Neil Lawrence ao vivo: Inteligência Artificial não é assim tão simples

a [Music] hello and welcome to a brand new episode of It's not that simple from the Francis man Foundation today a special one we're in front of a live audience in Lisbon and a special one as well because we're delving into a topic that's so important for all of us humans all walks of life all Industries and that is artificial intelligence we're going to try to break down this topic and explain why it's not that simple we're doing it with a world-renowned expert who's here with us today special guest Neil Lawrence Deep Mind professor of

machine learning at the University of Cambridge you've dedicated great part of your career to this topic you've got a great book that's been out for a while called the atomic human where you talk about the role of us humans in this era of uh increasing technological presence and evolution there's so much to to break down here today Neil but I think the appropriate place to start would be to go along with the name of the show which is it's not that simple so as we sit here at the beginning of 2025 why would you say

that at this moment artificial intelligence is not that simple so I always like to look at historical examples and perhaps the one I'd choose to explain why it's not that simple is um a bit related to this apocryphal story about Henry Ford who's supposed to have said okay if I had asked my customers what they wanted out of an automobile they would have said uh faster horse and I think we can ask the question what do people think they're going to get from artificial intelligence is the answer a smarter human and I guess at that

point I would say it's not that simple you've mentioned humans straight off the bat even though I asked you a question about artif artificial intelligence and I think what will be interesting is for us to explore the relationship between humans and and and AI um I have to be honest with you whatever industry uh we talk about with the emergence of of artificial intelligence and I think it'd be pretty fair to say that most of us heard about AI through chat GPT um that there were a lot of fears a lot of fears about our

place on the planet but mostly in the workforce uh there are some who see AI as as a threat there are others who see it as a vehicle a tool of empowerment uh I had a chance to speak with with yal NOA Harari over a year ago and there are a lot of fears there from his side but you you like to look at the possibilities that are there and you don't think that the emergence of AI will threaten us as as humans and our place in society in industry in economy well it's a it's

a disruptive technology there's no question about it it's this You could argue that it's perhaps the most disruptive technology we've seen but but on the other hand it's the continuation of a technological Revolution particularly a revolution in the way that we share information through the computer and on that basis I've been talking today the first time I was in Lisbon I think was 1986 or 1987 when I was 13 years old and so if you imagine talking to I don't know a 17 or an 18y old in 1986 1987 about what jobs they should do

trying to imagine what the world looks like today this is an era before there's an internet uh an era before I guess just on the cusp of uh Portugal joining the EU EU and you think about the disruption that's occurred across those 40 years further if you go back 40 years before that to 1946 I mean this is uh at the time when the computer was first being developed so we you're in one of those challenging areas where and I it's certainly credited to Bill Gates this statement but it maybe other people have said it

as well that technology is one is often the change that you expect in the short term is often slower than you expect whereas the change you expect in the long term is often much greater so one of the challenges I think we're facing at the moment is there's a lot of heat about this topic a lot of hype but there's not so much light on the topic and and uh very few nuanced opinions being heard about what it might look like and I think that that's extremely damaging to people's confidence and disempowering to people in

their decision making um fundamentally yes there's going to be enormously disruptive change but that changes the norm has been the norm for the last 200 years when we look at the this process of the industrialization of Europe I think when you look at that change and you think well what do we want from that well there's a number of things that even with the last 40 or 50 years we see haven't taken us to a very good place and the sort of questions we should be answering is well how do we address that how do

we address social inequalities how do we address the fact that technology is not being deployed in the areas people ask us to deploy it like Health social care education better security and is mainly deployed in areas which are solving problem s people didn't know they had like no one knew 5 years ago that it was really important that we had very short video clips of other people to watch the whole time right you know that turns out to be a majorly yeah and and you know these things are great they're very entertaining I think of

them as like Greenfield Innovation but what most people care about and the struggles most people are facing are to do with you know the standard problems in uh English we sometimes call them Wicked problems the problems that don't go away and when I look at what we've done in terms of our technology to address those problems my answer is it's not good enough and so part of the steerage part of the dealing with that is accepting that people are right to be angry about the quality of the technology we've been given and people should be

questioning why isn't this technology doing more to make my Hospital operate better or make my education better and I think that there's an opportunity this is the optimistic part of me with this techn with this wave of technology to change that and to re-empower our nurses our teachers the people that make Society a better place because the last 20 years or so has seen a massive disempowering of those groups there's so much to unpack uh from what you you've just said in so many different directions to go but since we're talking about change and we're

talking about Evolution I think something which scares a lot of people it certainly scares me is the speed at which this is happening right and artificial intelligence is constantly evolving and evolving faster than us humans can adapt to it and and regulate it yeah how important is this conversation as far as the regulation is concerned because otherwise I think the internet was fast but there was a before internet after internet and I still remember the first websites that I ever accessed they were really slow to load and then before the modems got faster it was

a process now it seems that every day there's a new chapter what every day there's new functions that help you with every day it's so fast Neil how are we going to regulate this and how important is it so it's interesting thinking about the early days of the Internet isn't it because I was watching a um a YouTube video about by a guy called bbby broccoli about Nell I only watched the first half on the the plane on the way across um and the CEO of Nel who I think was called John Roth um in

in the 1990s talking about the early internet then talking about how he realized the internet was a big thing when he managed to buy a part for his 1969 e Type Jaguar from a shop in England that was a sort of run by sort of four people and it just reminded me of what we thought back in the 1990s that we just genuinely felt this was going to be a democratizing technology and it was going to bring people into contact with each other and the it was going to empower the little people because all of

a sudden that garage this you know classic Jaguar Garage in the UK was was able to sell into um Canada now of course that's happened to some extent but it's also what's also happened is it's it's provided a playing field for massive agglomeration of Power by a few organizations on a scale that is utterly unimaginable I mean like Portugal and the United Kingdom you know two countries with Colonial histories we can't always be proud of and what were those Colonial histories about often commercialization at an enormous scale but that commercialization required that you undertook extremely

risky sea voyages in fragile ships around the gate Cape of Good Hope so you were always constrained by the physical communication possibilities what we're seeing with this current wave of companies is they're utterly unconstrained in the information Arena and what we see now why is the the fear there I think is because as a result of that it doesn't matter who is in power whether you respect that individual or not you know just slightly clumsy behavior from them being deployed at the global scale overnight which is what these companies can do they can change their

mind about what to do overnight is is very disturbing um and when it when we look at you're talking about regulation of this new technology but we can see we still haven't understood how to regulate the wave of technology that emerged 20 years ago the sort of the emergence of social media um the challenges that's brought to journalism the challenges um you know that's brought to Online safety um what that means when you have advertising ecosystems where the taxes are being offshored um how it undermines High Street businesses these are present challenges today and my

argument would be that if you want to think about dystopian Tech Futures just look at the present you know this is a technology that is in some ways fundamentally undermining us doing the opposite of what John Roth said it was going to do back in the sort of mid 1990s and and and that process occurred because you know when Mark you know we grew up expecting the internet you and I were born at a similar time expecting all these wonderful things from this technology whereas someone like Mark Zuckerberg grows up thinking well no this is

a route to global domination um you know you see it in a different ways and that's one of the challenges we face is that the people that you know the equivalent of you and I um for the AI era the people who are sort of 20 23 years old today even they won't see or understand how this technology is going to be used in the future so when we think about regulation the first thing we have to realize is that the problem is more about the agility and empowerment of our existing regulators rather than the

idea that we can sit in a room centrally and decide what everyone's future looks like because to be honest you know Mark Zuckerberg deciding what it's going to look like or someone in the European commission deciding what it's going to look like neither of them should know what we really need is The Regulators who are on the ground closer to the businesses to be put in a more agile position better understand the problems that are being faced and have more capable people who understand what the technological actions are and in many cases that might not

require legislation that just requires better informed more agile support for existing regulators and it would be nice to be getting that right first before we talk about what regulation looks like for AI itself we don't talk about what does regulation look like for the computer I mean that that would be an odd thing to say we talk about it in context in deployment in finance or health or in educ a and in each case the answer is different what's particularly fascinating for me about the term machine learning is that it's learning and it's learning from

us but it's getting to a point maybe you can expand on this where it starts learning from itself and that I think is also what what scares a lot of people how you lose control of a weapon that's so strong and the regulation also in the sense from a military standpoint for example where let's say the EU passes some kind of Regulation that only allows artif icial intelligence to have influence over a certain number of topics or issues whereas maybe that's not the case in the states or in China or in Russia and um this

this evolution of of of AI and and allowing it to run free um H how how do we deal with that and how do we make sure that AI continues to learn from Human values rather than learn from itself well I think without those being present second without those being present yeah for me it's a it actually this is already embedded spiritually in some existing legislation um I know that the general data um protection regulation isn't the most popular form of regulation for many businesses but if you look at what it's trying to achieve and

I think a lot of it is about the um the sort of enactment of the legislation rather the spirit of it I think it should be called good data practice rules because what the legislation is really about is so defining the notion of a consequential decision and uh when that consequential uh decision should have human involvement so we already have it's it's not the war example but for things like Health Care um sort of getting a loan or education decisions these consequential decisions are are defined in existing legislation um I think that the real problem

there is the enforceability around that um and you know it's all become about accepting or rejecting cookies or something rather than holding companies to account it's also become a form of legislation that is easy for large companies to manage but sometimes feels like a barrier for small companies and that's also because an industry has emerged around gdpr of Consultants who sort of explain what you can and can't do so it's a very good example of a legislation that I think at its heart has a really good intent but in practice the way it's been enacted

is not achieving that intent I think it was one of the people responsible for the legislation I saw them talking about it recently it wasn't about stopping your butcher doing the business it was about dealing with the power of large companies it's not worked out like that so the first the first rule is is don't do stupid things and it's an amazing thing how hard that rule seems to be for policy makers to follow because as soon as a new technology comes the response is to panic but decisions made while panicking tend to be poor

decisions now one of the things that makes me very sad is if you look at Europe you know we're rich in human capital with the diversity of cultures and capabilities so you just think about the position Europe is in to understand the different nuances to to sort of have some understanding of say American perspectives but also some understanding of Chinese perspectives that that should imply that we can conduct a conversation that should come out with some of the most mature responses to these Technologies with our histories of our different countries we understand the Perils of

autocracy we understand the importance of democratic institutionalism but when we speak with an international voice I don't hear that diversity and that strength in diversity being reflected and I think it's really really important to address that and I think you know I don't think I've really answered your question because I think one of the problems with the question itself and the poor quality of the international debate not your I understand question it's okay it's but one of the challenges we have in answering that question is that it encourages this Panic oh what are we going

to do you know when what we really would like to do is is convene the experts the people who understand the situation listen to them have multilateral conversations between businesses public sector academics and try and learn from that what the best answer is try different things in different countries learn from different countries um the question is do we have time now I don't know but I don't see any other way forward I mean when we think in the other area you mentioned like Ai and weapons how hard have we struggled to even manage landmines right

this is a technology that's been around for decades and and it still causes enormous problems in Conflict so when people naively State well we should just have a ban on this or a ban on that not taking into account the situation on the ground in Conflict zones one of the things uh uh groups I've been working with is the um International computer and AI Network something we've been setting up to try and ensure the AI resources are deployed in the right space and we've run a few pilot projects and through those pilot projects I've got

to know colleagues from the international committee on the Red Cross you only have to spend a bit of time talking to them about what it is like to be in the middle of a conflict Zone where Warlords are contravening the Geneva Convention and your role is to go to those Warlords and explain that you know showing a prisoner's body you know after you've killed them in the town square with a signpost on it is is not good behavior you know and and you realize this is so separated from the international debate around what's going on

in conflicts and we need um this is uh you know Philipe is was my colleague I'm talking here we need people like Philipe around the table and I think the problem is our rush to legislate means we're not convening those type of quality conversations I had another question regarding the Insurgency around AI before we move on and that is again access to information and when we utilize these tools uh making sure that we're getting correct information true information right because I think one of the challenges of the social media era has definitely been facts and

we grew up looking up facts in encyclopedia compas and in books and there was some kind of vetting process and yeah we had official news networks as well that could have been biased as well but at least there was a check fact checking process before we got fed anything now you know it that's not the case if anything it's the opposite there's very little checking so as we get into an era of AI considering we haven't managed social media particularly well when it comes to information how are we going to guarantee that the information that

AI will use and uses to make decisions and uses to give to us is factual is true yeah I mean of of course there's a lot of subjectivity around this as you already hinted but I think sometimes when I think about this um I think that you know in those early years of the internet certainly I I I kind of think there was this sort of interesting age where we saw the internet of like this sort of land of information in which there was some Lakes of misinformation you had to watch out for but by

default we we sort of treated it as took it as writ um and now the modern internet I think is the Sea of misinformation with some islands of information in it and and that's been the sort of change that happened slowly and one of the things that I think is interesting with the increasing availability of AI is I think that that change is going to be happening quicker and I think it's become more obvious to people that that's true so if we go back to um what then that looks like for an information ecosystem that

we can trust well I think those islands of information start to look a little bit more like what the traditional information infrastructure looked like going sort the pre- internet where you had certain institutions that were respected perhaps universities or or or certain newspapers it was worth building up your reputation um and I think what's interesting now is the nature of those institutions changes with the changing nature of digital information technology like just as it took a long time um and I think I love Jill the Paul's book these truths on this talking about the history

of the United States but alongside that the history of the changing media it took a long time to go from the invention of the printing press to a a media where people were not just publishing opinion that they were trying to sort of reflect some better truth and then of course that was also Disturbed with the development of radio and propaganda on radio thinking about the second world war and what's happened is that we grew up in this very stable ecosystem where that was mostly settled and understood and then that's been massively disrupted there are

some interesting hopes like I think that if you look at something like Wikipedia it's an extraordinary new thing that couldn't exist without the internet that is not perfect but no institution is perfect but is it remarkably well regulated in terms of its quality of information and I think very interestingly almost impossible I would think it's impossible to design you have to get lucky with it it has to start with this burst of excitement where everyone creates articles but now it's more like there are these sort of people who monitor truth on it and they do

that they're not paid it's extraordinary you know imagine telling anyone that you're going to start that if it didn't exist they would think you're an idiot but it's one of the main sources we have for human information and it actually emerges from this constant desire to have a public andopen repository of information and one of the very disturbing things about we're seeing with large language models is of course they are being built on top of that repository they are absorbing the public goods within that repository and giving nothing back to the ecosystem when they're closed

Source um that also means that it's very difficult for you know so so there's an interesting question what does a large language model look like which has the Wikipedia likee property that people can edit and improve it and that's very difficult to imagine at the moment but let's bear in mind that Wikipedia took five or six or seven years to emerge out of the early internet so these things do to these institutions take time to emerge and we may just be in this very difficult time where we'll be subject to the worst of it a

bit like the first people to receive radio were subject to the worst of it and the first people to receive the Press what can we do about that well we can constantly look at what things are obstruct in the emergence of that ecosystem and try and loosen them up and encourage that ecosystem to come about and I should say when you look at it poor data is part of that so there's an example of something where you're curating a country's data putting it into one portal making it easier to access and this is incredibly important

because the more it's accessed and used the better the data quality becomes um and once that data quality is being used a lot and improves in quality it becomes a trusted resource so these type of activities are things we can expect and hope to see moving forward how quickly are chat Bots evolving uh we I mentioned chat gbt earlier but obviously deep sea came on the market and was incredibly disruptive if nowhere else then certainly on the stock market uh with China coming in and and and becoming a big player in the international game how

how quickly are they moving and what you think they'll be able to do you know in one two three years time with which is in Old Times like 20 30 years time because they evolve so quickly right yeah um I mean in some sense the rate of progress has been um pretty constant with increasing data but I think that there's these thresholds they cross where the the capability is just wow so the emergence of chat GPT itself I mean we' known about these language models and they' done some cool things but at some point it

crossed a threshold where I just think for any member of the public it's like wow yeah whereas previously there was this sort of wave where it was a little bit lower and it was like oh now it's talking nonsense and and and that it was very hard in the community like we were criticized I was at the time I was part of the AI Council for the UK government and we were criticized by the Tony Blair Institute of all people for for not warning um the government uh of these large language morals and it sort

of like well look even open AI didn't know the effect it was going to have at launch um Deep Mind had no investments in this space we were all aware of this technology but the most interesting thing is what people do with the technology so as a technologist most of the you can sort of conceive that it's going to get to a certain point but the hardest and most difficult to predict is what happens when people assimilate and that's why the question you're asking is very hard because what you're seeing is after that initial wave

at excitement what we've had since then is I refer to as AGI vaporware so what do I mean by AGI vaporware well we're constantly being told by big tech companies that they're developing artificial general intelligence which to me is just an absurd idiotic concept right and I wanted to describe why I think that so I don't know how you all traveled here tonight but I suspect none of you came in an artificial general vehicle um and what is an artificial General vehicle well it's the vehicle that you can use for any Journey it's sort of

like so you explain what vehicle you've got and I say no mine's better than that it's kind of like a something we do in playgrounds as a kids no my vehicle can also fly oh okay your vehicle can fly but okay my vehicle's going on the water no my boat my vehicle goes on the water oh it goes under the water too it's a submarine as well so you've got this sort of crazy um I don't know if you had it here I suspect you did hanab Barbara's Wacky Races yeah uh yeah remember that Professor

Pat pending he had one of these vehicles um it had it had everything right and of course it's an absurd notion because the right vehicle depends on the context right so if I'm going to work in Cambridge often a bicycle is a good choice but maybe if it's really raining hard people take a taxi you certainly see taxi use going up and certainly if I'm going to get back to the UK I would love to cycle I don't have quite enough time to cycle back to the UK so I'll take a flight intelligence is the

same right the right form of intelligence and and we can think of for example some intelligences are more improvisational I talk a bit about this in the book some are more planning and the right form of intelligence is dependent on the context so if you're in a more uncertain environment there's no point in doing a lot of planning if you're in a in a less uncertain environment it's often worth planning to get ahead of the game but often you don't know what the type right amount of uncertainty is so having a diversity of approaches turns

out to be the right thing so where does this notion of general intelligence comes from well it comes from Eugenics so Francis gton talked about intelligence as a rankable thing like height and he said we should breed more intelligent people and then they struggled to measure intelligence Francis gon thought intelligence should be measured by performance on the Cambridge mathematical tripos um I'm not sure that's a very good idea great though our mathematicians are you they're not good at everything so when they're talking about this AGI they are talking about a fundamentally eugenic concept that doesn't

make sense in practice but when I say AGI vaporware by talking about it in this way they they talk about the sort of scale of the thing that eliminates anyone else from participating it's like well if it costs you 100 billion pounds to create that technology that eliminates any other business from entering the market and it's a it's a tactic that is um and I'm not saying it's a conspiracy they're not sitting around thinking let's do this but it's a tactic we've seen from Silicon Valley before when producing products so the Intel itanium which was

a 64-bit processor so Intel kept saying they were going to make this extraordinary processor that pushed everyone else outside the market and then they never delivered and then all of a sudden there was no processor and it's going to be the same with AGI like this is a transformational technology don't don't uh it's it's an utterly transformational technology but it's a technology that means that everyone in this room for the first time can talk to a computer and that's extraordinary because up until now in order to use a computer to do something that you wanted

it to do you had to rely on a software engineer and waiting for that software engineer to write exactly the program you wanted they never did they just gave you Excel you know I was over hearing Excel yeah or word yeah and then you have to fill this form in in word the administrator sends you can you please fill in this word document you're thinking that's not a form that's just a word document you know that's what we get from the technology because it's all one siiz fits all but but if there's anything to this

technology at all it's a technology that allows everyone in this room to talk to and control the computer which means you're getting access to A system that has access to 300 million times more information than you can possibly imagine and that's a massive shift in the information infrastructure and I think the real challenge we face at the moment is this battle where these big tech companies rightly see this as an existential threat because in the world that we would like the power to build and control Information Technology should be redeployed amongst the wider population meaning

that a nurse can you know design their own own interface to the data system so they're not spending two hours a day doing data entry they're spending that time with patients and that the computer is supporting them in that they're doing the bit of their job that they want to do rather than the bit of their job that Microsoft or Fujitsu or Google forces them to do and right at the moment this is the most dangerous bit we're in we're in a sort of a a a piece where those existing companies are trying to preserve

their slice of the market which means persuading everyone in here that this is a technology that's beyond you that you should leave to us rather than doing what they should be doing which is producing products that Empower people to use this technology it seems that that not a month go goes by without one of the biggest established companies in the world announcing that they're investing x amount of billions of dollars or Euros or whatever it may be in in AI projects we've heard about Stargates in the states yeah uh 500 billion over the next four

years um on new infrastructure for open AI in the US is there an AI race and if so who's winning it well that's that's part of the vaporware agenda isn't it that they're basically saying um there's going to be this thing that's going to exist it's going to replace all of you and it costs 500 billion I mean you know that's what their argument is Right whereas I'm saying well no it's impossible for the machine to replace humans that's what the book is all about that that can never happen and not um in terms of

there are things that are intrinsic to who we are that are our vulnerabilities and our weaknesses and the way we overcome them that can never be fully replaced by a machine that's actually already what's enshrined in the gdpr um and that whole argument that we're seeing deflated by Deep seek so deep seek for those who haven't seen it is yes the Stargate which is Sam Alman saying that he's going to get 500 billion from soft bank and these other investors a lot of this is Sovereign wealth fund investment from countries that are in massive trade

surplus with the United States so Saudi Arabian Sovereign wealth fund or historically it's been I haven't checked all the figures in this Abu Dhabi so they're trying to find out what to do with their spare dollars they have enormous amounts of spare dollars China's puts theirs into treasury bonds and uh a lot of the Surplus uh from the sort of oil wealth is is in Ed in Tech but it's not invested cleverly in a distributed way in Tech it goes through these vehicles into these sort of like Stargate light schemes that seem self-fulfilling because the

point about this vaporware idea is it takes all the oxygen out of the room and means it feels impossible to small businesses in Portugal to innovators who are graduating from your universities to believe they can compete now what's amazing about deep seek and it's absolutely true and it will be the Deep seek is only the first deep seek we'll see other moments like this it was a relatively small you know still large by UK or Portuguese startup uh size uh startup company that was able to get to state-of-the-art Performance with only I think they claimed

$6 million I think maybe it's more but with a much smaller amount of money an achievable amount of investment um and it deflates this whole idea that you need $500 billion do to compete in this space now I believe that we will see many more examples like that but the the interesting thing is you saw this initial collapse in Tech prices in the States but then you sort of see them rebounding again because this whole vaporware idea sort of operates a bit like a pyramid scheme that that you know it's not in the Investor's interests

to say that that the the emperor has no new clothes know in that story it's why is it the kid that says the emperor doesn't have clothes because everyone else in the town is employed by the emperor they have a vested interest in keeping the emperor going and you know so if you've got the whole sort of Wall Street and all of these big tech companies are in on this sort of scheme that AGI is going to come in this way then when deep Seekers who are saying Oh look The Emperor has no new clothes

they did exactly what the people of that town would have done shut up kid bundle the kid away get it out of sight you know don't all just be illus and that's what that's sort of the the large scale effect that's going on at the moment so I think it will take a few more examples of that but I think what's really exciting and what we should really encourage is for young people Innovative people that want to do interesting things with this technology in each of our countries we can encourage them to go for it

and support them in bringing this technology to the area it's needed because the real proof of the pudding will be in the eating as it were we're we're approaching the the the close of our conversation we got a a little quick fire section at the end and to get us ready for that in a couple of sentences I'd like you to tell me what is getting you the most excited about AI right now and I'm going to ask you to tell me what are you most fearful of so in two or three sentences what what

are you most excited about right now yeah I'm I'm most well I think it's probably the same thing I'm most excited about the possibility for individuals to interact directly with computers writing their own programs making the computer do their own thing and the diversity of products we see emerging from businesses that are will enable that and that's also what you're most fearful it's also something to worry about right because it's a very very powerful technology and you know it's people will be putting it to good use and people will be putting it to nefarious use

um and what most of this is about and what I think you know the foundation here is helping with is the way that we can win through with the good use is just a little bit of coordination a little bit of support the reason why we can have confidence and and optimism even if it's only a small minority that is putting it towards the good use the nice thing is that that Minority they effect work together without having to collaborate or conspire because they're working independently but in cohesion for better outcomes if you're trying to

put things to nefarious use you're constantly creating fictional worlds um and your you know like the fictional world wec with the AGI vaporware and of course that's a really difficult example but eventually it will collapse because it's a lie and we can be confident about that and it will collapse quicker by showing more light and more support to those who are pursuing I don't want to say the truth because none of us have access to the truth as individuals but as a society by each of us individually doing the very best we can to bring

about a better world we get as close as we're likely to be to that all right quick fire time um I'm going to improvise one I'll ask three um what is the best P piece of advice you can give to your children regarding utilizing artificial intelligence in one sentence follow the things you're passionate about and work out how you can use the technology to help you in delivering them okay what is one character trait that a good leader could really benefit from having patience which there's not a lot of it around AI is there um

what would you say uh is the most important learning from your life and career I wanted to say patience again um yeah I guess it's the thing Beyond patience trying to judge when the time is to act because those two go together being patient and then acting quickly in the right moment and if you could change one thing in the world by by Magic what would that be and why my mind immediately went to Sheffield United leads United uh this evening's match uh which doesn't seem like quite quite the right answer no must be praying

on the mind um I think um yeah I think it would be that every time someone says AGI we we laugh in the way that we would laugh if the minister of Transport was claiming that we were going to solve everything with perpetual motion machines because that's what is being said and it needs to be mocked it needs to be mocked it's a ridiculous notion and the quicker we start mocking it the faster we can get on to serious conversations about what we would like Society to look like Neil it's been an absolute uh pleasure

to have an opportunity to delve into your mind and and try to extract as much information as possible about this this topic which is so relevant to to all of us our generation especially the generations to come right so continue the great work you're you're doing educating us and and what we need to know that helps us deal with our fears and and and with our uncertainties so thank you very much a big hand for Neil Lawrence thank you Pedro thank you everyone ah [Music]