there there was nothing, you know, necessarily permanent necessarily in that foreclosure, these these could be reversed or or shifted in certain ways.
76:54 So, you know, so that's sort of one criteria you might look at like whether or not a given policy forecloses other strategies and then the cost associated with it. But also related is what strategies does a given capability open up, right? And that's what Matt was just discussing in some of the the things Joel was talking about the the the no kidding most important things we ought to be prioritizing because they help us advance the entire strategy space right across all seven strategies. And one way that a given policy or capability might do that is if it provides information, right? If it helps us better track the pivots or provide information about capabilities or you know what China's up to or the like that that that all you know provides provides real value of given capability you know that that might offset some of the cost associated with the capability right if it if it just allows you to make better decisions down the line well I mean those those sort of upfront costs might might be you know very much worth it. And then I I think a last sort of criteria that grows out of this way of thinking is is what does it signal to the rest of the world other geopolitical actors and and you know that might be where again you know the your point about export controls is relevant. Maybe the real cost associated with that was it signal that we were not interested in some forms of partnership or or something like that. you know, maybe. But are there other things that we could do to signal that we're interested in in various forms of partnership if the time is right and we have the right verification mechanisms and we're not worried about, you know, someone defecting from their agreements and the like. And so, you know, those those criteria looking at again at, you know, foreclosure of strategies, looking at what strategies are opened up, looking at what information we might get from a given policy, and then and then lastly, what does that policy signal to other geopolitical actors? I think for for me that's really how I think through like what what should we be prioritizing, what policies make sense, what are the costs associated with it and the like.
78:42 >> Do you have any other thoughts? I mean so a couple quotes that I pulled out of the report were on the topic of international relations and signaling that the US should quote signal to hold deterrence and cooperation credibly open at once. Another quote that I particularly appreciated given my relative doubbishness is that we should weigh carefully any action that would harden a rival's resolve. I think my sense having spent a couple weeks and you know by no means a China expert but the take I got there from people this summer was like yeah these chip restrictions are like hardening our resolve that seems to be happening at least somewhat. I do agree that you know it is nice you can reverse that policy and start selling chips again or you know come to some new agreement if if circumstances or attitudes change. How would you think kind of high level though about like this balancing of deterrence and cooperation at the at the signaling level? How do we how do we even know if we're getting that right?
79:39 >> Yeah, that's a great question. I mean I think you know the idea here is you want to demonstrate capabilities selectively. You want to share the existence of capabilities selectively. Often you're going to want to withhold your intent behind capabilities. Sometimes you're going to want to signal your intent behind capabilities. And so an example here would be, you know, if you have a credible independent evaluation and verification regime, it actually serves both sides of that posture and it doesn't actually foreclose anything. So that's where building this capability and and announcing it and sharing it can be actually really useful and actually preserves your optionality. I think in areas that might be more confrontational. So like offensive counter AI tools, you might not want to disclose that you actually have those or you might want to disclose that you actually have them, but not explicitly what the purpose is because that type of signaling can actually foreclose some of your options.
80:29 >> Yeah, I can just state like coherent enough to do some of these things today. One thing I I look out at at China and I'm like I don't know what's going on behind closed doors, but they definitely present a pretty unified front whereas we are like counter signaling you. we're like signaling to ourselves so much that it's like I think very hard for us to have any sort of coherent signaling strategy on the international stage.
80:51 Does that feel right to you or do you think we actually can do that more than I give us credit for? >> I I think we can do that. I think you know it requires some policy discipline but I think sometimes that ambiguity is actually beneficial and so if you look at the ambiguity around Taiwan and what US intentions would be there I think some of that ambiguity can actually serve our purposes. you know and just I mean very briefly on this I mean you know we have the example of the cold war where you know our open society the freewheeling public discussion the critiques of leaders right you know there was times when that was seen as a disadvantage but I think you know in retrospect you know that actually made our system much more resilient much more able to you know take on different possibilities you know that that there was more just legitimacy and support for ultimately what decisions were made and so I you know I I personally see that like actually has as has a huge advantage. I know it can look a little messy, but but but ultimately I think that's that's to the good. You know, just just going back to the other question about deterrence and cooperation, like I think there's ways in which and we've heard this critique that the freedom of action strategy is kind of asking us to do it all, right?
81:57 It's like have your cake and eat it too or walk and chew gum at the same time or whatever analogy you want to use. In some sense that's that's true, but that doesn't mean, you know, that this is somehow like feckless or not particularly demanding. You know, we think that there's lots that can be done that advance both deterrence and cooperation as as the example. I mean, Matt was just laying out, you know, parts of the evaluation ecosystem or verification tools. I think th those are are very clear examples that support both of those. Another one just to throw it out is like incident response both domestically but also with crisis communication channels with with adversaries like China or with with other countries around the world. You know that that again I think supports multiple strategies at once. We we don't have these yet up and running. And so even if we only advance the things that serve all the strategy space that gives us quite a lot to do and you know then we don't have to you know wrongfully commit to something that you know would would be premature and would deprive us options later or perhaps you know signal things to China that would be overly provocative. We can do the things that that enable both like the deterrence and and the cooperative approaches at the same time.
83:04 >> Nathan on this amendment point of weigh carefully any action that would harden a rivals resolve. I mean, I think this is one that's that's really important where, you know, sort of moving towards some sort of suppression capability can basically set a signal to your rivals that the window is closing for them to actually take action, right? It can push them to accelerate. And so, there's been an argument that, you know, chip controls basically fed China's push to really create their own domestic, you know, AI GPU accelerator market and and basically, you know, start producing those because now it's existential, right? They can't procure GPUs on the international market as much as they could. So, they have to produce them themselves. And I think there's some logic to that. I think one of the we might have mentioned this earlier, but I think one of the most dangerous things if if there's a dominant strategy being run, one of the most dangerous things that, you know, the United States could do is actually signal that that dominant strategy and that monopoly would actually be used to threaten, you know, the Chinese Communist Party. I think anything that turns existential for a particular rival then really puts the pressure on them and and hardens their resolve and and is going to focus their efforts and that's where you could get some sort of you know dangerous consequences from that. So I think that's where you want to be very careful when you're signaling to make sure that the message is being received in a in a particular way and you're not boxing your adversaries into a corner. Yeah, that was definitely hardened at the level of they are going to have a chip industry at this point and you let's hope it's not hardened too much beyond that. Are there other kind of hidden gem policy ideas that you think just deserve more attention right now? Like my friends V always hammers on the Jones Act, you know, repeal the Jones Act as like his sort of quicksotic quest in life. Are there things that are maybe not the you know the headline but are sort of among you guys in the no like your kind of favorite things that are neglected when you know there's not enough time to get into the weeds like we have today. I mean, I've written on some of this already, you know, but I do think a lot about both how we're going to respond to incidents and prevent incidents from happening. And some of this it gets a little conceptual relating to what it is for us to hold AIs at risk to ensure just as we are vulnerable to them that they they remain vulnerable to us. going back to the the conversation we were having earlier and I think you know ensuring various capacities for for shutting down models and being able to to credibly threaten to them that we we will do such a thing and some of that gets to our resilience and ability to live without them to make those types of threats credible but also ensuring we have the right technical approaches the right authorities in place the right crisis processes so that we can act quickly. I think all of that is is also pretty important, especially if we're we're worried about the the very high-end risks from advanced AI that that something about our ability to to hold them at risk to be able to shut them off, I think, is is is pretty crucial and and and maybe doesn't come out as explicitly in this particular piece.
86:11 >> Yeah, I'd be interested in any double click you have on that or even just any sort of advice for, you know, the rest of the country outside the government. Yeah, >> I mean I just had this experience yesterday where I was I'm kind of try I've like put a lot over the last nine months or 10 months into my own personal AI infrastructure and it's gotten pretty good and I've just happened to be given a couple talks this week and one of the things I'm going to leave people with is like hey here's a couple public versions that are sort of depersonalized of my setup. Not that you want to copy it, but like go talk to your agent about it.
86:46 Maybe you'll copy some, maybe you already have some. Whatever you can maybe learn some. And I'm al also going to offer sort of agentto agent exchange which I think is going to be you know potentially interesting. But one thing that happened as I was doing this depersonalization is that Claude decided to rotate the passwords on the agent console software that I use to talk to the agents which is increasingly how I'm doing all the work. And there was a moment where I was locked out of my own system >> and I was like oh I don't even know where that password is. It just changed it. Claude can change my passwords. like this is an an issue I hadn't really prepared myself for. Now, COD is not fortunately scheming against me at this point. So, I just went to a terminal and asked it and it gave me the password and then I logged back into my thing. But it was a kind of a little wakeup call of like, wait, I've intermediated between myself and the actual levers of, you know, action in the world so much already just in October 2026 that like if Claude turned bad on me, it could go really bad. Like it could block me out of all my You know, this is like getting real for me at least already. Do we have any do you have any tips for me?
87:54 Do you have any best practices you want to promulgate or habits of mind you want to cultivate in people? Yeah, I mean I I think a lot about this less in terms of which passwords I want Claude to manage, but more which type of like analytical work, writing, other types of critical thinking I would like, you know, Claude or other LLMs to do. And this is also something I'm I'm I'm a teacher, a professor, it's small kids, and you know, thinking about which capacities it's important that they maintain and and the best practices for how they're interacting with AIS. and and what I advocate for is always starting first from your own considered judgment and and and point of view. So before I ask a model for its opinion on something, I want to already have real clarity about what my question is and what sort of good or bad answers would be. And so going into it already with a with a clear vision of of sort of what I'm hoping to get out of the interaction.
88:47 Then of course like letting letting the model cook and letting it expand my thinking and you know build the tool or or do all the things that I wouldn't have been able to do without it. But then on on the on the the other end I'm ensuring that that critical judgment and falcony comes in again and that I'm you know really being thoughtful about the response. I'm not just you know putting my name on it and passing it forward.
89:07 I'm I'm really thinking thoroughly about it. And so so sort of the the the human comes in at the beginning and then it comes in at the end. And and and thinking of that sort of like cycle as as being something I really want to you know inculcate both in myself and my students and my kids and the like you know that that's that's a best practice and sort of like personal levels in terms of these like more institutional decisions about when we should be relying on AI you know in terms of like I mentioned critical infrastructure like things like you know managing your your desktop or whatever it might be. you know, it it's it's hard for me to have like one sort of best practice to suggest for everyone except I try as much as possible to continue to ensure that I can operate with without these things. And some of it is just like the cutesy like I'm going to try to, you know, navigate myself to, you know, my friend's house without looking at GPS or something like that, right? Just to try to try to, you know, flex that that type of muscle. But but I think more institutionally it is thinking about these human backs stops you know various fail safes that allow you know manual operation of of different systems you know still having paper copies you know of the health records or the like you know because I think all of that both ensures that you know if there's some you know incident we can we could perhaps withstand it you know that's where the resilience comes in but also makes our you know our our certain threats to shut shut these systems down much more credible right it makes them a lot less costly to us society and again on these personal levels and and so I I really think of it as being very clear about which dependencies like in our institutions have thinking about what we can do to lessen those dependencies or at least have again these these fallback options and and then and then you know exercising them practicing in them make sure that they're they're they're really available if and when you need them but you know not to say I've like perfected this it's very much you know it's very much a practice something I I I strive to do and and often fail at That that's that's sort of how I think of it.
91:01 >> Yeah. Almost seems like we need like an AI Sabbath day. Even if we're not observing religion, we need like a day a week where we just like don't do AI stuff. >> Exactly. Yeah. And you know, you can extend that to to other examples. So like already, you know, in LA USD or even the New York schools, these are now AI free. You know, that was an explicit choice made by the school districts. And I think it's sort of along the lines of of what you're suggesting, AI free spaces. And do we need to establish more of these, you know, community centers where, you know, AIs are are not welcome. You know, there's there's human activities happening, people are, you know, exercising or convening or these other types of activities that are that are very human and doing it free of the AI mediation so that we can sort of, you know, practice those those human abilities and again be prepared for a situation where we might not have AI or might need to stop an AI from from from intervening. So, I think a lot about those types of things. I do think things like, you know, an AI Sabbath, a you know, sort of a a human reserve labor force for critical infrastructure or hospitals or or the like, you know, might might be important. Yeah. And these other these other practices. Yeah.
92:06 I try actually to take an AI Sabbath. I I often fail unfortunately, but yeah, do do my best. >> Yeah. Well, that might be something I need to start >> when you're talking about things where I think there needs to be a little more attention. I think one of the things we were looking at is actually the typology of super intelligence. So when we were doing this assessment all of the strategies out there, we found that most of these strategies were actually focused on the idea that super intelligence would actually be very centralized either a singleton or maybe you know the United States has one and China has one and they became very brittle when you started talking about some of these other models for super intelligence. And so, you know, there could be sort of lots of different super intelligences out there at at varying levels of capability that are very widely distributed. There could be super intelligence that's very narrow and you know, you use architectural controls to make these more like an oracle where they don't have a lot of generalized intelligence and there's lots of checks and balances on them. But one of the more interesting ones was this idea that you could actually get super intelligent types of behavior from agent swarms that are sub ASI or even sub AGI. And this is something where, you know, we went and did some initial research on this.
93:15 There's really just not a lot of thinking about this. And I think what we've seen with the agent swarm hacks this summer is really maybe a preview of what's to come. And this is less like, you know, dealing with sort of a singleton or, you know, even a constellation of super intelligences and more like looking at these things as like an ecology. And so some of our colleagues at Rand, Edgeist and a number of other authors put out a paper called the ecology of AI risk. But I think this is really an emerging area of research and artificial intelligence. It is enormously important as we start getting more and more agents out there. You know, they may be interacting with each other, forming different coalitions, having these types of emergent behavior that we don't understand. We spent thousands of years really figuring out all the checks and balances for how human agents interact with each other out in the world. We have very little time and very little experience with how these types of agents are going to interact with each other out in the world. And this is very very important for I think how things are going to turn out.
94:12 >> Yeah. ecology of AI is it's wild times but I've had a very similar impulse that like you know we've sometimes of course but China is always at the end of a lot of AI conversations but I also sometimes think about like but prime intellect you know it's like what what is going to happen at the distributed layer in a community of you know participatory fashion in some stories or in some timelines that may be just as important as what the Chinese government decides to another. Okay. So, kind of opposite side of the coin, if you will, of all the investments that you recommend that we make to maintain option value. I wonder if there are any things that you would say we have enough information right now that we should all be able to come together and agree. Certain things we just shouldn't do. And like one that I like to float, which you may or may not accept, is I don't think we should train AI generals. If we train AIs that are capable of winning military campaigns, you know, the foreshadowing is like so obvious, right? That it's like we we are training the thing that is going to then be fighting us at some point potentially, right? And like I don't want AIs to be that good at fighting military campaigns. Yes, they might beat our adversary hypothetically, but like I just kind of worry about our ability to fundamentally control that technology.
95:35 Is there anything you would put into the category of yes, this is just obviously such a bad idea. We definitely just shouldn't do it. If if I may just try and understand your example, Nathan, which is interesting, is the idea we just shouldn't even train such a model or an agent versus, you know, maybe we train it, maybe we leverage it, but it's still going to be humans that are fundamentally going to make a determination about whether to deploy such a thing, whether to act on the recommendations from this this general.
96:04 So, so is your is your objection just in the in the very just the training itself or in how we're deploying this decision? >> I would probably like to stop a little bit upstream of the final decision to deploy. I feel like if we all of a sudden, you know, hand that over to the military decision makers, it might be pretty hard to resist. I honestly would probably say the same thing even about just like training AIs to go out and make money on the internet and just rewarding them for how much money they make. It seems like that's a pretty clear way where we're going to get all kinds of lying and cheating and conniving AIs and yes, they'll be economically valuable, but like they'll be at least as with the techniques we have right now. You know, possibly, you know, as you alluded to earlier, methods not yet known could come online that could, you know, salt in with this and make it okay. But like at present, you know, what I've learned from like Anden Labs type stuff is like already even the models we do have which have not been trained in that way are like, you know, sometimes pretty dicey to deal with.
97:01 >> Yeah. >> So yeah, I just think if we go down that path, it seems to me we're going to have a very hard time not deploying and we're going to have a very, you know, very dangerous situation where it's like, you know, that the military is now run by the AI. Now we have to fight that. Like it's a really really tough place to be. >> Yeah. No, I I'll I'd be interested to think more about this one, but I you know to to my mind the the line is more around giving access of this agent to various tools or capabilities that allow real world impact. Now in the making money case, I feel like that ship has sailed. You know, I I don't see how we can we can pull back. I mean, these these companies are designing systems that are supposed to have some economic value and run for, you know, over long horizons and the like. And so that one I think might be harder. But in the military context, we are accustomed to complying with various restraints and obligations that that we do for for the good of humanity. You know, and these are codified in the law of armed conflict, international humanitarian law. And and I think it's those same types of legal constraints should apply with respect to our use of AI in military applications. And they require for instance you know reviewing operations to ensure that they don't have you know disproportionate collateral effect you know ensuring that we are distinguishing civilians from from you know combatants you know and ensuring that someone remains accountable for these decisions and and that to me just suggests that even if you're training these agents and getting some value out of them in the planning process you know we still need to create institutions where the humans are remaining fundamentally accountable and we do have practice of that in in the military with like clear hierarchies and we just need to double down on those and that might mean you know bringing in more of these military lawyers to review operations bringing in more just people thinking about what's appropriate or not but I wouldn't necessarily say that we are completely unprepared for such an eventuality we just have to I think go back to what's what's worked historically and and really ensure that those existing institutions adapt and are and are relevant for today's challenges but sorry that's not answering your question I'm just I'm I'm actually just interested in that as a I guess a side side side issue.
99:11 >> Yeah, that's good. Yeah, I mean I offered it for you to to beat it up a little bit. Anything else though that anything come to mind in the definitely thou shalt not category? >> Yeah, I mean I think one of the things that we talk about a lot is that you know rushing to put AI into your nuclear command and control is probably a very very bad idea. And you know this needs to be done in a very deliberate fashion if at all with a lot of demonstrated safeguards. And so anything that you know you're touching on the the nuclear domain is an area that you need to proceed very very cautiously. But I think there's also you know the use of of agents in military affairs. You know some of the research has shown that you know and this is all you know I think very almost anecdotal but you know people have found that some of the models when they're put into these simulations basically tend to escalate a lot. And so, you know, if you have agents that are basically doing reward hacking and escalations on the table, that seems like something you want to be very very careful with. Also, >> yeah, chilling, honestly. All right, last one. What What are you tracking most closely, right? So, just to revisit the structure of this whole discussion a little bit, we want to maintain freedom of action. To do that, we're going to make kind of parallel investments in a lot of different things. The reason is because we have seven archetypal strategies and it's too early to decide and we have these five pivotal questions that we're going to have to kind of work through and get answers on in order to decide when it's time, you know, what strategy we're ultimately going to commit to. So with that in mind, what are you guys tracking most closely right now that you consider to be the kind of, you know, likely next data points that will start to shift your sense of where these pivotal questions are going to shake out and then obviously, you know, what strategy we want to adopt downstream of that. I mean, I think the things I'm tracking are probably not not a surprise, probably what a lot of the community is tracking. I mean, one is continuing to see developments along the pacing proposals. What actually gets enacted? What type of access do independent evaluators actually get internally at the frontier companies?
101:22 What sort of shared standards are decided on among these companies? You know, and whether or not various security controls are sufficient to manage internally deployed models. So far, we've seen that the companies have not created sufficient security internally. And so, I'm very much watching to see whether or not they they will get a handle on that. and if they don't what type of accountability or consequences can come and then also very much watching for signs of recursive self-improvement or what what the trajectory of the AI R&D space it looks like since that can really change timelines across the broad range of decisions we are tracking and then the last thing I might mention of course is the the USChina dynamics from the broader diplomatic standpoint but also more narrowly on the AI issues whether or not some sort of incident management channel can be created or there's more shared understanding of the risks. That's yeah, I think that's the stuff I'm watching.
102:19 >> Yeah, I would just add on to that. You know, I think the the important things to watch or as I said, danger close is probably the most important question we can answer right now. And for that, we want to really be looking at the pace of recursive self-improvement. How are the labs managing that? You know, they're doing their own pacing. Are they actually able to release new capabilities safely? Are we continuing to have you know more and more and escalating incidents around agents or are we effectively developing the standards around how agents are used? I think one of the interesting things to look at right now is cyber capabilities.
102:54 So mythos was released about 9 months ago now and typically open source open weight lags you know the frontier by 6 to 18 months. So those types of capabilities are now starting to be available in more of these openweight models. So, we're need going to need to see, you know, how are malicious actors using these capabilities? How are state actors using these capabilities as they become more and more available? And I think, you know, when you talk to people in the in the cyber world, I think there's a lot of concern about the next couple of years. I think over the long term there is a view that you know in the long term AI might be positive for cyber defense but a lot of that depends on whether the defenders can actually uptake the the frontier capabilities quickly enough and if there's these openweight capabilities that are out there that enable you know malicious actors terrorist groups criminal groups to go and hack it may be that big tech's fine but what about the small and medium enterprises that aren't able to actually uptake the latest frontier models as quickly and so I I think we may be in for a rough patch in the cyber world, you know, over the next 12 to 24 months.
104:01 And hopefully it's just a a narrow rough patch, but that's something I'll be keeping a close eye on. >> Yeah. Offense can stay dominant longer than you can stay in security through obscurity, I suppose, is the >> Absolutely. >> Yeah. Well, this has been great. I really appreciate all the time and and the hard work that goes into making you know this is really one of those documents. I would encourage people to to share and you know just go through the the hierarchy because this is one of those you know you actually took the time to write the short version and I think it it it clarity does really come through. So I I appreciated that a lot.
104:38 Anything else you guys want to touch on that we that I didn't think to ask or we just didn't get to before I let you get back to work? >> Well, just add one thing. So, I was just at the curve last weekend, which is one of these big AI confabs in Berkeley, and I presented on the freedom of action strategy, and a lot of the feedback I actually got was that this really enabled people who are newer to the AI debate to really understand the different dynamics at play in looking at and choosing strategies for advanced AI.
105:07 And so I I think folks out there in the audience shouldn't look at this as like, you know, just a an academic rand report. I think there's people kept citing this table that we put in there that outlines the five pivot factors and how they lead you potentially to the different archetypal strategies as something that they were going to take back to their stakeholders and constituents to try and explain, you know, how sort of AI strategy works. So for folks out there who wouldn't normally read a RAM report, I would encourage you to go just take a look at that and take a look at the pivot factor table. It might actually help you understand a little more intuitively sort of what these strategic choices are all about and what the underlying assumptions for most of the AI strategies out there are based on.
105:47 >> Yeah, I agree with that. I think for people who've, you know, been with us on this feed for a long time, like all these things are kind of in the air and familiar and I was at the curve too actually. I'm surprised we didn't have a chance to meet there, but for folks who are catching up now, I think the the structure and the the way that you guys present this information is an excellent primer. So, I do think it is definitely something that should be forwarded on to people who are kind of just entering the discourse at this possibly late hour. I guess we don't know yet exactly just how late it is, but yeah. So, download it and you'll be glad. You'll have something that you can send to friends when they need a a primer. I think it's excellent for that.
106:26 Joel Pratt, Ben Bro Budro, and Matt Chess. Thank you guys all for being part of the cognitive revolution. >> All right. Thanks for having us, Nathan. >> Thank you, Nathan. >> I'm building the capacity to commit >> a little every day. I'm building the capacity to commit. Now run along and play. A boy brought roses. Monday, another brought a ring.
107:00 I wore the ring to lunch with the third. He didn't notice a thing. They say I'm fickle, flighty. They say I'm hard to pin. I say a girl should keep her door wide open. Do come in. I'm building the capacity to commit. Ooh, a little every day. I'm building the capacity to commit.
107:33 Ooh. >> Now run along and play. You ask me if I love you. Why, darling, more or less. I love you like a brand new hat. On Sundays, I confess. You say you'd give me everything. How generous of you. I'll take it all and I'll take my time and I might take him, too. Oh, I'll settle down some morning when the moon comes up at noon. When the last of you stops calling. So don't stop, dear.
108:20 That's too soon. I'm building the capacity to commit. >> She's building it a little every day. >> Every day I'm building the capacity to commit. Now run along and play.
108:51 If you're finding value in the show, we'd appreciate it if you take a moment to share with friends, post online, write a review on Apple Podcasts or Spotify, or just leave us a comment on YouTube. Of course, we always welcome your feedback, guest and topic suggestions, and sponsorship inquiries, either via our website, cognitive revolution.ai, or by DMing me on your favorite social network. The Cognitive Revolution is part of the Turpentine Network, a network of podcasts which is now part of A16Z, where experts talk technology, business, economics, geopolitics, culture, and more. We're produced by AI Podcasting. If you're looking for podcast production help for everything from the moment you stop recording to the moment your audience starts listening, check them out and see my endorsement at aipodcast.
109:35 And thank you to everyone who listens for being part of the cognitive revolution.
Summary
- The seven archetypal strategies include: acceleration, dominance, co-development, preparedness for coexistence, moratorium, deterrence, and continuity of society.
- Five pivotal uncertainties must be addressed to determine the best strategy: proximity to dangerous AI, feasibility of coexistence, potential for decisive strategic advantage, ability to restrain development, and the possibility of using force to prevent AI development.
- The authors argue for investing in foundational technologies and institutions to keep options open while uncertainties are resolved.
- The conversation highlights the shifting political landscape regarding AI in Washington, D.C., and the need for a coherent strategy amidst rapid advancements.
- The importance of resilience and maintaining human agency in the face of advanced AI systems is emphasized, advocating for a balance between deterrence and cooperation.
- The authors suggest that the U.S. should build capabilities that support multiple strategies, allowing for flexibility in response to evolving circumstances.
- They stress the need for careful signaling to avoid hardening rivals' resolve and escalating tensions.
- Key areas to monitor include developments in AI capabilities, the effectiveness of governance frameworks, and the geopolitical dynamics between the U.S. and China.
Questions Answered
What are the key components of the new Rand perspective paper on AI strategy?
The paper outlines a taxonomy of seven archetypal strategies for achieving geopolitical advantage in AI and identifies five pivotal uncertainties that need resolution before determining the best strategy.
What does the continuity of society strategy entail?
This strategy acknowledges the inevitability of super intelligence and focuses on resilience, suggesting humanity may survive as a relic with limited agency.
What challenges exist in defining a decisive strategic advantage in AI?
Defining a decisive strategic advantage is complex due to varying interpretations and the economic forces driving technology, making it difficult to prevent adversaries from crossing critical thresholds.
How do perceptions influence the outcomes of AI strategy simulations?
Players' prior beliefs about strategic advantages significantly affect the outcomes of crisis simulations, indicating that perceptions can be as impactful as actual facts.
How should we evaluate AI's role in human decision-making?
Evaluations should focus on how AI influences human agency, including its ability to persuade or control decision-making processes, rather than just task performance.
How should the US approach deterrence and cooperation in AI policy?
The US should signal both deterrence and cooperation credibly, demonstrating capabilities selectively while being cautious not to harden rivals' resolve.
What are the risks associated with training AI for military applications?
Training AI for military campaigns poses significant risks, as it could lead to the development of systems that may act against human interests in the future.