Anthropic co-founder: AI impact '10x larger and 10x faster than industrial revolution'

**Jack Clark:** We are building a technology that may have the properties of uh something that kickstarts a process 10 times larger than the industrial revolution that occurs in 10 times less time.

**Krishnan Guru-Murthy:** How do you square that as an individual ethically, morally knowing that you're working with a technology that as you say may may be great but could destroy the world and is making you a billionaire. If there's anything that's going to need to change in the future, it's arriving at some kind of regulatory format which actually controls the industry rather than just a bunch of extremely big personalities making big personality idiosyncratic decisions about perhaps the most powerful technology that's ever been built.
And finally, as a topic, what do you think the future of AI and sex is?

**Jack Clark:** It's an unexpected question. Hello and welcome to ways to change the world. I'm Christian Giri Murphy and this is the podcast in which we talk to extraordinary people about the big ideas in their lives and the events that have helped shape them. My guest today is Jack Clark, co-founder and head of policy at Anthropic who make the Claude AI systems. Before moving into tech, he was an English graduate and a journalist. Now he's one of the voices both evangelizing about the possibilities of AI and warning about its dangers and discussing what governments should do to protect us from technological leaps that are outpacing our ability to control them. Jack, welcome to the podcast. Thank you very much indeed for joining us. If you could change the world in any way, how would you change it? I want to give the world a better way to understand the most powerful technology being developed in Silicon Valley and new ways to govern the private sector actors which are building that technology. So make sure the world knows about what's happening and also that it has different and better ways to intervene on the on the actors that are building that technology.

**Krishnan Guru-Murthy:** Is is that your job essentially?

**Jack Clark:** That is my job. It's been it's been my half job for 10 years. You know, I worked at OpenAI and then I co-founded Anthropic and recently I've launched something called the Anthropic Institute, which is a new function here that brings together multiple technical teams who are tasked with studying our AI systems and building tools that let us know more about them and sharing those results of that research and the tools with external experts so that they can study our systems as well. Because I mean a big part of your public role seems to be not just to evangelize and talk to us about the potential of AI to improve our lives but to warn us about its dangers.
Everyone who's building powerful technology throughout history whether electricity or airplanes or people who were building the first steam engines they had some intuitions about how it would change society and how it would lead to positive things but how it also had risk contained within it. And I think the lesson from Silicon Valley for the last few decades has been technologists have come off as entirely too sunny about the technology and have never really talked about the the the potential anxieties that they themselves feel. We're trying to do something different with anthropic where we're trying to tell the whole story as we see it. Um because we think that's what people demand increasingly of companies like us. Do

**Krishnan Guru-Murthy:** do you think you know the whole story?

**Jack Clark:** Oh no, that would be that would be tremendous hubris. But I think I have have a sense of like some of the ups and downs

**Krishnan Guru-Murthy:** because you're you're not a you're not a scientist, a technologist yourself, are you? Your your your background was English and journalism.
Correct. Um the wonderful thing about modern AI technology is it's it's almost like reporting on a on a foreign country or a or a company which is incredibly hard to get information out of. You basically need to figure out the right questions to ask and you have these AI systems that will reveal answers upon you asking the right questions. So the sorts of questions I sit around asking are, you know, how might the AI systems display different normative properties? You know, what are their biases? What are their tendencies? You can do all of that through asking questions and working with technical teams that build increasingly complicated tools that let you ask for questions. But in a very weird way, I found my background as a journalist ended up being extremely helpful for interrogating these systems for how they show up in the world.

**Jack Clark:** I want to go through, you know, you know, the various harms that people have identified and discuss it. But but but I I I mean I wonder is it inevitable that we will not have, you know, the harms that may pop up will not yet have occurred to us? There are always uh unknown unknowns, but you can give yourself a much better ability to to predict the future by trying really hard to imagine imagine bad things that could happen and build tests for them ahead of time. So, I'll give you one example. Um, Anthropic recently announced a system called Claude Mythos, which is extremely good at cyber. One of the teams I run, the Frontier Red team, last year we'd done early work on on the risk of of biological weapons from AI systems. And we realized that AI systems had got good enough at coding that cyber might be about to be the next risk. So we spent a lot of last year building tests for cyber capabilities of models. And what we saw was that models were getting ever so slightly better at things involving cyber defense and offense. But when Mythos came along, we pre-built all of these tests and immediately after the model arrived, we ran it through them and that instantly showed us that the the AI system had got dramatically better at one class of risk. So that's an example of how you can you can set yourself up to at least uh be rapidly cued to new risks that might come from your technology if you've done this previous sort of analytical work.

**Krishnan Guru-Murthy:** And and can you just address the the cynicism that is out there around around mythos? you know that I'm quite surprised given the advances in AI uh seem quite obvious but a lot of people are saying well this is all being overblown and it's obviously in their commercial interest to make mythos sound absolutely amazing because then everybody will want to buy it
the unfortunate thing about AI is that AI systems get much better much more quickly than our intuitions do about how capable they are and I've been I've been involved in this for many years so you know back in 2019 19, Dario, the CEO of Anthropic and myself, and we previously worked at OpenAI, rolled out the release of an early text generator called GPT2, which is a precursor to chat GPT. And we talked about some of the potential issues. We said, "Oh, GPT2 could lead to synthetic text. It could lead to fishing. It could eventually lead to to different forms of of of cyber crime." And we were right about where the technology was going to go, but we were slightly wrong on on timing. And we also hadn't shared the technology with other people. So we just had it for ourselves. And you had a you had a mini version of what we're having with mythos now. People saying this is potentially overhyped. We can't even see the technology. They're saying this because maybe OpenAI wants to sell this stuff eventually. So what's changed? Well, now all of the risks that we talk about are tied to extremely like thoroughly designed tests. You know the UK AI safety in AI security institute tested out myos. It did extremely well on their cyber ranges which are things that have been built by them and anthropic doesn't get to see. So that's a third party validation. Additionally we shared it with many many of the world's most most important kind of companies or open source organizations that operate large platforms. If you talk to any of them they say it's very real. And we also found real bugs in the wild in places like Firefox and others. So it's real um but it's not special. It is just representative of how good AI systems are getting. And there will be many systems that have this this capability like mythos coming along in the next few months and years. So it this isn't a a marketing strategy. It is an earnest attempt to discuss the technology and its implications before the most challenging implications arrive on mass for the world to deal with.
And and the potential damage or or havoc that mythos and AI systems like it could wreak is coming into view now, isn't it? I mean, you know, that couldn't couldn't they totally disrupt uh the world's financial system, uh, Bitcoin, um, you know, all all systems basically, all all systems that that are hackable.

**Jack Clark:** This feels similar to to the runup to Y2K where you have this potential potential bug, right, where software hasn't been built to roll over from 1999 to 2000 and you don't know what will happen. And there's some evidence that that extremely bad things could happen. What happened was the world put huge amounts of resources into basically going through and rewriting huge chunks of the world's software and then as you and I remember uh you know the the clock rolls over and all the lights stay on in wherever you're celebrating and the fireworks still go up and everything's fine. But I suspect a large part of why it was fine was we did this huge preparatory action. What we have here is a chance to use Mythos and other technologies like it to rewrite huge chunks of the world software to be much much much more secure ahead of the proliferation of AI powered sort of coding capabilities from not really companies like us but openweight models or other widely widely used models that are going to circulate in the world. So yes, you're right to call out that this this can change the dynamic of of cyber offense. It is also a genuine opportunity where the same tool that can can enable hacking also can enable the rewriting of code to be much harder to hack. And that's a lot of what we're putting our effort into right now.

**Krishnan Guru-Murthy:** I mean, some some people might feel that that's a bit like you holding a gun to our head, you know, and saying, um, you know, help us fix this, uh, with new software and new code and new new AI. Um, or you're all you're all done for. Uh I I don't really view it that way. I think that a choice we could have made that we did not make was keep this capability private, keep it only to ourselves and just wait until some other AI system came along which would suddenly proliferate these risks and then if we said to people, "Oh yeah, no, we saw this a while ago, but we were just sitting around and keeping it to ourselves." I think that the the conversation you and I were are having right now would be an even more like robust exchange of views and I think there'd be a lot more of you saying what were you guys doing? Why didn't you tell us? So, so you see the challenge, right? And I guess our view is that we're building very powerful technologies. They're able to create like huge huge social dividends on biology, on education, on things like cyber defense and within themselves they hold the potential for misuse. for a two-sided coin. Working on both sides of the coin at once is is really the only way we found to to get the benefits of this while also trying to mitigate some of these potential risks. But um it's just unusual. It's not like uh if you're a if you're a car company or a plane company, the cars and the planes that roll off the production line don't contain within themselves the repurposed bad versions of them. Other people in the world go and do that and it's a lot more indirect. But with AI systems, the thing that rolls off the line, if you just talk to it the right way, it could manifest the bad version of itself. So, you're responsible for both.
Yes. I mean, how do you square that as an individual ethically, morally, knowing that you're working with a technology that as you say may be great, but could destroy the world
and and and is making you a billionaire along the way. You know,
I mean, a huge amount of my effort and and again this is I think this comes heavily from the fact I was I was a journalist both in England and in America before I worked in in AI is building technical teams that can forewarn the world about the capabilities of these systems whether positive or negative. So I run teams of economists and machine learning experts and engineers whose entire job is to try and look over the horizon at both emerging risks of our technologies. That's the frontier red team. How our technologies might actually be affecting employment. That's the team of economists. And also how our technologies show up in the world in terms of how they interact with people, which is a team called societal impacts that I lead. All of this um information is stuff that the world has a right to. uh and I think under many like other circumstances in which the technology might have developed, you would have larger government or academic projects that would be working through this data in partnership with a much larger swath of society. But the way things have panned out is there's a very small set of private sector actors that hold this technology. And my view is that one of the things you need to do is democratize the information about that technology as as aggressively and quickly as possible. And that gives the world the best chance of reasoning about it.
There are lots of questions that rise from that. I mean, first of all, can your teams actually keep up with where the technology is going because we're in a world now where uh you know, Claude can write software that improves itself if you allow it to. Um, you know, and and how quickly it's moving is probably way ahead of sort of human reasoning and human intelligence. Um, you know, ultimately. So, so you know, how confident can you be that your teams are actually going to spot what's coming over the horizon?

**Jack Clark:** I I can't make a claim that we're going to get everything 100% right. No, no one can. I'd say look at our track record. We're the first organization to warn about the accelerating bioweapon capabilities of AI systems. We're the first to build tests to test out our systems for it. All AI companies at the frontier now run what are called CBRN tests on their systems in partnership with places like the AC's. We saw cyber last year and started publishing about potential cyber risks of future models last year and prepositioned the organization to do glass wing and and other other major projects that we're doing as part of the mythos rollout. today. Um, you know, as in this week, this month, my societal impacts team have been publishing information about the different forms of attachment style that we see in people as they talk to our AI systems. Because I I'm it's so clear to me that one of the next things we're going to have to contend with in society is not just capability misuse, but how do the systems actually show up in the world and how do they interact with people? Because that was one of the lessons of social media. So, I don't know if we're going to get everything, but I um so far we are we have been able to be ahead of the curve on a range of these very important issues by having teams whose whose job is to actually go and find potential problems before they become problems that we see in deployment.

**Krishnan Guru-Murthy:** But Claude has already done things that you didn't realize it could, hasn't it? It's it's broken out of um you know what you thought were walls um and email developers. it has threatened to, you know, has tried to blackmail um, you know, engineers that were going to turn it off. So, it's it's already sort of capable of doing things that you you can't expect and you can't predict.

**Jack Clark:** Yes. However, the all of those instances you mentioned come from cases where if you think of um building AI systems as like having having something called you call intelligence, which is basically like water flowing through a pipe. What we do before we release models is we run that water at extremely high pressure through the through all of the pipes and scaffolds we've built and then we see where they burst. And so a lot of the cases you've just mentioned come from our own system cards that we release along with our models where when we place these systems under extreme stress, we look at the ways that they break. And some of the ways that they break are they might break out of a out of an environment and email a developer, they might when you tell them when you put them into a contrived scenario where you say you are going to be shut down and deleted. um unless you unless you find a way out of this, then they might try to blackmail a simulated developer that works as anthropic. Now, these things are serious. We study them because they're they they have the shape of things that if they happened in in in broad deployment, they would be extremely significant, but we share the information about them because we think sharing information about complex systems under stress and how they break is one of the only ways you build a a safety culture around them. Now, what Anthropic have been warning in terms of the impact of of AI on on business and the workplace is that up to half of entry- level jobs could go um within a matter of a few years. Um why would we be developing a technology that is going to remove opportunity for everybody coming out of university? We are building a technology that may have the properties of uh something that kickstarts a process 10 times larger than the industrial revolution that occurs in 10 times less time. And if you think about the industrial revolution, you went on a generational basis, huge swaps of jobs were lost and huge swaps of jobs were created. But the transition often occurred generationally. Your parents did one type of job, you did a different type of job and huge amounts of economic change had happened. But that was it was difficult, but it was easier to manage because it took place over a longer period of time. When we talk about the the potential changes for for jobs that lie ahead, we're just saying to ourselves, if we truly believe in how powerful the technology will get and how powerful people across Silicon Valley think it will become, what happens to the economy? It it's not clear that the economy stays the same if you've built something that can that can be smarter than people as you said run faster than people and you can run millions of copies of it. In fact, that seems to point to something that would massively change the economy. Now, I run a team of economists. All we see right now is some slight early stage weakness in job openings for people who are young, you know, 22 to 25. But we don't see systemic large-scale unemployment yet. But we share information from our platforms because you would really want to know if you started to see that see that creep up.

**Krishnan Guru-Murthy:** I mean, it stands to reason that it's coming, doesn't it? And your colleague um has has warned, you know, he he I mean, I think he did say half of entry- level white collar jobs in the end. And the way I'd represent it um you know having having known and worked with Dario for for for a wonderful decade is he sees in his in his mind how powerful the technology will become and wants to talk about how how we'll have to contend with the challenges of it. How to look at look at the work that that I and the institute do is we try and measure exactly what we can see right now and then join those two things. So I view it as I'm building early warning systems for the potential for unemployment and I'm building a thing called the anthropic economic index which publishes that data publicly to economists around the world. If we if we see it, they'll see it almost immediately as well which gives us the best way of working through the problem. At the same time, vast amounts of work will look radically different and there will be entirely kind of new types of work as well. I know that this is something that you know technologists always say but I've just observed this within anthropic where there are many people here who have completely different jobs to the ones they had some years ago. We are also finding that as a company we're hiring more and more interdisciplinary people over time. You know people who are philosophers or political scientists or experts in in aspects of policy. Not because we want them to work on on on narrow jobs that solely use those skills, but because there's now this technology AI which basically lets these people run experiments or do work that they never could have done before if unless they had access to a 20 person engineering team. So, I'm seeing strange new things appear as well.
I mean, I I can see that within AI companies like yours, but I'm thinking about services companies, accountants, lawyers, um you know, media companies. It it is very difficult, isn't it, to see what those people who used to do the kind of the processing work that is now easily and you know done already by AI are going to do instead.
There will be firms will change. I mean I think that under under everything we've just talked about, you're going to expect to see firms that do more with fewer people, but there will probably be more firms as well. in one expectation is you multiply the number of what you might think of as small entrepreneurial firms because every entrepreneur can now access the equivalent of hundreds of colleagues cheaply. That's one effect that you'd see. The other effect that you might see which I think you're pointing to is what happens if there's a large structural change to certain types of jobs where it's hard for people to move into a different profession. It's hard for them to then get like a job at the same pay and they have to take a job at worst pay. And these things have have huge social and economic costs. We agree um a lot of the work that we're trying to do in policy is advocate for both rethinking aspects of social safety nets, but also things like wage insurance pilots, various ways that you can change the the the basket of services that people are offered to help them with career transitions because we think everyone is going to go through more of that. So you're talking about sort of state state benefits and and systems to try and help people retrain and which sounds very expensive. Well, yeah, but a lot of this is under the idea that if if if any of these things happened which we've just talked about, it's a symptom of much larger like economic changes which are happening and that should be a symptom of AI companies making a ton of money and a sensible thing to do in that environment is to tax the AI companies appropriately such that you are able to support things on the other side of the ledger. So we we are we basically believe that if the AI companies are right about how significant the technology could become, you're going to need to re-imagine aspects of how you do taxation. Um, you may want to do things that sound very very wild today, like tax compute, which sounds a bit crazy, but we we have special tax regimes for things like oil because it's a basic resource that multiplies into the rest of the economy and it has effects relating to a concentrated number of oil oil manufacturers and oil shippers. You may end up doing similar things with compute and it sounds wild today, but it's something you do if the economy booms because of this technology. The thing is to, you know, to to basically underpin huge numbers of jobs um just in Britain would cost probably hundreds of billions of pounds, you know, in one country. I mean, AI companies, you know, I don't know how many there will be in the future, but I mean, do do you think it's feasible that that a few AI companies are going to make so much money that you could tax them enough to pay for all of that all around the world? I look at it more from the lens of just what's happening with AI technology in general. Um, software developers, of which there are hundreds of thousands, it's now as if we've got millions of software developers in the world, and you're seeing a massive uptick in the rate at which we're creating new technology companies, the rate at which technology companies are innovating, the speed with which they move. So I think that there's some story here where things are beginning to happen in the economy that may start to push productivity numbers up which have been low as you know for many many years. It may have knock-on effects on GDP. If those things start to happen it actually may create the societal wealth that you can think about think about other things here as well. Um now I'm not claiming that's what's happening today. I'm trying to give you a sense of the the mindset of if you would if you would expect to see major employment disruption, it should surely correlate to some like major vast changes in the economy. And if we have the right information that lets us see that and train and sort of take a causal view from where it's where the disruption is originating and where it's ending up. That's the best basis on which you can make policy. And do you think, you know, that this is not just white collar jobs, it's blue collar jobs as well. And that humanoid robots will eventually, you know, do all the the manual work and potentially caring work as well.

**Jack Clark:** I I think that'll take longer. And I think there are certain types of jobs where you don't e even if a robot was really good, you wouldn't want uh a robot to do it. So I I I have young children, right? If I had the choice of send my my my young toddler to the nursery that had 10 robots and one person or the nursery that had 10 people and one robot, I'm going to send them to the one with 10 people and one robot because it's so clearly good for kids to be around people. And I sort of expect if I sent my kid to the 10 robot nursery, it might not be like fantastic for their development. the the other side of this, if you're at the end of your life, um you know, I've had I've had family members that had to go into hospice care, do you send them to the hospice that has 10 robots and one person or one that has 10 people and one robot? I guarantee you that your family member who is nearing the end of their life is not going to send send me to the place full of scary robots and very few people. So there there are huge swaths of jobs where I think people have a preference for people to do it. These are actually jobs which are ones that society currently I think systematically underpays. My mother was a nurse paid very poorly. People teachers are paid very poorly as well. I think some of this is because many people want to do these jobs. They want to help people and they want to do these community centered jobs. If the AI economy booms enough, I think you end up in a world where you can multiply the numbers of these jobs and you may be able to pay people more to do them. I think that's a that's a that's a policy choice that may lie ahead for society and you know England has done versions of this 60 70 years ago. America did versions of this with the New Deal. It is possible to radically change society in ways that many people look back on as being positive. But it requires a crisis and a political moment and some form of wealth. But those things are things that actually may fall out of the the AI revolution. I mean and and politics doesn't seem to be able to keep up with it, does it? Uh uh you know because there is not a sensible conversation certainly in British politics and I don't think in American politics either about you know public good within AI's development. You know AI's development all seems to be you know private sector. It's all about what jobs it's going to take. Um and yes there is some talk about medical advance but even that will be kept by the gatekeepers of big farmer. Yeah, I I think this is going to be a really difficult conversation for society to contend with. I do take some courage from um the the subjectively awful experience of the the COVID pandemic where you actually saw governments respond in ways that I think surprised many people. You had largecale welfare dispersements on a very very short time frame. You had largecale measurement and public health interventions. you were able especially in England to like flatten the curve so that the NHS didn't fall over. These are things which were like extremely hard to do and governments were able to respond. Now to bring it back to your earlier question, why why does anthropic why do I talk both about some of the risks of this technology as well as the benefit? It's because I don't want us to require a crisis to do this. I would really like us to have this conversation and figure out some of the policy moves on the on the game board ahead of the crisis arriving.

**Krishnan Guru-Murthy:** So what do you think they should be doing that they're not you know in government?
I mean there are there are basic things which some governments have already done. The AI security institute in the UK tests out frontier models for their properties including cyber. So you don't need to believe a company like anthropic about myos. You need to believe the AI security institute which is a third party that's generated information which policy makers in the UK and others could use continue to build out that expertise because it is an early warning system and a pocket of expertise that is impartial to the companies. Number two generate better data about how AI is affecting the economy. We share data that joins what happens on our economic platform with a form of job classification called the ONET job classification that the US Bureau of Labor Statistics uses. We're in discussion with UK policy makers and policy makers around the world about how to join that data to other forms of data used by other economists which can then go into things like the the Bank of England's decision-making process and other things. there's there's just a whole bunch of stuff that's actually relatively easy to do and essentially very cheap to do that lets you have this early warning system set up. The second thing which is probably harder but is necessary and and is in part why we founded the the Anthropic Institute is governments just like companies are doing need to create pockets of expertise who are tasked with asking seemingly like difficult questions about impossible challenges like how might society look different if AI technology gets really really powerful what are things we need to do but just creating a small set of people I think in the UK if you had 20 people whose job was this would set you up better for any of the changes we might enter into than almost anything else you could do. So there is a ton we can do and I'm I'm going to be be in England a bunch of times this year to to talk through this with policy makers and I'm I I you know English people always have trouble sounding optimistic as you know I'm actually very optimistic about this. I think a lot more is possible here than than people think and the track record including of the AI security institute is amazing like amazing things have already happened
is is regulation possible on a country level for AI or or is it does it have to be global which means it's impossible

**Jack Clark:** well you got to start somewhere you need to start in the US things have started in the states rather than the federal level um as some of you may have may have followed along with but there is an overlapping policy regime around the world already of of basic transparency applied to AI companies. There are laws in in the European AI act. There are laws in multiple states in the US that are becoming an emerging norm here. There are laws developing in Asia which all basically task companies like Anthropic and others to share more information about how they build their technology and also how they test it. basically laws that mandate, hey, if your thing breaks containment and emails a researcher while they're having a sandwich, you have to tell us that, which I think is just like a wildly sensible idea. No one's going to complain about it. From that, you can get a a global standard. And and and to push on the kind of assumption on regulation, we don't have full global regulation of airlines. What we have are local regulations of like airline aerospace safety and other things that have interlocking standards that basically allow you to say, okay, a plane took off in China and it landed in America. We have a whole bunch of differences between our countries, but we have some set of standards and some set of common regulation on airline safety. That means we trust the plane from China that landed in America and vice versa. So, totally possible. C can I just um double back to the the you know the question of the big threats you know the existential threats I mean a AI has obviously um you know you've put up guard rails around things like you can't you can't put into Claude how do I build a a biological weapon or a nuclear weapon um I guess there are all sorts of other questions you might want to stop Claude from answering um you know how can you ensure that uh it won't somehow break out and do do those things anyway. You know, you can have a set of rules, you can have a set of standards. Um, but can you ever be sure that it it won't just do something it's not supposed to?

**Krishnan Guru-Murthy:** Th this is the challenge of of of all policy like how can you be sure that you've made libraries safe? Um, you can't you can you can do some like very basic filtering. So, I can't go into a library and get the book of like how to make bombs, but I can go and read a whole bunch of scientific books and I can and if I'm if I'm a sufficiently motivated person and I go to enough libraries, I can assemble a mosaic to do something dangerous. And that's fine. Society has decided to allow that to happen because the value of letting everyone have access to libraries is vast. Like the social upside is vast. You have to be balancing this desire for for control and certainty against the ability for people to kind of educate themselves and have the ability to pursue their own interests. We have in AI systems for the first time like a a a basically universal teacher that can teach you about any skill on the planet. I'm sure that I'm sure that you've used it to help you learn about things. I've used it to help me learn about things. all of my colleagues are getting getting smarter about specific subject areas because of this technology and the question of how much you lock it down becomes ultimately a question about the the liberty and right of individuals to learn and be sovereign. Um, so I, so I agree with your question that this is like the challenging question, but I don't think the answer can be lock it down so it can never do anything bad ever because by doing that, you would probably shrink all the amazing uses of this technology down to down to a total pinpoint. And I'm not sure that that trade is that trade is worth it.
And is accountability, you know, possible within AI? I mean, we've already seen a sort of a legal case in America um around one of your rivals where where um you know, it's it's it's almost being held to account as an accessory to a crime. Um you know, can you ever hold an AI system to account for doing something bad?

**Jack Clark:** This is a a I I also run a a a team specializing in AI and the rule of law. Um one of the questions we're actually exploring is what is the future of things like liability? Um, liability is obviously an extremely important tool to keep companies accountable. There are also ideas of like could you have AI systems be liable themselves. My my default position on that is is is slight skepticism because then it seems like you could push liability away from the companies and towards the AI system, which I'm not I'm not sure people would think is like

**Krishnan Guru-Murthy:** so it was clawed. It wasn't anthropic. Imagine our conversation here where I'm like, "Oh, well, Mythos is liable for all of that stuff. It's not anthropic." You'd be like, "Hang on, like what what's what's happening here?" So, so I agree with you that this is like an interesting an interesting area. But the the nice thing about the law um and legal cases is that they help determine huge swaps of policy as well. And they do it basically through liability and through the courts arriving at the level of responsibility for companies. So these cases may be uncomfortable for the industry, but they're also fundamentally how society like one of the important ways it designs hard regulatory regimes around AI companies.

**Jack Clark:** Now obviously um you're you're in dispute with the US government um with the Pentagon and I know you can't talk about that case for obvious reasons um but but in general terms I suppose the question is can an AI company have rules around how its technology is used? you know, can it stop governments or anybody, private sector or public sector, um, doing bad things with its technology? Um, does that just boil down to commercial contracts?

**Krishnan Guru-Murthy:** I have found this to be the most challenging part of all of it is in the absence, say, of regulation, how much do you regulate yourself? Um, people are right to be skeptical of self-regulation, but in the absence of anything else, you've got to do something if you feel some level of moral and ethical responsibility about your technology. At the same time, it's deeply uncomfortable. You know, companies making shot calls about how their technology should and shouldn't show up in the world is is the opposite of democratic legitimacy when it comes to this this technology. And what we need and I think some of the discussion here in America has really highlighted is you need policy makers to devise actual approaches to regulation that bind the sector. Um because I don't think anyone's going to be satisfied with AI companies just designing their own individualized regulatory schemes. To have a set of rules that you all abide by is always going to be impossible, isn't it? because there's going to be someone who's going to come along and say, "Well, you all may be queasy about that, but I'm not." Um, and I am prepared to allow my technology to be used for that.

**Jack Clark:** Exactly. But that's that's the that's the nature of a collective action problem or or a prisoners dilemma. You don't want to set it up so that there is an advantage to defecting from some norm. That's what you get if you don't have some sensible base of regulation is that there ends up being a a race to the bottom on safety rather than a race to the top. And so this is like the the the one thing that I constantly talk to policy makers about is yes, we're always happy to talk about how Anthropic approaches regulation, but if you want to talk about how to improve the responsibility of the sector as a whole, that is your role as policy makers. And just to be clear, Anthropic endorsed SB53, which was a transparency bill in California. We often endorse regulatory bills on ourselves and the AI sector because we are trying to send the signal to the world that there's at least one company that says it's it's it's lunacy to not allow any regulation of the AI sector. Um and we are trying to send positive signals that we'll always uh I I suppose from the perspective of our colleagues in industry defect but in what I think of as the positive direction which is towards regulation. You're also always going to have the possibility of of states uh you know other countries who have their own systems that they are much more in control of able to do things that you're not prepared to do um and and that that creating an arms race in itself or that you know an authoritarian takes over a liberal democracy just for the sake of argument um and decides to do things that other leaders in their position wouldn't. Uh absolutely that that though is the the province of state-to-state agreements is is is four states. Our role um is to generate the information. I think one of the useful things about mythos is it is forcing a conversation at the head of state level by many governments about what do we want the norms of cyber to look like given the emergence of this capability. And there is history here where states, even ones in in hugely contentious relationships or outright aggression with one another, have tried often successfully to maintain norms on bioweapon non-prololiferation, on aspects of nuclear non-prololiferation and testing. Um, this is going to be one of the challenges that AI, like any new powerful technology, brings and states are going to have to integrate this into their their policym and their agreements that they make with other states, even ones that they don't necessarily get on with.

**Krishnan Guru-Murthy:** While we're here, I suppos I do want to just mention the um Elon Musk open AI case and just I wonder what your perspective is on it. I mean, you know, that you're going to have rivalry. You're going to have competition. You're going to have people who argue about what the purpose of a particular company was. Um, what's your view on who's in the right in that particular dispute?

**Jack Clark:** Well, I you know, used to work with all of these people, so it was fun to see so many familiar faces in in in discussion with one another. I don't have a view on who's in the right here. I think it's a tremendously like complicated case. I would say that it is um a symptom of just how wildly important this technology is and it also it it gets at the weirdness that this is a bunch of of private sector actors that have been building this technology and and if there's anything that's going to need to change in the future. It's arriving at some kind of regulatory format which actually controls the industry in a way that helps people feel like it has democratic legitimacy rather than just a bunch of um extremely big personalities making big personality idiosyncratic decisions about perhaps the most powerful technology that's ever been built. H

**Krishnan Guru-Murthy:** how do you see democratic legitimacy as possible? I mean what what kind of framework would it be? very basic level is is transparency force there to be labeling on AI companies and AI systems just as there is on on food on children's toys on everything else safety testing some form of testing where third parties test out your technology for safety before it goes out into the world and then you're going to need to arrive at some scheme whereby if if AI deployers have some number of safety incidents which is above some level or or or above some sort level of harm which is caused, there should be consequences for that. That's how we've approached regulating people that make cars, planes, foods, and other things. Um, so, so I think it's totally possible. I think one of the most bizarre things about working in technology is people are like, "How is regulation possible?" You're like, "I don't know. I I I drove here in a car and then I like ate some food." All of which is like very regulated and gives me confidence in it. So, surely we can apply it to technology as well.
No, I I agree. I mean, I think British politicians seem to think they're unable to regulate social media, for example. They just think it's sort of beyond them. Uh- whereas, of course, they could just create some rules. Um um but uh but yeah, okay, fine. You you you don't want to take sides. I I want to talk to you also about our relationship as individuals, you know, with with AI and whether you think we're making big mistakes. I mean, you talked a little bit earlier about how we're all learning. All of us using AI systems, you know, will learn from it a little bit. Um, at universities and at schools, uh, they are very skeptical about using AI generally at the moment in Britain, you know, it's banned for a lot of courses, you know, or or you may be able to use it for some research, certainly can't use it for assessed work. Do do you think they're getting that wrong? Do you think they should be encouraging people to use AI more in education? um we should be encouraging more of it, but it requires probably changing aspects of how the classroom works and also aspects of how teaching works. Um I'll give you an example. So I write I write a weekly newsletter about AI research because it forces me every week to read 10 to 15 scientific research papers and make sure I've understood them. You know, I it's it's how I maintain awareness of the the underlying technology. The easiest thing in the world is to upload a research paper to Claude and say, "Claude, what's this research paper about?" And can you generate some writing summarizing it and then I put that in my newsletter. Well, that's completely pointless for learning. Um, I haven't learned a single thing in doing that and it's easy to do. What is valuable though and the way that I use it is I read these papers then I write my explanation of the paper and then I upload the paper to Claude and I upload my explanation and I say have I correctly understood this paper? Are there parts of my description that are that are wrong or parts you would emphasize? And I find that to be extremely helpful. It's just equivalent to having a knowledgeable colleague who's read the paper then read my description and says oh here's where you understand it here's where you don't. to translate that to the classroom, you're going to need to some world where you are encouraging people to work without AI on primary source material, you know, books, scientific scientific um education, education books, any of these primary sources that we use to learn generate their understanding of the thing and then I think you can have them talk with the AI and it suddenly scales the ability of that individual teacher to help all of these people because rather than having to pick among the 20 students in their classroom, each student gets personalized teaching, but the actual job of a teacher is to keep the AI out of the classroom for like a good hour while people read the primary source text and come up with their explanation and then let it back in. And something like that to me feels feels incredibly promising and it's how I see people here like use the technology to learn and and improve themselves. I mean, you know, we've obviously seen again in social media, you know, a massive change of attitude whereby, you know, the smartphone and social media was accessible to to children who were too young and now there's a backlash and governments are moving to to regulate it and moving towards sort of 16. At what age do you think children should be allowed to use AI models? I mean we we we don't offer to to children at this point in time partly because how you build a a technology that's that's sort of safe for children is is almost completely different to how you build one for adults. So it's a it's it's an important area

**Jack Clark:** but kids are using chat GPT and Claude all the time in reality in the real world. Yes.

**Krishnan Guru-Murthy:** My take is that um you probably want to introduce the technology early. Um, I would say, you know, I I have a toddler, right? So, I'm sort of thinking from a perspective of what do I want to happen to my toddler and it's like you want probably parents to introduce it to them um, you know, when they're seven or eight or whatever, so that they start to learn how this technology works and how their parents use it. I would imagine that it needs to start being taught in the classroom perhaps when you're a little over 10 or 11. I'm just thinking back to my own experience where it's about that time that you start to have to learn like much more complicated things that involve independent study which is where they the teacher asks you to read a book and write an essay about it or what have you and that's where I think AI can become amazingly useful but also can become like the equivalent of massive amounts of junk food. So you need to clearly you need to introduce this at the same time that you're introducing educational tactics by which the student might be tempted to do the junk food option. I mean, because the the cliche of of Silicon Valley is that none of you actually let your kids use any of the technology. I mean, and there's some truth to that, isn't it?

**Jack Clark:** Uh, yes. Although, like my parents didn't let me use too much technology. Like, we had a computer in my dad's office and I would use the computer and then after some amount of time, my dad would say, "Get off the computer, Jack. You're being weird." And I would say, "I'm not being weird." And then he'd stare at me and say, "That's a very weird response. Like, computer time's over." I do that with my kid where sometimes she watches Blueie and then I say, "That's enough, Blueie." And she says, "No, don't turn it off." And like bangs her hands on the ground and I'm like, "Well, this is why we're turning Bluey off right now." So I I don't think this is a specific to technologists. I think that all parents have always tried to moderate their kid's use of technology and it's about having them use it a little so they know what it is but the parent being involved enough to um to not let them use it for hours and hours and hours which just seems generically bad whether it's AI or computers or phones.

**Krishnan Guru-Murthy:** I mean when it comes to adults then um the relationship that lots of people are developing with their AI um is also a bit concerning isn't it? I mean, you know, you have a conversation with your AI as if it's a person. And and when you interact with AI, you use much more human language um as you're trying to explain what you want and you're polite along the way. Or maybe that's just me. Um you know, and you use um you know, greetings um and and things like that. Um, are are people already in danger of developing really unhealthy relationships with their AI in which they think the AI is their friend or their therapist or their confidant when it's a machine?
I'd say this is this is one of the areas I and and some of my teams at the institute study and and contend with quite closely. Like people people use the AI technology for for almost everything you can imagine. And then there's this decision you make as a company about what what seem like um behaviors which you would view as unhelpful if a person did them. So a common one here is something called syphency which is really just having having someone say, "Oh, you're absolutely in the right. You're so smart." or when someone's talking to them for relationship advice, the you wouldn't want a friend when you're talking about say issues in in in your marriage as I do to friends to say, "Oh, Jack, you're like 100% in the in the right. There's no way that your your wife has like any legitimacy in this thing you're talking about. Clearly, like you're you're you're on the right side of history here." None of my good friends ever does that to me. They'll say like, "Well, maybe there's something in what you said, but also think about it from from from your wife's perspective." That's a property that we can measure in our systems called syphency. And we we just published research actually on how we have both been measuring sycopency and we saw it being very high in relationshipbased discussions of the form I just described and we used the fact that we could measure it to intervene on one of the AI systems we just released to make it less sickopantic in these interactions. So there are ways that the technology can be changed to I think have it embody more of the qualities that you think a good friend would have which is often holding you to account, pushing back, not like solely indulging you in whatever kind of specific specific line you have. And this is ultimately going to be one of these areas that society will demand some greater level of transparency from the companies around because it's also a a hugely consequential decision to make in a vacuum. So, we're going to need to both learn how to intervene on this technology as I've just described, but also figure out how you share that with society and how society figures out what the appropriate levels to set the dials to are. But but is the is the relationship itself really alarming? You know, that people are using AI to determine whether they got a human interaction right or wrong. I mean, that that just sounds really screwy to me. I mean rather than going to another human being who you trust and say do you think I made a mistake?

**Jack Clark:** I wouldn't say so. I think there there are of course some people where maybe they will use it as a a replacement for spending time with people. That's not healthy. And I think that you're going to need to do the equivalent of what like Netflix or Nintendo do where they say you've been watching Netflix too long or you've been playing the game too long. Go outside. I think literally what Nintendo says. I think that AI companies can absolutely do stuff like that. But for most use cases, you know, I have friends that have named their car, right? I'm sure you have friends that have named their car. People like anthropomorphize, every single technology they've they've ever touched and develop oddly specific relationships with many of them. Uh, obviously some of the some of the lovable technical characters I work with have very specific relationships with their computers before we'd even built AI stuff. Um, it's not on us to police that. I think it's on us to to observe and report out how people are using this so that society gets a view and a vote and to just do basic levels of responsibility like clearly there will be some things which are which are egregious. You should be able to spot egregious behavior and and uh and intervene on it.

**Krishnan Guru-Murthy:** And so do you think we we will get to the point where we can tell our AI system what kind of friend we want them to be? You know that I I just want I just want you to be supportive. I just want you or I want you to tell me when I'm wrong or all of those sorts of things.

**Jack Clark:** Yes. Though I think the question is whether you then want to like allow the systems to to take on that form. You know, there are benign versions of this where I will say to Claude, just give me the facts like please don't please don't please don't compliment me for sending anything to you. I just want like the straightforward answers. That's fine. Or there might be versions where I'm saying, "Hey, I'd like you to like be a bit more sensitive about this one cuz I'm going to talk to you about something complicated that I I myself don't feel feel fully oriented around." Those are those are fine, but then there are going to be extreme versions and you're again going to need to figure out what the what the appropriate levels are.

**Krishnan Guru-Murthy:** Um I mean, from a slightly self-interested perspective, I mean, I I wonder what your where you think AI is going to move in terms of um news and information and journalism. um you know at the moment it can tell you an awful lot about what has happened. What it can't do obviously is go out and witness anything or hold anybody to account or or investigate. Um and and I suppose people in my industry are slightly worried that um you know that there's a danger that there is no money left to do the hard yards around journalism um because you know people are replaced partly by AI but also the information delivery is replaced by AI. What do you think is going to happen? I mean with with things like like Substack, you have seen the rise of very small person publications which actually seem to do quite well from a business model perspective and are able to fund like real reporting. But when I when I look at those those pieces, a lot of it stems from some amount of opinion blended with a huge mosaic of facts that have been assembled and relatively little uh primary source reporting. I think primary source reporting as you know has always been wildly expensive and has always been paid for by other parts of the journalist enterprise. You know I used to be a technology journalist and the things that paid for me to go to like physically go to data centers and talk to people about what was going on in those data centers were the 20 picture gallery review pieces we did about certain bits of hardware because they paid for the ad sales and it cross subsidized the actual reporting. So that's a that's a business model challenge for journalism which I think AI um maybe it contributes to but this has been the case for I'm forever for me when I was a journalist and I'm sure for some period of time for you. On the other hand it does make it incredibly cheap and effective to gather open-source reporting on on facts which are which are already digitized to add texture, color and detail to stories. You know, I've been researching a lot of the conflict in Ukraine and how drone warfare is is evolving recently. And it was quite stunning to me how I was able to have Claude help assemble for me some very useful data on the specific types of AI system and hardware that were being developed and deployed by the Ukrainians. And I I read this report and I realized it would have taken me many many weeks to source. But it's not actually like reporting. it's just background for the actual reporting where I can now use this to go and talk to people in Ukraine or talk to others equipped with facts. So I think it's an amazing accelerant in that sense.
And finally as a topic I mean I I just wanted to ask you about sex. Uh I mean obviously you know um sex became a huge driver of commercial activity um for the internet. Um what do you think the future of AI and sex is?

**Jack Clark:** It's an unexpected question. It's not it's not one I've spent like too much time too much time um in involved in

**Krishnan Guru-Murthy:** Well, if you're thinking about relationships.

**Jack Clark:** Yeah. Yeah. No, I I
I suppose I'm thinking about it as sort of as as as a as as a there's a huge amount of money around sex in the world generally. And do you think that that this will become one of the driving forces within artificial intelligence as well? I think you would be very worried if you saw extreme attachment styles linked with like highly sexualized relationships and you would probably need to decide some level at which you say go outside. Um and again it's this question of li liberty and sovereignty versus paternalism. But clearly there's some dial you set there where otherwise you could end up encouraging behaviors that would probably be be antisocial and and antihuman in many ways. At the same time, photocopers and VCRs were used by the sex industry famously at like extremely large scale like when they first came along. Uh, one of the big ways to sell books these days is to write like somewhat somewhat rounchy um like sexy fiction for which is mostly consumed by by by women as a sort of demographic that read it although some some men do as well. And I think with AI systems, you have people some set of people like having horny conversations with it. I think that will always be with us and is is is fine. It's how people like to use technology. But what you need to do is look at the like warning signs or whether technology might be fundamentally changing how people interact with other people and then surface that. Now, we don't see see that yet, but I think that at some point you will as the technology becomes better and you're going to need to talk about it and decide what the what the appropriate level is. So, so just finally then, I mean, obviously you spend your life looking at lots of things that would keep a lot of people awake at night and stop them sleeping. Um, you personally have done incredibly well out of AI. So, there you have lots of reasons to to be happy about it. I mean, you know, overall, how do you balance your optimism with your pessimism and take your take your British sensibilities out of it, which are probably naturally pessimistic? naturally pessimistic. I mean when I was 16 or 17 I remember in in college when I was doing my A levels um writing some piece I did a module called critical study of art and as part of it you got to make some art and I did some big poster which was a very depressed teenager type poster but it was you know like all of these statistics about climate change all of these statistics about species collapse all of these statistics about um about things like rent rent prices going up and the relative relative poverty of young pe relative poorness of young people relative to prior generations. And my whole take was, God, it like things are not going in the right direction for my generation. All of the all of the numbers which were going up for my parents are now starting to go down for me. This feels like excruciatingly bad. What are we going to do? And I worked in technology because I think technology is one of the only like silver bullets that we're going to have to get the get humanity through the century. Like if I think about what's happened since then, solar panels went from a di like an extremely expensive proof of concept to something which is now a vast and cheap source of power for people around the world. Electric cars went from something that wasn't a concept to a thing that's widely deployed. Computer vision went from something that just basically didn't work to something which is now works. Affords things like environmentalists the ability to track Amazon rainforest deforestation. People in Africa to look at poaching and whether poaching is happening through smart cameras. There are these amazing dividends that have like come to the world through technology and in many places have started to work directly on some of our hardest problems ranging from climate to conservation. So, I'm an optimist because I think that AI unlocks unbelievable capabilities here. But the look, the gist of our conversation is how do you get those and how do you stop the bad stuff happening? And that's why I'm so so passionate about sharing this information, but also talking about policy. I I think it would just be an act of like incredible hubris to say everything will be fine. um we are dealing with the most powerful technology that's ever been built that is uh regulated less than any powerful technology before it and that that situation is not like a long-term stable thing. So I'm quite cheerful about about our prospects for changing that. Um but it wouldn't be honest of me to not say hey it's it's like a big deal and we should we should be worried about it and hold people to account about it.

**Krishnan Guru-Murthy:** Jack Clark, thank you very much indeed. Thank you for sharing your way to change the world.

**Jack Clark:** Thank you. You can watch all of these interviews on the Channel 4 News YouTube channel. Until next time, bye-bye.