The Hardest Problem AI Ever Solved, with Google DeepMind CEO

[00:00:00] Speaker 1: something's obviously not quite right about the definition of intelligence if we play this out what's the limit here the best use case of ai was to improve human health it was the moment i'd been waiting for that could achieve something no other system could i want to use ai as a tool to help us understand the nature of reality around it governments are going to use ai what would you hope that they use it for there's two things to worry about one is that's demis hasabis the ceo [00:00:27] Speaker 2: of google deep mind nobel prize winner he is one of the most important people alive on what is quickly becoming the biggest technological leap in our lifetime because the biggest way that ai is going to impact our lives isn't something that we can see it's not a chatbot it's not an image generator it's tools that are invisible to us in drug design and natural disaster detection and nuclear fusion and quantum computing tools that he and his team are building here he is winning the nobel prize for just one of those tools so who he is and what he chooses to build matters a lot for you and me and he's fascinating he's a childhood chess prodigy who at 17 turned down a reportedly million dollar job offer from a gaming company to go to college instead and then got a phd in cognitive neuroscience he founded his company deep mind with a mission to solve intelligence starting with beating video games he then sold that company to google specifically because they promised to let deep mind focus on scientific research but as this has turned into the most intense technological battle in recent history demis is now in charge of much much more he's now behind basically everything google does in ai he's making decisions that affect your life and millions of other lives every single day so what is he planning to do with all of that power my goal is to show you the future that demis hasabis wants to build so that you can decide for yourself what you think of it thanks so much for doing this it's great i really appreciate it you already know that huge conversations is a different kind of interview i'm not going to ask you about financials i'm not going to ask you about your management style i all well covered elsewhere what i'm hoping to do in this conversation is think about it more like an explainer that we're making live together and i have some props this is not actually meant to be a jenga game um each block represents a project or a model and i want to talk about them and how they fit together wow so they were meant to be visual aids but as we were setting up we started playing jenga with them and it turned out to be way more fun than anything i had planned [00:02:53] Speaker 3: also i know that you like games yes i love games so this is great first in an interview anyway yeah so [00:02:59] Speaker 2: my hope in this conversation is to make this explainer together and to help people see what's happening right now in ai really and what is the future that you see coming what are you hoping to [00:03:11] Speaker 1: do in this conversation a lot of the reasons that i got into ai 30 plus years ago now is to um advance science and medicine and i've always thought of ai as potentially the ultimate tool to do that so i'm hoping we're going to talk about that today and really that's been my passion for what to apply ai to although of course it can be applied to many things oh this is going to be a lot of fun yeah so in this [00:03:34] Speaker 2: jenga game that we have a lot of these are blocks that people will have heard of right these are you know this one is gemini right but i would argue that the ways in which ai is meaningfully shaping people's lives most are the things that are invisible to them most of the time so i want to start by talking about the project that you won the nobel prize for yeah good jenga playing i want to tell the story of alpha fold with all of its drama because some people might not have heard it but then i want to get really quickly to the cutting edge of this sort of category of science why did you decide to tackle this problem out of all of the many well i came across it actually [00:04:15] Speaker 1: uh as an undergrad in cambridge so i had a lot of biologist friends and one of them specifically was obsessed with what's called the protein folding problem so proteins are what everything in your body relies on they make biology possible and living possible and what's important about them is their 3d structure so in the body they fold up into kind of 3d structures and those structures determine what function they have or partially determine what function they have and so the protein folding problem is really about can you predict this 3d structure just from the one-dimensional amino acid sequence so that's the kind of 50 year grand challenge of protein folding so i love challenges i love puzzles so i couldn't resist it from a scientific point of view as this probably you know is described to me as the equivalent of fermat's last theorem but for biology so who couldn't be interested in that but also when i first heard it i i thought um the kind of problem it was would be suitable for ai one day even though we of course this is in the late 90s we didn't have any kind of ai that would be possible to work on this but i thought one day that would be possible and then the final thing was just the impact it would make if you cracked it because it would open up all these downstream possibilities for research and especially in things like drug discovery and understanding disease so which i think is you know the most important thing to apply ai to is improving human health and the reason that [00:05:35] Speaker 2: this would be huge for human health is that up until now in order to develop new medicines we'd had to spend hundreds of thousands of dollars and years of human effort to find out the structure of a single protein by shooting x-rays at it so we had figured out some protein structures but it was slow and expensive so i'm skipping over an enormous amount of hard work here by you and your team but i think by the way that i'm asking the questions it is very obvious to people that you solved it yeah so there's this moment where you realize that it is genuinely useful and you have solved what had been called one of the most important unsolved problems in modern medicine and it's 2021 you're in a meeting i am so glad that there was a camera in this meeting yes randomly it is one of the most incredible moments i have ever seen can we use alpha fold to solve it i think you're talking with your team about setting up a system where scientists could send in a request for a specific protein like a website and then get the protein folded yes and then someone else has a very different idea yes can you walk me through what happens in that meeting and then you your reaction is incredible uh and i really want to know [00:06:47] Speaker 3: what you were thinking yeah sure well look we we it's it was funny that the cameras were there happened [00:06:52] Speaker 1: to be in that in that particular meeting it's crazy that was that day you know they very rarely followed us but it was for that meeting and um normally what happens for these sorts of uh prediction models is you the traditional thing is you kind of set up a server and then other scientists send you their protein sequences and they say oh i'm interested in this protein can you send me back the the predictor structure so you know uh and that's how it'd been done in the whole field for the last 40 plus years um and the reason is is because most of the prediction algorithms are quite slow so maybe it would take a few days and then you'd get back the you know you'd email your email back the structure um and then you'd ask for you know you'd ask for the next one but once i realized sort of in that meeting actually that how quickly uh not only how accurately we could fold the proteins but how quickly you know in a matter of seconds and then i was just sort of doing the back of envelope calculation like how many proteins are there known to science known in nature 200 million and then how many computers do we have and uh how many would we need and then you know per and then if we fold one every 10 seconds and i sort of realized sort of in the middle of that meeting i was fiddling on my phone that it would be possible in a year and so why go to all the effort of building the servers and the data betas and and you know the emails uh uh client and all of that when we could just actually fold everything ourselves everything anyone could ever request and ever want and then put it on a database somewhere for free for all the scientists in the world to use so just suddenly hit me we should just do that now why don't we just do that well so that's one of the options like we right uh you know there's this we should just what we should that's a that's a great idea they should just run every protein in existence and then release that suddenly all these things must have been going on the back of my mind and i suddenly realized that that would be the obvious thing to do and it would be actually probably less effort than standing up the server so actually it would save us time and you in that meeting your [00:08:53] Speaker 2: reaction is something like why don't we just do that that would be way better we should clearly do that and then you do all of a sudden this crucial process that had been so hard is suddenly fast and easy and it's being used by scientists all over the world this huge unsolved problem now solved is it correct to say that we have now predicted the structure of [00:09:18] Speaker 1: almost all proteins known to science yes and we keep updating it so every time you know somebody scoops a pale out of the ocean somewhere and then you know there's loads of different types of organisms in that bucket of water seawater and then they sequence them all and so the sequencing technology has obviously uh improved many orders of magnitude since the human genome was sequenced so now the problem was the structural biology the finding these 3d structures was taking was far lagging far behind the genetic sequencing so now with these computational resources like alpha fold 2 we can actually keep up with oh here's a new million uh genetic sequences from some new strange organisms we found uh oh here are the structures and so uh we have a kind of a small team uh at the at the european bioinformatics institute that keeps updating every year you know all the new sequences that have been found that year so we're now you know always at the cutting edge like we know what all of these different uh uh protein structures mostly uh uh look like that's so awesome it is pretty amazing it's especially amazing actually for the people researchers that work on um slightly more obscure you know organisms or animals and things like uh for example wheat i found out a lot of plants have way more uh genomic data than mammals and humans which is very strange they seem to have multiple copies of of of of their genome and things and it's uh it's it's it's it's it's a kind of strange and bizarre world i think the plant world but my plant scientist friends of mine is like you know they they they don't have the resources um like with you know human genome there's a lot of works being done on that but some of these more obscure organisms um that's still really important for humanity you know like crops and things like that uh now we're able to um immediately jump to the science around what they want to do with the proteins um maybe you know help them be more resilient to to climate change things like that and uh and and and they can jump straight to the problem they're actually interested in rather than getting bogged down with trying to crystallize the proteins that they're interested in another boon is for um researchers who work on neglected diseases uh that affect primarily the more developing parts of the world things like malaria or chagas disease or leishmaniasis these that affect you know hundreds of millions of people around the world and um but there's not uh a lot of money in that if if big pharma um try to research that and find cures because they're in the more poorer parts of the world so they tend to be neglected the research that goes into that so there's these amazing uh non-profit organizations that do the research on that uh but they don't have a lot of money or resources so um giving them the structures of the proteins that are involved in say malaria virus uh is a huge boon for them too because they can go straight to the drug discovery phase that was one of the hardest [00:12:11] Speaker 2: things to figure out as i was doing this research because there's this moment where scientists all around the world have access to alpha fold you can see like the map lights up you can see that people are using it yeah but i wasn't easily able to figure out what a great example would be to talk to you about of a scientist using alpha fold and then that speeding up a drug process that results in a drug that i could now take yeah what is your favorite example of a scientist using alpha fold for [00:12:37] Speaker 1: something the audience might understand or have seen so over three million uh scientists are now using alpha fold we think it's pretty much every biologist in the world at this point uh and one actually scientist that uh at a pharma company said to me uh that uh you know almost every drug developed from now on will have probably used alpha fold in its uh in its process which is sort of you know mind-blowing really and amazing that that's that that that impact it's having so um but it still takes time with drug discovery so we're still mostly in the fundamental biology stage of understanding the disease what is the protein we're targeting is that the right biological mechanism and then uh as i understand it some of these uh drugs are now in kind of the the the clinical trials phase and then uh hopefully we'll see in a few years time mirror you know whole dozens of drugs that were partially helped by at least uh alpha fold and um in terms of my favorite breakthrough is more that so far has happened with the help of alpha fold is there's this protein called the nuclear pore complex and it's one of the biggest proteins in the body it's huge for a protein and what it does very important job it's basically the gateway that opens up and closes to let nutrients come in and out of the cell nucleus so it's it's basically the gate it's like a big donut ring and that opens and closes and um but we didn't know until very recently what the structure of this was because it's so big and complicated uh it's pretty hard to crystallize and actually see and so uh almost i think it was pretty much six months or a year after we put alpha fold out uh some teams used it along with experimental data to finally work out what this beautiful shape was of this uh this gateway protein and uh that was amazing to me it's you know one of the biggest proteins in the body that's um and and alpha fold was was very useful in helping determining that structure and so perhaps we can design drugs or [00:14:34] Speaker 2: treatments that better that that use that somehow that yes better access the yes potentially i think that [00:14:40] Speaker 1: was more for fundamental biology understanding but but uh obviously there is uh i mean we ourselves we've spun out a new company isomorphic labs that uh actually uh tries to build on alpha fold and uses it to indeed in this in this block here uses it to uh as as one of the pieces of the puzzle to um massively speed up drug discovery so on average you know it takes like 10 years to uh uh to to kind of develop a drug it's crazy long time unbelievable amount of hard work very expensive and huge failure rates you know only about 10 percent of drugs actually get through all the clinical stages so we need to vastly improve that if we want to improve human health i think and i think the way to do that is by using uh in silico methods alpha fold two being uh one of those components um but knowing the structure of a protein is only one small part of the drug discovery process you need a lot of chemistry like what compound should you design to bind to it all of these things so we're trying to isomorphic um build these you can think of them as uh adjacent systems that work with alpha fold more advanced alpha fold alpha fold three alpha fold four you could call it and um and then uh end to end basically create these drugs that have you know very minimal kind of side effects and uh and incredibly effective at addressing uh uh the type of disease we're we're trying to help with and we're working on you know i think at this point like 18 19 different drug programs um across the gamut of things from cardiovascular heart disease to cancer to immunology so um i think eventually these these types of technologies should be able to help um across almost every therapeutic area [00:16:27] Speaker 2: in prep for this i did a background interview with your fellow nobel fries winner john jumper he really stressed that it's one part of a larger problem of drug discovery and so that brings us to the cutting edge today i've taken some of the some of the examples that i want to talk about yes um [00:16:43] Speaker 1: what is the cutting edge now sure so we're building many different components that can kind of go together so alpha fold is one of the linchpins so that's the structure of the protein but if you think about it let's say you understand what the shape of the protein is okay and then now you know which bit of the protein is the important part that does its function so now if you think about drug discovery so say you want to block the effect of that protein or enhance it in some way you know you need which part of the protein surface do you have to bind to so now you have to discover a chemical compound that will that will attach to the right place on the protein right and and you want to know how strong will it attach and will it and then on top of that even more important is not just will it attach to the thing you're interested in make sure it doesn't attach to other things because if it does that will be toxic toxicity so you don't want it to have these side effects we call it side effects with drugs so you want to minimize those so but because now we have all of these amazing um algorithmic tools we can sort of do a virtual screen of like oh here's a compound one of our ai systems is designed it binds this is our prediction of how strong it binds to the protein surface and then we can check that very quickly like in a matter of hours that particular compound how does it attach to any of the other 20 000 proteins in the human body and so we can just do it like that you know within a few a few minutes and then um modify keep modifying the compound so that it has less and less side effects ideally none on any of the other proteins but um increasingly strong effect on the one that you want so you can see i've just outlined a self-improvement process or self-modification process and this is extremely fast and efficient if you can do it in silico and then um you know on on on computers and then at the last stage um only at the final stage do you check it in the wet lab so you still have to validate it you can't you make all these predictions you do all your search in silicon but then at the final stage you check your final uh uh proposed compounds in the wet lab um and then check it really does what the predictions say but that you can imagine would save you can search thousands of times more compounds or maybe even millions at some point more quickly and efficiently that way and then just at the end check that they're correct that's so much more efficient than doing the search in in the wet lab which [00:19:07] Speaker 2: is what effectively is done today oh one of my favorites also is alpha genome yes so i reached out to yet another noble prize winner um dr jennifer doudna who i've had on this show fantastic and she sent a question for you so i'm going to read this question from dr down okay okay so she says crispr the gene editing technology that she pioneered can now target nearly any dna sequence but for most genetic diseases we still don't fully understand which changes in the dna are actually driving the problem especially in the 98 percent of the genome that doesn't code for proteins with tools like alpha genome starting to decode that 98 percent how close do you think we are to the moment where ai can reliably point to the exact genetic change causing a patient's disease so that technologies like crispr can fix it yeah what [00:19:55] Speaker 1: awesome question so um you know i've discussed this with her actually in the past and and it is really exciting i think with alpha genome which is exactly that kind of technology it takes the big long genetic sequences and then it tries to predict you know if you have made a mutation to this particular single letter singles position in the genetic sequence will that be a harmful you know mutation that might cause disease or is it benign and it won't do anything and alpha genome which we just released is the best system in the world for predicting that so that's exactly what you then want if it got um it's still not probably good enough yet but you can imagine a future version of alpha genomes that are accurate enough to sort of really know like oh that particular mutation in combination with this other one that's the hard part is what if they're multigenic diseases where there's cascades of mutations cause the problem those are even harder to detect but actually perfect for sort of ai to try and help with then um you could go in with something like crispr maybe one day and go in and fix that mutation um and then fix the problem so that would be so a kind of combination of things like alpha genome and crispr could be incredibly powerful and hopefully one day we'll be you know collaborating with with the likes of [00:21:08] Speaker 2: jennifer on that last year you said something to the guardian that i found really interesting you said that if i'd had my way yes i would have left ai in the lab for longer oh sure and the quote is done more things like alpha fold maybe cured cancer or something like that from the outside it looks like the story goes you found deep mind with the mission to solve intelligence and use it to solve everything else yes and then you sell to google specifically because they will allow the freedom to explore science in this way yes and for a long time that's your exclusive focus yeah and then chat gpt comes out google goes code red and you become the head of all google ai including the consumer products that you weren't spending as much time on before and it feels to me like watching that from afar it mirrors somewhat the larger experience of ai which is just this incredible change yes um in the last couple years yeah what was gained and what was lost in that change yeah i think um that's exactly right what [00:22:13] Speaker 1: you describe is is sort of what how it felt from the inside too and um for me uh as i mentioned earlier was that the ai uh the the best use case of ai was to improve human health and accelerate scientific discovery in fact for me i i i've got i got into ai in the first place because i was interested in all the big questions in the world the nature of reality nature of consciousness these kinds of things and um i felt we needed a tool to help us even the best scientists to help us make sense of the amount of data and information out there and find insights in that and that's happening which is amazing and obviously alpha fold was the our first and you know so far best expression of that let's say uh and i always had that on my mind and many other and other problems like that um so it would have been great i think to and and given how important agi is and how transformative a technology is maybe the most transformative one in human history um then uh i thought it would be best to approach these kinds of uh the the sort of latter stages of building it which we're in now in we're using the scientific method very carefully very precisely very thoughtfully um and rigorously with all the best scientists kind of in my ideal world collaborating on um in kind of cern like way effort uh on making sure each step we understood each step each as we got to the final goal of building agi um seems to me like that would make the most sense with the technology like this um and then and of course you don't have to wait so that might take a lot longer maybe a decade even a two decades longer but i think that would make sense given the enormity of of what we're dealing with and and then my other idea was but we don't have to wait till agi rise to get start getting the benefits of ai we could use uh more specialized systems that maybe make use of the general technologies the general algorithms we're developing for agi but are not in themselves general intelligences they're narrow ais if you want to call them like alpha fold which does a specific purpose and only that purpose and we could we could have we could and we still are i'm still doing this uh create you know many types of alpha folds and isomorphics uh while we're building agi in this careful scientific way and then benefit the humanity could benefit from the from the proceeds of that like cures for cancer uh or maybe new energy sources or new materials and so i felt that that would be you know maybe looking at this from 20 30 years ago when i started out on all of this that would have been the ideal way for it to play out um in my opinion now it didn't happen like that because technology is unpredictable and in fact it turns out that things like language were a lot easier than we were all expecting even those of us who were obviously optimists about the whole technology and eventually will crack language but it seems funny to think of it now but language and concepts and abstractions things that the current models foundation models like gemini do incredibly well we thought that maybe there will be one or two or three more breakthroughs needed before we could get there but it turned out transformers which my google colleagues invented and some reinforcement learning as well on top was enough to crack things like language and and uh we were sort of playing around with that so with the other leading labs but um of course with chat gpt and fair play to open ai they scaled it and then they put it out there and i think even they say it was a sort of science it was kind of a research experiment they didn't realize it would go so viral and i think none of us did and we had sort of fairly equivalent systems at the time because i think when you're building that technology uh you are so close to it you you're very aware of the things it can't do the flaws it has and you don't realize that actually people out there would find use even though it was hallucinating and doing other things that we're obviously all still trying to improve on now still not completely fixed um but there's still interesting use cases like summarizing things or you know brainstorming things like that that people use you know everyone uses chat bots for today now the downside of it is is that um we're in this sort of of ferocious commercial pressure race that that everyone's sort of locked into currently and then on top of that there's geopolitical issues like the us china race and so on so there's sort of multiple levels of rate of of of of pressure to sort of move fast so the benefit of that of course you get faster progress obviously so you know the progress is just like at lightning speed that these days um so that's good for all the good use cases um the second benefit is that um everybody all of the viewers out there everyone you're all getting to use the most cutting edge ai technology perhaps only three to six months behind what is actually in the labs so that's kind of mind-blowing it's also great because i think it gives everyone a feeling for it's democratizing ai it's giving everyone a feeling for what it's like uh to interact with cutting edge ai and what it can do what it can't do and i think that's good for society to start um getting normalizing itself to what is going to be an enormous change with this technology coming so it's probably better that we get to sample that in incremental steps rather than it's just a shock to the system here's a no there's no agi and then here's agi one day probably that that that's not good although i think there could be many ways it could have rolled out and then the final thing that's actually on the benefit side is that um uh you you can you can't really fully understand your systems until they're stress tested by millions of people so it doesn't matter how good your testing is and you know your in-house testing obviously millions of smart people trying out things and then you seeing what bubbles to the top or the feedback you get um is really important for building more robust systems and um better systems so i think there's positives about uh and negatives about how the way it's gone um it's not the way i dreamed about years ago where we would be sort of contemplating this philosophically and and and and sort of um carefully considering each next step um we're not in that world and i'm i mean although i'm a scientist first and foremost i'm also a pragmatic engineer so um we you know we have to deal with the world as we find it and make the best of that and we try to do that by advancing the frontier but also trying to be as responsible as we can with doing that as we deploy these you know very powerful technologies um like gemini and alpha fold [00:28:46] Speaker 2: there's another story happening at the same time as this and i want to get back to your concerns and how you weight those concerns and the cost um in order to understand that i think we need to tell a story about ai being very creative unexpectedly yes and that story begins let me find my jenga block that story begins here so let's go back to march 10 2016. yeah there's a very famous go player that sits down to play against a system that you designed and at this point computers have beat humans at all kinds of games but go is really interesting because there are more potential moves in go than atoms in the universe they're they go back and forth they're playing and then your system makes a move that is so surprising because it is incredibly unlikely that a human would figure out a move like that yeah move 37. yes and you see lisa doll sitting there he's just got this shock on his face he's got his head in his hands like this and it really was this moment where i think people like yourself saw ahead to the creativity that we would find in ai systems that are very different than the systems that we've talked about so far so there's a category where you're giving a huge amount of data and you're asking to make new predictions and i understand this is much more complicated than this over simplification but then there's a category where you're not giving data you're giving rules like with math or physics or games like go and it has this incredible opportunity for creativity yeah where were you when that moment happened and what future did you see ahead yeah it was an incredible [00:30:29] Speaker 1: moment that you're describing and it's actually almost exactly 10 years ago now which is feels like a century ago actually but i think in many ways it was the dawn of the modern ai era because until that point there were many ai programs that could beat world champions at games things like chess but they were done with what's called expert systems so they were systems where the team of smart programmers with a team of smart in that case chess grandmasters came together tried to distill the knowledge the chess grandmasters have into a set of rules and heuristics and then the programmers would build a system kind of a brute force system that would use a lot of compute like on a supercomputer like ibm did with deep blue uh to be gary kasparov and they would sort of encapsulate the rules they were given by the chess experts and then the the system would sort of dumbly execute those those rules and heuristics and do millions and millions of searching of moves and then uh try and work out against those heuristics which is the best one to do now the thing with that is um for me that was not satisfactory when i saw that in the 90s um i was doing my undergrad at the time i didn't feel like that was proper ai because that system let's take deep blue okay it's it's it's world champion level at chess but it can't do anything else not only can't it do you know language and robotics or any of those kind of things it can't even play strictly simpler game like tic-tac-toe right so something's obviously not quite right about the definition of intelligence right in the sense of like no human you could imagine a human grandmaster um not being able to learn how to play tic-tac-toe it would make no sense because it's strictly simpler so so there's something sort of wrong about um its generalization capability and the fact that it didn't learn it was just given the answer right so where if you could ask for something like deep blue where did the intelligence reside of the system well it wasn't in the system it was in the minds of the chess grandmasters and the and the programmers they solved the problem of chess and then implement and then implemented the solution the program just dumbly executed the solution now go as you mentioned is um the sort of final frontier for games it's it's the most complex game humans have ever invented it's also the oldest game so it's just amazing in many ways and it's also very beautiful so in asia where they play in china and japan and korea instead of you know it takes sort of they play instead of chess basically occupies that intellectual echelon but it's a much more intuitive game sort of artistic game almost so you you play patterns that look beautiful and they turn out to be you know really strong which is why the game has a little bit of a mystical element to it like almost encapsulate the top go players would say to you encapsulates the mysteries of the universe in the game i think that's how the ancient uh uh chinese thought about it and so um and and also just its raw complexity as as you mentioned has more possible ball positions 10 to the power 170 than there are atoms in the universe so what that means is there's no way you can brute force it in the way that we did with chess um and furthermore because the game's so intuitive and so esoteric um there aren't really these rules that you can encapsulate easily for a machine to follow so when you talk to a go master unlike a chess master they'll tell you things like why did you play there they'll say it felt right okay that boys a chess player will never say that they would say like i did it because i'm calculating this this and then they'll tell you the calculation so that intuitive intuitive feeling is obviously very hard to encapsulate in a system you can't really program that directly so it's the perfect um proving ground i would say for these new techniques that we were pioneering in the early days of deep mind of deep reinforcement learning can you use build systems that learn from themselves directly from experience so in this in the in the case of alpha go alpha go started by looking at all the games on the internet that humans have played and learning the types of moves humans would do but then we overlaid it with a monte carlo tree search that allowed it to sort of discover new branches of the tree of knowledge if you like in go starting with what humans knew and then going beyond that and that's what we hoped was going to happen so so the amazing thing about that match which was ended up being watched by 200 million people around the world was that um not only did we win the match for one that was the main objective but in game two specifically it played this famous move 37 that you talk about this creative move that was um it was on the fifth line of the board and early in the game and it's it's sort of a big no-no to do that and go right like go i go if you were being taught by a go master they would slap your wrist playing on that because it's just regarded as a bad move and and and but not only was that um a great move it ended up winning the game for alpha go like 100 moves 200 moves later it was in the right place as if it sort of presidingly put the stone there so it was the critical not only was a surprising move it was the critical move for later for it to be exactly in the right place to decide the game so obviously it's changed the way all go players play go but for me it was um the moment i'd been waiting for uh in terms of building a system we'd already spent six years by then building these types of learning systems that could achieve something no other system could you know this sort of mount everest of games ai you know the final frontier if you like of can you beat the go world champion um but also not only did it win the match but it was how it won and with these creative new ideas like move 37 and that for me was the signal that we were ready to turn it to scientific problems like alpha fold to say this back to you the [00:36:09] Speaker 2: reason why it's important that this audience that wants to understand the future understand what happened with move 37 and go yeah is because the implication is if deep mind can build a system that can do that it can also perhaps build a system that can play any game yes it can also perhaps build systems that can um figure out in real world problems what is the best solution in quantum computing or in nuclear fusion or in matrix multiplication or what else do i have chip design so many projects or um etc etc could you tell me about the cutting edge here yes pick one of these systems what is the move 37 of yes the surprising [00:36:55] Speaker 1: creative element yeah going on i think um alpha zero is very interesting to talk about which was the evolution of alpha go so after we we we we won um you know got to the pinnacle of go and showed that it could come up with new ideas at least in go move 37 and actually many other ideas that it came up with which has revolutionized how people professionals play go now uh we then generalized it further to a system called alpha zero which i think is going to turn out to be a very important system uh for today as well um where with alpha go uh we started with all the human games that we could find on the internet um and also there were a few other couple of things that were specific about go that we were built into the alpha go system like the symmetry of the board and things like that so we wanted to get rid of all of those assumptions completely and actually start from scratch um as if the the the the program and the algorithm didn't know anything about what it was trying to do to start off with and that's why this is what the zero refers to in alpha zero is sort of like alpha go but now removing any knowledge human crafted knowledge both in the data and in the uh any of the kind of heuristics that we've given the system so alpha zero starts like tabula rouser almost obviously it's a has a learning system it's a it's it's got a neural network um we we set up the parameters but we didn't give it any domain specific knowledge about go or any other game and then what we tested alpha zero on was first of all could it learn go from scratch and then be alpha go right so and we managed to do that so uh it takes 17 evolutions of the program so you can imagine what happens is alpha zero starts off random uh to begin with it just it only has the rules of the game plays randomly obviously it's terrible at playing it creates its own data set by playing a hundred thousand games against itself right and then it can see what which moves won or lost even though it's playing more or less randomly to begin with there'll be some moves that are slightly better than other moves okay so now it takes the hundred thousand it we train a new version of itself on uh version two now of alpha zero with that new data that version two is slightly better than version one so now it's not random anymore but it's not great it's not good but it's playing like okay moves and then those okay moves end up um being better and so then a version two gets trained a version three a version four and so each time that new system gets played against the old system and sees is it um significantly better or not and it turns out that are at least in go and chess and things like that around 16 17 generations of that is enough to go from random to better than world champion and at least in the case of chess which i actually once watched live happen because i was fascinated by obviously playing chess myself is you know it starts in the morning random then you know by lunch time i could still just about compete with it myself and then by tea time it's better than all grandmasters and then by dinner time it's better than the world champion and you've just seen the entire evolution of that from scratch and also it's playing interesting new chess that that even chess computers like stockfish um you know with the more the kind of expert system brute force ones uh haven't discovered those types of new types of moves so alpha zero was the full generalization of the alpha go ideas and interestingly i think we need these types of ideas back in now with um our foundation models the new you know gemini and these kinds of things which you can think of a generalized models of everything language the world around us not just a game obviously like go but we still need the this ability to search and think and reason on top of those models and sometimes we call those world models and um i think that still hasn't fully been cracked yet how to do that bring back bringing back some of these alpha go ideas but now instead of just a narrow game applying it to that but to the whole world and maybe interestingly parts of science uh too like material design um and things like chip design and uh quantum computers all of these cool projects that you know so many i've just when i see all these bricks i can't believe we're actually working on all these things but it's true is and this is sort of the dream is like i get to i love all so much of so i mean i love every branch of science and i get to um indulge myself in all these different areas of science because ai is such a general tool it can really uh make a huge difference to all these areas so maybe that one example i give is just designing new materials you know if we want a material with a special type of property can we go beyond uh what is currently known uh in material science and i think [00:41:32] Speaker 2: alpha go like processors could be very useful there and the equivalent of a move 37 would be like alpha tensor finding a new algorithm that makes you know matrix multiplication faster exactly yeah exactly so you [00:41:46] Speaker 1: can apply it in algorithmic space which is quite exciting because then the algorithm itself gets faster so there's some circular circular sort of improvement there and yes alpha tensor just making the matrix multiplication which is the the basis of all neural networks it turns out everything's matrix multiplication uh you know if you just make that five percent faster that's a huge you know the tens of billions being spent on training that's a huge cost saving and uh and so these are good examples of of of um ideas and i think we're still early you know also like things like the design of chips on a on a on a on a uh on a on a die you know making it as efficient as possible the routing you know it's a kind of uh np hard problem you know in terms of like the the traveling salesman like what's the shortest distance you can wire up all of these things and alpha chip and programs like that are really good better in some cases than human chip designers at dealing with that so i think we're just scratching the surface i would say of what's going to be possible in the next few years with um today's kind of more general systems combined with these types of ideas from from alpha go and and alpha zero i think are going to come back [00:42:54] Speaker 2: these two categories the story that starts with alpha fold the story that starts with alpha go these are the kinds of ai that make me feel really optimistic i also think that being really optimistic and you do this a lot in public which i appreciate is fully thinking through the ways in which something can go wrong and what we can do to prevent that yeah so i want to insert one other in here sure do this on purpose okay this one yes and the reason why i bring up this game is this is a real-time war game yeah and in the videos where this system is absolutely crushing humans you can see the engineers cheering for the victory of their system but of course as someone who didn't build the system i'm thinking to myself what if that's real and we're speaking right now during a time when the debate about militaries and governments using ai is a huge topic of conversation i want this conversation to last for 10 years i want it to be useful for that long so i don't want to talk about specific companies specific terms of service i also think people are in some way missing the forest for the trees here because bigger picture governments are going to use ai and so what i want to know from you as someone building these systems is if you could wave your magic wand what would you hope that they use it for well look i think [00:44:26] Speaker 1: governments and governments should be using uh um ai and you know we want to support all sort of democratically elected governments and i think um the things i would love to see them use it for and what we're trying to build our systems to be good for is uh things like improving public health uh education i mean all of these things need to be rethought the efficiency gains and the amount of good we can do with it governments could do with it for their citizens could be incredible and i think some countries are doing it like singapore and uae i think are leaning into uh these types of use cases i would love to see it being used for uh things like energy like optimizing energy grids um we did that with our data centers and save 30 of the you know energy used for the cooling systems i think there's enormous societal gain from applying ai at scale to these types of areas so that's what we you know um i've always thought about and and and and um hope that um governments will pick up and use for and we you know want to support all of that um of course you know geopolitics of the of of the world is very complicated right now and these are dual purpose technologies and um you know i worry about a couple of use cases things that can go wrong with ai that you know in the bigger picture as you say i think sometimes the the is the the the kind of details uh uh get people get bogged down in the details but actually there's big the big picture there's two things to worry about one is bad actors whether that's individuals or all the up to nation states using uh repurposing these technologies that we're trying to build for good like curing diseases and advancing material science and energy and so on um for harmful ends right and whether that's uh inadvertently or intentionally uh and then the second branch of things i worry about uh is the ai itself uh going rogue um or going off the rails if as they get more powerful that's not today's systems but maybe in the next two three four years especially as we go towards more the agentic era which we're entering now and by agents i mean systems that are capable of completing entire tasks on their own so you can of course we want those because they'll be very useful like as an assistant or something like that but also that means they'll be increasingly capable and autonomous and so um how do we make sure as as one of the frontier labs and the frontier labs all have to think about this is the guardrails are put in place that they and that we can ensure that they do exactly what they've been told to do or the goals they've been given and they've been specified clearly enough and there's no way of them circumventing that or accidentally breaching those guardrails and that's going to get that's an incredibly hard technical challenge if you think about how powerful and how smart and capable these systems eventually going to get so i tend to worry about those are that you could call the medium term now even though three four years is not really medium term but those are the things i think people are perhaps not paying enough attention to at the moment i think will be the biggest uh issues that we're going to have to contend with if we're going to get through the the agi moment in a in a way that's beneficial for for humanity yeah one of [00:47:36] Speaker 2: the biggest questions i came in for you with you know if i get an hour with you in my life was next time i read a headline how do i wait the concerns that we're all going to have over the next 30 years yes you know like what are the things that people are worrying too much about and what are the things [00:47:55] Speaker 1: that they are not worrying enough about yeah so i think the two things i just mentioned are the things that maybe the average person is not worrying enough about but even i think some of the experts and the scientists in the field i feel like those are the key things that are more societal affecting that if we if we there are other things that we need to worry about too like deep fakes and and and we work and we try to help those are immediate term worries right misinformation deep fakes these kinds of things and you know we work on this system called synth id which is uh you know a watermarking system actually an ai watermark probably somewhere one of these bricks yeah one of them and and and uh we need ai to sort through them and and and it's it's it's uses ai to actually digitally watermark any generated image so all the things all the google technologies vo and everything else and nano banana uh they all they all they all have this uh uh um uh watermarking technology so we can detect and flag to the user or or government or whoever that these are fake um and i think actually i would advocate all uh uh companies working on genitive uh ai uh should build in something uh some kind of technology like that so at least uh it can be detected or they can detect which things have been built with their uh technologies and i think that's going to be increasingly important but i think that still pales into into into you know as a small issue compared to some of these bigger issues around um agi itself um uh becoming very capable and how do we put make sure that you know guardrails are put in place that we understand uh what that those types of systems are capable of as we get towards agi and you know i think a lot more research a lot more effort needs to go into that from from everyone and actually i would love to see international cooperation and cooperation around amongst the you know leading labs around these safety issues and um and including you know with with with places like the ai uh safety institutes and and also academia um to help kind of work out how we navigate that next step because it's unprecedented to create technology like that if we play this out what's the limit here [00:50:06] Speaker 2: what are the things that you think ai cannot do that humans can do you've called this the central [00:50:13] Speaker 1: question yes your life yes it is and it's very related to um you know uh some some scientific thinking of some of my all-time heroes like alan turing you know he described turing machines which were these theoretical constructs that actually all modern computers are basically turing machines that are able to um compute anything that's computable um so anything that can be described as an algorithm uh that this this type of machine can compute and i think that the systems we're building are approximate turing machines and potentially a lot of neuroscientists including me think that maybe the brain a good model for the brain is an approximate turing machine so the question is and but there are others like um friends of mine like roger penrose uh and uh who you know believes there might be some quantum effect in the brain right and i'm sure you've probably done videos about that that and he you know we've had some very uh good natured debates about this but so far neuroscience uh hasn't found any um quantum effects in the brain um doesn't mean they won't be found but so far people have looked quite carefully and they haven't we haven't found any so it looks like most of what's going on in the brain is kind of classical computation and so therefore um it's not clear what the limit would be in terms of eventually what an ai system could do and could mimic but um you know i think that's an empirical question i think that's one of the um you know the questions around consciousness i mean i don't think it's very well defined what it is but we will intuit what it is and um i think this journey we're on of building an intelligent artifact i think we'll have almost like a controlled study comparison uh to the human mind and then i think we'll see uh in this journey like what are the differences and what's unique about the mind and i'm very open-minded about that i think there could be uh unique things and um certainly unique connections between humans that will never be replicated by uh you know these ai systems but i think a lot of things that um we currently are not in reach like long-term planning and reasoning and maybe some forms of creativity i think eventually ai systems will be able to do i want to be honest about what's happening [00:52:24] Speaker 2: in my mind right now and it is that i am doing exactly the thing that humans have done throughout history i am trying to find the reason why we are special yes it is that we have to be at the center of the universe oh wait we're not we have to be the ones that are emotionally attuned oh wait elephants have funerals oh we must be the ones that can be creative and create art oh wait gemini can do that or like what um oh we must be special do you find yourself doing that as well that's my reaction as you're [00:52:52] Speaker 1: describing this future yeah no i think i think we are special and i think there is some there's a lot of deep mysteries about how the universe works and including a lot of things that are in our minds but also things out there in physics you know i i think that's why i got i think i decided from a very young age to do ai is because i was obsessed when i was a kid uh at school with with with the big questions and normally when you you know physics was my favorite subject at school because that is the subject you're supposed to study when you're interested in all the big questions and um but the thing was i just realized uh uh i guess as a young teenager reading all these science books and biographies on the best scientists richard feinman is one of you know my all-time heroes that they actually although they just we discovered a lot and we know a lot about the world there's so much we don't know like this is incredible like we don't know what time is i mean this is this is insane to me like we you know we don't we can't even describe something as that it's just we're swimming in it but what is it we you know of course it's you know entropy and things like that but it's nothing it's nothing satisfactory about what it really is and um you know we don't understand a lot of quantum effects and gravity properly and and consciousness all mostly most of the things we we care about and and but we just sort of i feel like most people we just distract ourselves all day with you know tv shows and games and things and don't worry too much about it but i'm i've never been like that i it's just it's sort of um it's it's it's these deep mysteries kind of play on my mind all the time and i think um i'm quite open-minded about what the answers might be eventually about what's going on here the nature of reality i think that's ultimately what i'm after and i want to use ai as a tool to help us understand the nature of reality around it and i'm quite sanguine about whatever the answer might be i'm not you know i guess i'm a true scientist in that sense of like i don't actually i don't really have any pre-described notion of what the answer should [00:54:46] Speaker 2: be i just want to know the answer me too one way to describe what you're trying to do is effectively this which is to say to create a system that wouldn't be especially good at one thing or another thing but rather to create as you've been saying agi artificial general intelligence that would be good at all yes i know you're a fan of sci-fi i am too um could you play out for me the plot of the sci-fi movie in your head yeah that is the future where you actually do this yeah i can i i think um i love [00:55:24] Speaker 1: sci-fi too and i probably i read too much of it when i was when i was a kid may explain a few things but one of my favorite series was the culture series um by ian banks um i think it just paints a really interesting actually post agi world he didn't call it agi but that's what it was describing like a thousand years in the future but i think even 50 years some of this could happen where we've we've got through the agi moment safely it's built it's it's it's um helpful for society and it's it's and and you know it's it's here and maybe we will have it in our pockets even and um we've used it to crack some of these what i call root node problems in science alpha fold was one of those right so these are problems if you think of the tree of all knowledge these are kind of root node problems which if you cracked it it would unlock a whole branch of new research or new applications and i think there are other things like fusion we briefly mentioned uh or better maybe room temperature superconductors at atmospheric pressure that you could then combine with optimal batteries and things like that i think though there will be a solution to the energy problem so free pretty much free renewable clean energy one way or another fusion or you know better solar uh and then that will unlock uh us to really travel the stars because the main cost of you know elon does amazing work with spacex and those things but the main cost is still the rocket fuel right the energy cost so if that's sort of zero somehow um because we can just make infinite rocket fuel out of seawater because we have we've cracked fusions we can have you know catalyst plants and desalination everywhere um then you know that unlocks the really unlock space and then we'll um be able to get a lot more resources because we can mine asteroids all of these things that are the purview of science fiction become i think very plausible in the next 50 years dyson spheres around the sun uh mercury's sort of conveniently in the right place actually made of the right material which is kind of amazing if you think about what's going on in the universe and um and and then that should hopefully lead to you know maximum human flourishing and we help um cure all these terrible diseases so we live much longer healthier lives and traveling to the stars bringing consciousness to the rest of the galaxy that would be i think uh an amazing outcome and i think could happen within the [00:57:44] Speaker 2: next 50 years i believe you you're saying these things that i like when you're saying them i believe [00:57:51] Speaker 3: you that's the that's what we're i'm trying to do at least yeah so this is my last question [00:57:58] Speaker 2: if i were a fly on the wall at my own funeral after they said she loved her husband and her family and her friends i would hope that they would say that she spent her life trying to help people see optimistic futures so that they can be part of making them happen that they can make them happen more quickly or better for more people or whatever it is that people decide to do with the vision that they see and so my last question for you is what do you hope that they say about you i would hope that [00:58:26] Speaker 1: they would say that you know my life was of benefit and service to humanity that's i think what i'm trying to do so that's maybe will be the best thing thank you so much for your time thank you [00:58:38] Speaker 2: really appreciate it fun thanks awesome if you want to play jenga yeah exactly well you seem pretty [00:58:43] Speaker 1: good version of jenga you did that very well so yeah this is actually awesome i can't believe how many projects we've done it's really crazy when i saw the break so they all got apple yeah they have all got our projects on them did you just memorize where everything was yeah okay of course so the uh [00:58:56] Speaker 2: the game is you pull it out and we were playing this it's unfair to play with you but it would be you have to say what that project was and you don't get the point if you get it wrong so for example it would be gnome this is material science yeah it's a little bit unfair on you i mean i would hope i [00:59:15] Speaker 3: would win this game but um although you're probably way better at jenga than me yeah let's see what this one there you go okay alpha code yeah that one that one's clearer right code forces yeah for [00:59:27] Speaker 2: my sense this is uh genetics but the two percent that codes for proteins yes we have to do this now [00:59:35] Speaker 1: i've got time i can push back my next great what wait i have one more question you know alpha evolve [00:59:40] Speaker 2: alpha evolve is coding yeah we can be used for coding programming it's it's combining genetic [00:59:46] Speaker 1: algorithms with um with uh gemini so this is this is our this is one attempt at doing like alpha go stuff beyond what is known so i wouldn't get the point for that no half a point half a point okay one [00:59:59] Speaker 2: more question for you then while i have i'm just going to keep going sure i have you because why not we're still we're still rolling uh obviously um okay what did i not ask you that you think is [01:00:08] Speaker 1: important for people to know what did they not ask me i think we covered a lot actually um gencast this is yeah weather prediction yes oh yeah we didn't cover that navier stokes i completely forgot about [01:00:19] Speaker 2: solving solving that whole bunch i completely forgot about that thing i did solving so that was one [01:00:24] Speaker 1: interesting thing is simulations we didn't talk much about that or genie which is um the the role of simulations to um dqn of course started it all off the atari stuff um simulations to help you uh understand some area of science that um or even social science like economics that you can't a very hard to run either expensive to run experiments or you can't run controlled experiments in so i've always loved simulation oh yeah i said cover that there we go um we're both very competitive i think so this is gonna this is gonna be quite serious actually in jenga you is the rules if you touch it [01:01:05] Speaker 2: you have to move it do you or not we are playing a loose a loose the easier version also because we uh we were doing a creative thing where you're allowed to like push them together you can use two hands [01:01:14] Speaker 4: also oh okay you're not allowed to normally do that right right okay i'm just going to take this one i'm [01:01:19] Speaker 2: going to cheat with alpha code again one of the um questions i think people will have for you is if they're watching this and they you know are very optimistic gemini everybody yes they're very optimistic about the futures that you've described there they have all of your same concerns they generally have gotten to the end of this conversation and they're thinking i believe in this future and i want to be part of it how would you codemender i think that finds bugs in code yes very good how would you advise them to participate if this is all right people participate [01:01:56] Speaker 1: in the future i would um when i do sort of talks at universities and schools um i would say [01:02:04] Speaker 4: you've got to just go with the flow of the direction i would immerse myself in every tool available [01:02:11] Speaker 1: and just become almost like super powered fantastic um super powered with uh those tools and those um those capabilities because i think my impression is even at the frontier labs we are um there's so much work has to go into just making the next versions of these frontier models and then all the adjacent models so for us like vo and anna banana and gemini um that we even we can only explore a fraction of what that at the applied things you could do with it the applications you could you could make with it so and i think that gap's getting bigger and bigger in terms of like the overhang of the capabilities all the cool stuff and the latest models and that that time the the release schedules are getting faster and faster on that so i think the opportunity space is getting huge if for people who are really expert and um at using those tools and then apply it to some new domain so um i think a kid these days could probably start a multi-billion dollar business in some ways using these tools in some new way that no one had thought about and i think things like um open claw is a good example of that yeah yeah maybe we should call it a draw because i think uh i think i don't think i don't think either of us could [01:03:28] Speaker 3: bear to lose that right so it's your move it's your move yeah it's your move we can end on your move [01:03:33] Speaker 4: i will try my my enemy move go on then [01:03:37] Speaker 2: you're gonna make me i know it was gonna get me another move yeah in 2016 you had a sticky note on your board that said solve protein folding smiley face yeah what is it yes okay what is on the board now [01:03:50] Speaker 3: in your proverbial oh my gosh uh i've got a pile of about 100 sticky notes yeah desk so what's [01:03:57] Speaker 1: what's on it alpha chip this is what's on it what's in it i i can't actually remember it will be a list of um about 30 things that need to be done by like this evening so i better probably get to them [01:04:12] Speaker 3: but look great should we do you want it you want to actually i'm gonna keep going until you stop so [01:04:18] Speaker 1: you can stop yeah let's about to what time is it okay i'll do one more move but now we're now we're kind of cheating we're good we're using our the pieces that already [01:04:29] Speaker 2: oh you're gonna you're gonna try i'm gonna go ambitious in our last oh come on come on come on if i get this one i get another question okay that seems fair oh god [01:04:44] Speaker 3: how is that going to balance surely not no that was awesome thanks that was a great great idea to have that [01:05:04] Speaker ?: Thank you.