Well, it's exciting al having actually Rishi Sunac do this global AI summit to try and shape the agenda on how you regulator and how you use it safely.
I think people that are really grappling for the extent to which this is positive and the extent to which the truly alarming headlines around the damage it could also do the extent to which they are something real going concerns.
And it's interesting to see the UK. Actually, I think it's the first summit of its kind, right, global summit looking at the risks of AI, and it's interesting to see the UK trying to gather everyone together to talk about it.
It is interesting, and I think it feels convincingly Rishi Sunac, doesn't it. It feels like something everything about his backstory suggests he has a kind of credible take on AI, having worked in the States where many of these tech companies are based, and also at the same time wanting to just rein in and check some of its influence for you know, everyday voters and people. So I think there is It does feel like it's convincingly Sunakian. Not
a word, maybe a word. I don't think it is, well, just any think it feels like he you know, but we we you know a lot of people are asking what actually are the aims for the summit, what actually do they want to come out of it. We've heard talk of this institute being set up. That's probably the easy bit. What is the institute do is the heart of it.
And that's really what we're looking at this week. Welcome to In the City, Bloomberg's podcast, connecting you to the conversations and the story is shaping the world of finance. I'm front Sine Lackwell with Rastratton in the London studio. Now with us this week. Nick Langson, his global editorial lead for Tech and Gaming and co host of the Bloomberg original program AI I R L. In Real Life, Nate is the pace the fastest we've ever seen it at the pace for change.
It feels like it, absolutely, and I think that is symptomatic the fact that the amount of money that now exists in order to invest in some of these technologies has never been greater. We have seen an enormous speeding
up of tech development. Absolutely because a lot of the companies that are able to invest enormously in these kinds of technologies in artificial intelligence are the ones that have made the most money from the most recent technological innovations, ie cloud computing, search, and social media.
Some of us are freaked out about AI. Others think that this is the best thing that could happen to us. But because of your time, why do you've seen these changes and really have chronicled them from the very beginning.
I remember when DeepMind was founded and everyone got very excited about this weirdly secretive company that no one knew anything about. That suddenly was all the rage in computer science and deep learning, which is what underpins the vast majority of what we currently see as artificial intelligence, although that is only a fraction of the overall body of work that is AI, which has been in development since the nineteen fifties, and we've seen it go from a
conversation about making better use of big data. You know, that was a huge topic about ten years ago. But obviously it goes without saying almost that the release of chat GPT, which bear in mind, at the time we speak right now isn't even a year old. I think that's sometimes quite shocking to people just to realize how
relatively recent that all happened. It wasn't until that release was made public that people and businesses and investors and governments and regulators started to realize, Okay, now we get it. Now we can see the potential impact returns on investment but also harms. And it really, I think motivated everybody to think, Okay, now we need to really take this serious.
And Nate haven't covered this for eighteen years when the last few weeks feels like the headlines and the warnings sort of end of the world is nigh have reached a bit of a kind of fever pitch. Do you think it's hysterical or do you think these are proportionate warnings.
I think they are well meaning but wildly doom focus, and I think history teaches us some lessons here about how to prepare for these kinds of changes. My favorite comparison to make is that of the dot com boom. You know, in the very late nineties, there was an enormous fear that the millennium bug, as it was coined, was going to cause a complete collapse of society, planes
falling out of the sky. There was at least one airline that put one of its senior managers on a plane just to prove that they weren't worried their planes were going to fall out of the sky at midnight. Banks printed additional cash because they feared people would do a run on the banks and stockpile cash. There was a legitimate fear that this was the end of all times. And then of course the millennium came to pass and
everyone realized that maybe they'd slightly overplayed the worries. But it did encourage people for those years leading up to that to at least check and prepare. And that's why I think that the questions about is this going to usher in an era of homegrown chemical warfare? Will this cause the color of society? Because an AI becomes self
interested in preserving itself and growing itself. And it is useful to ask those questions, but I think it's more useful to keep it into context around what is physically possible and what are some of the actual real world requirements for that to even happen.
I e.
An enormous amount of power and computing resource that either doesn't exist or exists but is fundamentally inoperable because it would just cost too much money.
And you can see Richie Sunak and his government trying to straddle that sort of looking both ways. Huge risks coming, but huge opportunities too. Yeah.
Absolutely, I mean it's a totally realistic concern and it's smart to think about these. It's very smart to have a public debate. And one of the biggest challenges that we have with social media now is the result of largely unchecked, unregulated growth of technology companies that eventually resulted in the consolidation of power and influence in the hands of a very small number of extraor ordinarily powerful entities,
all of which are commercially motivated and publicly owned. And therefore, you know, we can't get the genie back in the bottle for that. We want to avoid that this time.
I mean, that's my question when I'm looking ahead to the summit that they're hosting, which is in many ways, you know, is smart, and I think it probably is a decent moment for the UK to show some leadership in an area where there is great investment in the UK.
But the question I keep returning to is, as you say, when you look at the evidence of getting big social media companies to do the right thing and to police themselves and so on, have we got the kind of track record from them that suggests that they will also police themselves in managing these huge language sets.
Absolutely not no, there is absolutely no, they're absolutely not. But there are good intentions. But some of the good intentions also come with the caveat that they are self interested in protecting their business. They want to release the technology because they want to commercialize it and monetize it and be seen, as you know, as a leader in that field. But at the same time, you don't want
to be caught out. And it's why you see a lot of companies calling for regulation, because it gives them boundaries and guardrails and it allows them to say, well, look, we've played by the rules you set, so it's not our fault that this happened. And that is a challenge in its own right.
But it'd be interesting to see what you think actually the overarching goal for this AI summit is, and when we understand the goal, whether other countries like chrying to share those goals.
Well, so yes to different different questions, but you're absolutely right the key questions what's the goal for the UK? Well, there is investment in the UK by these firms. If we're asking the question that we are often asking on this pod and in the readout and all over Bloomberg. Whereas the growth agenda come from, Partly it will be around investment in tech firms into the UK. So the first thing, there's that. Secondly it's something which she soon
that cares about. And everything about his backstory points to a summit on AI. You know, if his ties right, you know, he's a you know, Standford Business School. He has spend a lot of his career in America, so he is instinctively comfortable with those companies. But equally he is the MP for constituency in Yorkshire, which you know, at the end of the day, normal people will be affected by this revolution that is coming. I think it is interesting the question around what effect it will have
on jobs Eventually. We've had warnings this week on job losses by twenty thirty and are people ready? And you have a lot of very sensible people saying in that context where many many low paid jobs are being automated, we have to think about a debate around universal basic income, which a few years ago was unthinkable. I think we all know and everyone listening though, is that there are and Nate's mentioned it already there's these constant fars that new tech's going to get rid of jobs and they
don't necessarily materialize. I feel a bit more gloomy about this one because you have seen increasingly over the last few years that lots of low paid jobs are being automated and they aren't being replaced by particularly high esteemed valuable jobs that give people, you know, a sense of
purpose in the morning when they go to work. I think there is a question mark there, But it is the case too that history does show us that actually it's the kind of the luod Oitte fallacy that essentially, in the end, different jobs are made and there's more of them. Well, I hope.
So is there illegitimacy for the UK to be hosting this? So by regulating it, do you also attract growth and investment because you're seen as the AI destination?
I think we need to work very hard to be seen as an AI destination. I think that in the sense, the UK has an opportunity here to play something of a Switzerland role in AI because let's be honest, the money is in China and the US we do not have a huge AI tech scene in that you know, we are investing a tiny amount of money in AI. We are making a very small amount of money available
to AI development. Even the billion pounds that figure that gets thrown around that you know, SUNA can smaller than I mean, I think it was Amazon invested something like four billion dollars in one AI company. That goes four times the figure from one company in one startup, like
it pales into its significance. But I do think on my Switzerland comparison is that there is an argument I think, and it's not one the government I'm sure we'd agree with, but I think it's realistic that if you have a country like the UK that will be impacted or benefited by negative or positive outcomes of AI but doesn't actually have much of a commercial stake in its development, it actually puts it in a rather arguably neutral role in setting at least some of the agendas that other countries
play by, because we can be a little bit more impartial as long.
As whatever the UK's goals are shared with the people that matter, i e. China.
Yes, that's very true, and there is some argument whether we do or not, And certainly the US and China have a very different approach to how they foster and encourage the growth and development of a lot of technologies. China is very much top down, the party will dictate a lot of the time where the money should go and where the support for development should come from, whereas the US is much more like, we'll get on with it, private enterprise and we'll regulate if we have to.
And the pitch from the government will be, Okay, let's use these brexit freedoms. You've seen already tech bosses saying use your Brexit freedoms, right, So it's not just the necessarily a sort of comm's line from this government, it is it's also being echoed by people in the industry.
So that's so that's that's that's the pitch. But completely you know, Nate is obviously very lucid on you know, in reality, does the UK have the heft to convene this debate and necessarily pull out of it's that kind of not quite as regulatory as the EU, not quite as wild west as America, and definitely not China or whether actually in all of that, you know, then you know they get left behind.
But are they still aligned with the EU? So again it's I.
Think there is that here that here that they can actually push out further and that and they're being encouraged to by various players in this scene.
So they're forming an institute to try and understand. First, you have to understand it, I guess, right, So that's where we die. Where's that incident going to be? But you have to understand what you're regulating before you regulate. So the problem with regulators is that they're always a step behind. Because AI evolves, the chips evolve, it gets smarter, and the regulators need to catch up.
That's always been the problem with technology as well. And that's what's so interesting is that none of this is unique to AI. This is something that's happened every year, every decade, seemingly that I can remember reading about or talking about. The hope is that we collectively learn from past experience that when we spot something that maybe we should probably put a lot of extra time into looking
at early, we do that. I like to think that that's kind of what the world is doing with this, and that maybe that's a good thing.
So Nate, isn't the problem with deep fakes that actually you can't really tell them apart. So there's a picture of a leopard going around. There's three of them, and it challenges you to find the deep fake one.
The difficult it's very, very difficult. There are many experts trained in this that cannot tell a deep fake from the real thing, and so there's this debate around we should have everybody in the industry that is producing something that would qualify as a deep fake or something that is generated to water market. And that's not watermarking in the sense of if you look at a photo online and it's got like the brand slightly proof, yeah, proof across the top, or it's some audio water marking that
makes the video useless to share or sell. It's not that sort of thing. It's baked into the image itself. In a very similar way to right now, if you take your iPhone or your Android phone, you take a photograph, the amount of data that is stored in that image that you never see unless you look for it, it is staggering. Everything is in there where it was taken, the GPS coordinates, the timestamp, the model of camera, the model of lens in the camera, the aperture of the
lens attached to the camera. You know whether the flash fired, whether it didn't the model make everything. This would just be another piece exactly. It's all stored in. There is removable, you can delete it, but the vast majority are never going to be removed because no one knows they're there, and why would you.
For the most that feels in a very difficult debate. That feels promising, But the problem is does everybody agree to the watermark.
Everyone would have to agree to it, and everyone would have to sign up to support it in a way, and it's a slightly strange comparison perhaps, but in the way that if you buy an audio like an AV receiver for a home cinema or a pair of headphones or something that you know they all support Dolby Digital, you know that is a format, that is a standard, that is a thing that is internationally recognized and adopted, and so you would need a standard's body to say
this is the standard for watermarking whether an image or even part of an image is a deep fake, and that every other piece of software or hardware around the world can read and interpret that, and that at least gets you some of the way towards identifying whether is real or fake. But it does change because you can manipulate images.
And how quickly can that happen, because clearly we've got two major elections next year years so not for next year.
Very quickly it can be developed, but adoption and an acceptance internationally and roll out is is a many years long process, I would say.
So we'll see deep fakes in the next elections, right.
I spoke to an expert the other day and said, will will deep fakes decide the twenty twenty four election, and he said, no, it'll be one or lost on abortion rights, but it will be deep fakes that are the core driver underneath that or words to that effect.
It is by the time you've established that something's a deep fake, some of the damage will have been done twelve as later, and a lot of people won't even listen to the clarification, as we've seen recently.
Yeah, and the argument comes down to, well, is being used to spread disinformation and sometimes things are created with the goal of subverting reality as a disinformation item. But if that looks real and somebody unwittingly shares it, they are sharing it effectively is misinformation because they don't realize that it's not real. And with the speed of social media, it can travel way quicker than you could ever possibly counteract.
It, and so I mean, we're probably guilty in this conversation of focusing too much on the downside, and that's for justifiable reasons, but the upsides are massive. My favorite is, you know, for a teacher who has to mark thirty books or twenty five books every night, that can be aied. Ditto for difficult healthcare. AI will spot stuff that it was impossible for GPS to necessarily crawl over all that data.
So there's legitimate hopes that sensible, proportionate use of AI can help with productivity and release all of us to do what The question is, how do you adjudicate well?
I think one thing that's really important to keep in mind in all of these conversations is that there is this sort of common belief that you can take a problem and you can throw AI at it and that problem gets solved. And that's just not true. And it's really important that people understand that you can't just throw something into AI and suddenly get something better. It's a
significantly more complex series of steps to do that. One of the things I find most fascinating about what could happen now is if you look at the large language models, which is what if you take chat GPT, that is, you know, the poster child for a lot of this, even though the technology underneath it isn't that new. It uses large language models, it scrapes the entire Internet, it
reads every word on the Internet. Increasingly, now those sorts of products are able to look at every video and every picture and that's what's called a multimodal AI and it can process that, train itself on that and then give you usable outcomes. But there comes a point I think, and I actually think we're reaching that point now where everything that exists has been read and scanned. So in order to improve to the same degree, you need new types of data, and you need quality data.
Reliable data, I mean reliable problem exactly, deep fakes right, it has to be vetted data.
It does. And so the next step is, well, where does the next huge volume of data come from. It will come from sensors, it will come from us, it will come from medical trials and studies. But my point is is that if you develop some kind of a system that can try with molecules, looking at the interactions of molecules in a way that you cannot realistically do.
In a lab, you can run all these simulations to figure out how a molecule might interact, take the data that that simulation has created, and then apply AI to well, what could we do with this that changes the game entirely. That is not where we're at really right now, but that is probably one of the next steps that we're going to get to, and that's where things get very exciting.
An optimistic, I'm optimistic, but I mean I'm an optimist generally speaking, and I think that we have a when we talk about doomerism and AI doomerism, I think we take a really anthropocentric view of these systems in that we build things to kind of mimic our own inherent
abilities with the human brain. Neural networks are very loosely modeled on how human neurons work and interact with each other, and so when they do something, as in, when an AI system does something that looks slightly human like, it looks a bit like creativity, It's kind of like seeing a face in a piece of toast, you sort of think, well, is that the you know, is that the reincarnation of
the Messiah. It's not it's just very burn bread. But as human beings, we are evolutionary trained to see that, and I think we spot those in AI as well, and then we say, well, if it's going to do a little bit of human like creativity on this, humans do sadly have a tendency to start wars and kill each other, and maybe we should stop it doing that. But it's not necessarily coercentric.
That's a big word name. I think I have to google it it.
I mean, yeah, we obsession about human beings, and.
It's the obsession that humans are the end game, you know. It's this thing that we are the most evolved and the holy Grail, the final step in the evolutionary journey, and that anything we create is ultimately aiming to get to that point too. And I just think that's not necessarily true.
Does in the UK at the summit need a big win and saying we're regulating this way so that people are safe.
That would be great, but I mean, the technology doesn't respect borders, it doesn't respect geopolitical tensions. It just it is what it is, and the technology is global, so I think it'd be very hard for anyone to come out of that and say yes, we've set the agenda for global regulation because it doesn't work, and we saw that it doesn't work with social media. Look at Facebook.
It cannot come up with a content moderation policy that it can create in California, plot and apply worldwide when cultural sensitivities and laws are so vastly different outside of its own country, so that the idea is best principles and being able to come up with something that says, you know, if you run through this checklist of ten things that will at least get you closer to being able to answer a question that your own constituents and your own jurisdiction may be asking you, which is how
should we move forward? That would be quite useful. I think.
Thanks for listening to this week's in the City. We'll be back next week, but in the meantime, if you like our show, please head on over to wherever you listen to podcasts, rate, review, and subscribe. It doesn't really help people find the show. This episode was hosted by me Francine Laqua with Allegro Stratton. It was produced by Summersati, additional editing by Blake Maples and really special thanks to Nate Langxing