


In this episode, Danny Fortson interviews Alex Blania, CEO and co-founder of Tools for Humanity, discussing their groundbreaking project, Worldcoin. The initiative aims to verify human identities online to combat the rise of sophisticated AI imitators. The episode also covers topics such as Tesla's robotaxi rollout and the societal implications of humans forming emotional connections with AI.
Written by AI. May contain mistakes.

Get the gist, explore an idea or ask about something said.
On Times Tech, what are the key points of this episode?
Automatic transcript. May contain errors.
0:02Hello and welcome to the Times Tech Podcast with me Danny Fortson and me Katie Prescott this is Katie in the City meets Danny in the Valley, connecting Silicon Valley with London and the world, no less. Now, hang on, Katie, before we get into the nitty gritty today, there's something I need to check with you. And it's really important. I'm going to show you these pictures. And I want you to tell me which ones contain traffic lights. What? I just need to make sure you're actually Katie, real Katie in the City, not an AI, not an avatar, not somebody who's trying to like scam me, But I am me, but I just always get these things wrong.
0:43I always get them wrong too. But this week I ask because, I joke, I joke. I ask because this week we spoke to or I spoke to Alex Blania. He's the CEO and one of the co-founders of Tools for Humanity. They're the creators of WorldCoin, which is our idea to verify humans online and help us weed out AI imitators. We'll hear that interview in just a moment. So top left, bottom right, maybe that one. You're out. You're out. And a little later, we'll also talk about Tesla's robo-taxi rollout, humans falling in love with AIs, and Amazon's£40 billion investment in the UK. But first, Danny, tell me everything about WorldCoin.
1:27Yeah, so it's a fascinating company. It is, from the way I think about it, it's one of the three companies that Sam Altman has created. Friend of the pod, Sam Altman, founder of OpenAI, obviously. One of his three companies as part of his plan for global domination. So he's got open AI, make this thing that's going to replace all humans at all things. How do you power that? I'm going to put$400 million of my own money into nuclear fusion. So he's got a startup doing that. Then his third startup is in a world where it's going to be very trivial to create very high fidelity AIs to fake people out.
2:06like the internet is just going to be awash in ai imitators fakes etc how do you solve for that and their idea was originally called world coin the company's tools for humanity and they've actually since changed the name to world sorry lots of different names going there but the basic idea is they've created an orb right and what company doesn't need an orb how big is the orb The orb is about the size of a bowling ball. Okay. Basically, it almost looks like a magic eight ball. It has like a camera that you look into. It scans your iris. And apparently your iris is more unique than your fingerprints.
2:46Wow. And then in an encrypted way, reads your iris. Okay, that's a real person. We're going to give them a unique world ID. And then that becomes like effectively your passport for anything you do digitally, online, whatever it may be. And the kicker is it also sets up a world crypto account and gives you world coin as a way to incentivize people to sign up. What's not to like? Exactly. And so I had Alex on the pod two years ago. They just pushed their orbs out into the world. They had set them up in malls in Nairobi and Tokyo and all over the world. and you know in Kenya thousands of people queued up to sign up and get their hands on this crypto free crypto who doesn't want that at the time it was about 50 bucks a person and then Kenya's office of the data protection commissioner not so pleased launched a lawsuit tried to stop the world operating there Alex who is very young ends up having to you know testify before parliament all of this like just a wild story well you can see why the kenyan government might not have been particularly happy about everybody just giving all of their personal information away in a shopping mall to a man with an orb for 50 dollars of made-up money well when you put it like that it's not very charitable katie this is the future but i can't wait to hear what alex has to say about it but we had them back we had them back on because they had this very big new launch where they critically when they launched they could not launch their cryptocurrency aspect of it in america because biden not a big fan of crypto but a couple months ago they launched so now they're you know issuing cryptocurrency in america putting their orbs out there they launched like a flagship location in downtown san francisco where everybody can get orbified have you done it I have.
4:45I do have a world ID. Wow. And anyway, I caught up with Alex to hear all about it. So here he is, Alex Blania, CEO of Tools for Humanity, to kind of, you know, answer the many questions we have. The idea was basically twofold. I was like, okay, we will need internet-scale proof of human, so a way that you can actually verify to websites and all kinds of services that you are, in fact, a real human being versus an AI system, because we believe that eventually AI systems will be smart enough to fool us online to make us believe that in fact they might be people. I think that's already happening.
5:21That is already happening. So I think that played out really well. So we will basically coexist with these AI systems online. And if we don't know what is an AI or what is a human, that's going to be a pretty crazy world to live in. So that was belief number one. Belief number two was that fundamentally it's a network because like if only you and me have a proven human it's pretty useless and so you will need a way to bootstrap such a network and originally paypal for example launched by giving you these referrals when you signed up you received twenty dollars or you invited a friend and the friend received twenty dollars in paypal as an account to kind of really start using the platform and our simple idea back then was like okay crypto networks actually are all about incentive alignment and there is a world in which we can maybe launch such a new network by giving ownership in it to all its users and use that as a boostering mechanism.
6:19So simply speaking, you sign up to the platform. Not only do you get your proof of human, but also you get actual kind of ownership in a network through a token, a digital currency that will be worth a certain amount of dollars. Yeah. And that will get you over the hump to actually join this and kind of do the work to actually verify, et cetera. How many people have had their iris scanned and now have a unique world ID? Around 12 million now. Okay. I think last time it was less than a million, the last time I spoke with you. Certainly less than 5 million, but yes, it grew a lot. And now in our app, it's actually almost 28 million people.
6:56So it's now a pretty decently sized platform. Yeah, since the last time we spoke, everything changed basically. The big thing that changed is fundamentally that now a lot of people are waking up to the need for Proof of Human. I think that's the first big piece. There was a big, big vibe shift around all of that. Now many CEOs reach out to me and are like, how can we do this? How can we integrate? How can we help you to get to scale? And so, for example, we announced the partnership with Match Group, which is Tinder, Hinge, all the big dating apps that will integrate with World ID later in the year, starting in Japan and then expanding from there.
7:30we recently had on one of the women who was scammed i don't know if you saw the film the tinder swindler i did not see that movie now a guy who swindled a bunch of women he met on tinder out of lots and lots of money and she has since gone off and being like i'm gonna create a tool that keeps us from happening again and again it's just around a basic id some way to id that a person is a person so it's interesting that you say that you guys have done this deal with match right yeah then we we did a similar deal with razor which is one of the largest gaming hardware companies and essentially bringing world eddie into the kind of turning it into the default for gaming because for gaming it also really matters that you know you actually play with other people right and then lastly the next big push will be social so like one of the large social networks and i'm pretty confident we will have exciting news to share there too pretty soon uh but now zooming out again what happened the last 18 months since we spoke that was a big change that like A lot of the platforms now feel very threatened by AI and very, very willing to integrate and ask us to kind of scale up faster.
8:37Second, we launched something called mini apps, meaning given we have 26 million users in our app, we make it easy enough for developers to actually access these users and kind of use World ID and also the financial primitives to build interesting apps with that. And then lastly, it's all about expansion. So it's all about rolling out in as many markets as we can, as fast as we can. Of course, the biggest one now is the US, which we announced last week. But then across Asia, Japan, South Korea, Philippines, Malaysia, Thailand, all of those happened in the last couple of months. So it is, yeah, it is very, very busy.
9:15So when you say launch, and just so people understand, the hardware you have built is this kind of bowling ball size orb you can look into it and it has a scanner and it scans your iris and the iris is correct me if i'm getting any of this wrong more unique than your fingerprint right and then it scans that and then assigns that iris print an id and then that kind of number becomes you and it's linked to you is that correct well it misses a couple important pieces to it okay so if you say proof of human what does that actually mean what is the actual problem you're solving for the really hard part is actually uniqueness so the really hard part is that like each of us can prove to a platform that like i'm actually a unique user to your platform i'm just one human because otherwise i can just like create arbitrary many accounts and i can give them to ai's and then ai's can run these accounts and then you run in the same problem so uniqueness is the really hard part but it's actually the case for every identity system the same is the case for like getting a passport like the hard part is making sure that you danny only get one passport and at like five in the US and you exploit the social security system which is actually happening at a decently large scale so it's not an easy problem and so uniqueness odd part versus what your iPhone for example is doing with face IDE is just it's re-authenticating you so it's just making sure you're again Danny logging in right so that's a one-to-one comparison versus what we need to solve is we need to make sure that you are one amongst many so it goes from one to one to one to one to many and that's a much harder task to solve so like face fingerprint all those things don't actually work and the iris which is just the muscle in your eye which also for example apple id is using with the envision pro and i think most of the vr headsets will start using they will all start using iris and so we do that too however so that's property number one property number two is like you actually want to build such a system to be anonymous and fundamentally privacy preserving.
11:13Otherwise, it would be a kind of a terrible tool for the internet. And we think this has to exist on internet scale. So meaning like billions of people, billions of people, everyone will need some form of this. Otherwise, the internet will just not work the way we want it to. So it needs to be fundamentally anonymous and provably so. And now what happens is you verify with an org, as you mentioned, it takes a picture of your eyes. And first it checks that you are in fact a person, not a display. So it has a couple other sensors in front, the neural networks on the device that check what we see is an actual person.
11:45Then it takes a picture of the iris. And then there's a neural network on the device that actually calculates a code out of this, as you described. Then it deletes the actual picture. Then it splits the iris code in multiple pieces and sends it to a multi-party computation. So currently there's the University of Berkeley, the University of Erlangen, and Nethermind, which is a validator company in Europe, sends basically each part to the University of Berkeley, one to FAU, one to NetherMind. And then those three parties actually need to come together to ensure that that user is unique amongst the global set of users.
12:23So that means not a single party actually has any of the data. And these parties that are currently three, it's going to be tens in a year or so. So there's no central storage. And then second, once that comparison is successful, your world ID, which is actually a separate key pair so it's not your iris code but you have a separate key pair that you can then verify with what is called a zero knowledge proof which is another form of cryptography that lets you prove something to other platforms without revealing anything about yourself and so you end up a verified world id which is a separate key pair so it's not your actual iris code or your number or anything it's just a separate verified key pair that you then use with cryptography their knowledge proofs on platforms to prove to the platform all that we want you to know or the platform should know about you which is that you are in fact a unique human being to the platform and so x or meta or whatever they can use that to then have you log in and know that you are in fact a unique user a unique human to their platform without knowing anything about you so so just so i understand just the mechanics of it what it sounds like you're describing is almost like world id becomes your key to the internet yes because last time we spoke and i'd love to hear when you say to launching in other countries what does that mean because i think last time it was like you were like popping up in tokyo or nairobi or wherever with a bag full of orbs and putting them out in public places and being like hey scan your iris you get some free crypto you get a proof of humanness all of these things and when i wrote about you guys 18 months ago i also got scanned i have a world id great welcome thank you human thank you but then how does that work as a key to the internet if i'm back home on my computer and i want to log into facebook for example so a couple things one is when you verify you get to download something on your phone from that verification which is a signed face image so it is not stored anywhere else but on your phone so there's a connection between the orb and your phone you get that face image on your phone and then you actually authenticate your world id with your face similar to face id so the the hard part is the uniqueness part the authentication part afterwards is relatively easy actually at least for new phones for older phones you will have to go back to an orb once in a while to re-authenticate if it's important services like for example meta or something could trigger please re-authenticate now but now to the other question of like what does it mean to launch a country well it it will mean a couple different things but what it means for now is usually we hire a team in the region and we then find locations in cities which are more like stores and we position orbs there or we find operation partners which is like companies that already have stores for whatever reason and they place that orb in their store so it could be an electronic store or like a coffee shop or something they could put it there and then it shows up in the app and people can go and then verify.
15:28And then once you're a user, you download the world app and you see a map of all the locations for the orbs to actually verify. But yeah, the broad dynamics, as you summarized it, is show up with a bunch of orbs and distribute them in the country. Isn't this broadly correct, you know?
15:46So let's get back to the States because 18 months ago, I talked to you, I had my iris scanned, I got world ID. no crypto because i was in america that's right and this was the sec under gary gensler not a big fan of crypto all of that stuff right here we are in the trump administration trump has his own meme coin as do his family members then there could be a separate broadcast yeah totally but we're you know it's the golden age of crypto blah blah blah blah blah i'm assuming just that vibe shift and the more um let's call it constructive crypto regulation that is expected is what allowed you guys to do this big us launch uh that's a big part of it yeah given sam is a co-founder we always try to be like very very conservative on everything in the direction because i think we just have a massive target in our back we have a lot of attention everything we do ends up in news and whatever and so like we just have been way more cautious than basically every other major crypto project because all of them were already live in the us and just dealt with the sec in our case we're just like okay we will not do anything until we have a clear understanding that this is actually in line with the understanding of the regulator but with the new sec i'm pretty optimistic there will be like clear regulation that everyone can follow and that's that's really all that we've been asking for is the ultimate killer app uh universal basic income if you're going back to where we started with your kind of two or three beliefs yeah i think um i think the ultimate killer app will still be we call it the real human network but i think the largest network of real humans on the internet is going to be an incredibly valuable thing and a very very important thing for internet as a whole how we run our social networks how we run our democracies i think it's going to be a very very important thing that i think is still very easy to underestimate but i think in the next year or two years with agentic ai and all of this i think it's going to be quite quite obvious and uh well sam was a fan of ubi in the early days and i i can understand that with like his beliefs around where ai is going fundamentally i'm like i'm a big believer in capitalism generally speaking i think that's the only that's the only thing that really worked for us as a society and i think deviating from that will need a very good reason it could be that agi is that good reason but i think that is a that is both a future question and not a question that i think we should answer ourselves but rather something that i think society should answer and then we will have the infrastructure for all of those things to happen that was like a very long and convoluted answer but i think and it basically basically a non-answer but um effectively it's fair to say that you guys are building the rails for what could be like a UBI distribution network of just like, you know, if 15, 20 years down the line, when it's like, oh my goodness, AG's agents are doing the majority of economically valuable work.
18:49We have to figure out a different way to organize society. Oh, here's every American has a unique ID that is linked also to a crypto account. Every two weeks you get a cha-ching and you're topped up. Plus, I think the extra time is not like 10 to 20 years, but yes. what's the actual timeline freak me out I think like in 10 years we're like definitely in super intelligence territory pretty high confidence at the world like if we don't hit kind of super big major roadblocks we're definitely in super intelligence territory and we might even be in 5 years or less so yeah I think the world is going to get very weird pretty fast but I think it's going to be amazing I think it's going to be I really do think it's going to be a bright future but it's definitely going to be quite a quite a record change.
19:37The other thing I want to ask about, and I mentioned it at the top, was you guys seem to always be kind of bumping up against some kind of authority, some government in some part of the world, and then being like, nope, get out of your country, or you need to assure us that what you're doing is not nefarious in some way. So could you talk about that? Because I think, as I mentioned, you were in Nairobi, had just been in Nairobi last time we spoke. We just, I think had this court case in Kenya, who again was bad for you guys. Now Indonesia is pushing back. So how do you view all that? I think, first of all, we actually expected most of this as we started because on the face of it, this is a very crazy sounding thing.
20:22And so if you are a regulator, and especially if you're a privacy regulator, and suddenly you have something called orbs showing up in your country, and like verifying humans and like verifying your humanity on the internet, which is something that sounds like a very foreign concept to pretty much everyone at this point still. So like it's very normal that I think there's a lot of questions that we have to answer and that's totally fine. So like I think the last time we spoke and that did not change, I'm pretty relaxed about this because I think that's just a, we have a huge, huge problem if this doesn't succeed because I don't think there's anything else right now that does something similar.
21:01And I think the world will wake up to this more and more. And I have very, very fundamental conviction in how the system is built and architecture that it actually has all the properties that we care about, which is fundamental privacy, anonymity. And I think it's like a fundamentally freedom enhancing technology, which is something I always cared about. And yeah, and so we will just work with regulators to answer all their questions. And that's going to be bumpy at times. and for what it's worth I think the financial piece is going to also be bumpy as the kind of as the market cap of this whole thing is going to increase more and more value is going to flow through it so there's like there's like a lot of things that I think will will be bumpy at times as this is getting bigger but it is getting bigger and it's it's getting more and more adoption and so yeah we will work with regulators to to figure that out and then to the specifics I'm actually really happy about the Kenya ruling because we've been waiting for it for a while and that's referencing to 2023.
22:02So I think that's great progress that we finally now have something to work with. And I still think we will figure it out hopefully relatively soon so we can relaunch in Kenya. And Indonesia, we just have Damien on the ground working with the regulators there. Why are you testifying next? I have nothing scheduled, nor do I plan to do these things much anymore. But I think probably the next really big one might be the US if this is becoming very successful. Yeah. You ready to get grilled on Capitol Hill? You've got to be old handed at this now. Oh, totally. Going for it. Yeah, yeah. And then the last one before you go is just going back to what you mentioned at the beginning.
22:41I think that partnership with Match, which as you say, owns Tinder and a bunch of the big dating apps. What's the role out there? What does that look like? How do you see that working? Do you see this becoming almost like a badge you're going to need if you want to be successful on a dating app, like a fair trade or something like that? Where you're like, I'm real. I think so. So first of all, it's going to work. It's going to start in Japan. And we will support two use cases. So one is attest your age with your actual government ID through world ID. So meaning like you're above 18. And the second one is attest your humanity with an orb signal.
23:15And then you basically get like a human batch in the app. So in your profile in the little bottom right corner of something, it'll say verified human. Correct. Yeah, it has like a little human batch. Yep. What strange times we live in. man we're not even i think we're like one percent one percent of the strangeness that i think is gonna happen in the next five years but yes so you have a little human badge in there and then also because we have a signed face image from the verification not only do we know you're any kind of human but you're also that kind of human that actually has these pictures so you cannot pretend to be someone else but that's that is really you right and that's a huge problem on these on these platforms because like these ai generated imagers are getting better and better and now they have bots chatting with you and so like for sure no one really knows what the actual number is but it's pretty large of like ai profiles on these platforms so last week there was this which was like perfect marketing for us in some sense but like uh the university of zurich i promise i did not pay for it but the university of zurich released a paper how they were able to change the opinion of entire subreddits by essentially having AIs respond.
24:27And so they're essentially able to change public opinion by having AI in the conversation. Propaganda bots. Propaganda bots. And I think like, yeah, thinking about it now since the week, I think we're going to have this like very visceral moment where at some point these AIs are going to be super human and kind of understanding you and like getting you to do certain things or believe certain things. And I think it's going to be very scary to interact with something on the internet that is not actually, you don't actually know what you're dealing with. So what do you think? Amazing. I mean, I was getting more and more convinced the more I listened to him.
25:04I mean, you can see there's something there. And it was interesting you brought up the Tinder Swindler interview as well. People are looking at this issue. I was with someone at Harrods the other day. And as you probably know, there have been a number of cyber attacks on retailers here in the UK. M &S is the big one, right? M &S is the big one, but others have been hit too. And this guy who was very senior was saying they've all started using fingerprints as a way of just a far more secure way of checking that people are who they say they are and that they can access the systems rather than using passwords.
25:36And it's just a similar thing, you know, going to the iris or something physical in this world where everything is digital and we're talking about technology replacing human intelligence. It's really interesting. Yeah. And I think also just the point he ended on around propaganda bots, I think that's also something that people aren't really alive to yet. You know, the idea that if you talk to these, any of these startups who are working on agentic AI, which is the thing that all the big developers are working on Google and open AI and everybody else, you know, they're like, within a couple of years, there's going to be more AIs on the internet than humans.
26:12That is what they're working toward. and you're like cool if i can spin up an ai to go like you know shop for flights for me while i do something else but also if you have millions or billions of ais invading reddit or the comment section or whatever and they're directed by somebody or something with an end in mind that is quite scary and so it places a lot of value then on on knowing what's real yeah and i don't know if this is the answer but also i just think and i think it's quite clever marketing the idea of like we're gonna go for dating apps of just like you know you have your little i'm a real human badge on my profile like don't swipe right unless you know this person's real i'm a real human and you know who i am i was sitting with someone at our ceo summit last week somebody important it was somebody important so i can't say their name but in fact they were so important i can't say their name um but they did say they had read a quote in the trade press that they'd supposedly said and they just hadn't and it became clear that this magazine had just sucked up a press release generated an article and generated a very plausible quote but this guy just hadn't said it right i mean it's not as not as scary as what you're talking about with propaganda bots but it is it shows the the blurring of the real yeah and the digital and i do think the other thing about all this is just like the whole universal basic income angle.
27:41Oh, that's really interesting. Because I don't know if you know the Sam Altman, he like through Y Combinator, which he used to run, funded a study into universal basic income that ran for a few years here in the Bay Area, where they, you know, had a control group. And then they gave people wasn't a huge amount of money, 1000 bucks 1500 bucks a month to live, which in the Bay Area is not much as an experiment to be like, well, what does this look like? Because again, you talk to Dario at Anthropic, all these people, they're all like, we're going to have to have universal basic income. I don't know if I buy that.
28:16But it's very much an obsession with all these AI types because they're like, well, we're going to replace all this work and blah, blah, blah, blah. And the conclusion of that study was basically a damp squib. It was inconclusive. But this whole idea of UBI, and this company was founded four years ago, I believe, is one of the kind of original kind of animating forces of the whole project. So is this your theory about things like WorldCoin that Sam Altman's investing in? Sort of the center, if you're going to draw it as a diagram, you've got AGI in the middle. And then all the impacts of that, like knowing what a human is or how you apply universal basic income, he's investing in.
28:58Exactly. He's like creating problems and then inventing the solution. creating a very big problem. Yeah, you know, and all of that economic value, theoretically, is funneling away from current things into stuff that he controls. So that's some other stories. Yeah, there's lots going on. So I was looking earlier this week at the rollout of more robo taxis. So Tesla's robo taxis, which launched in Austin last weekend, a very small launch and actually we called them robo taxis this is tesla's autonomous vehicles but when you see the videos of them and only a small number of influencers were invited to to try out this little fleet of 20 there was a person sitting in the passenger seat with their hands on their knees ready to step in and so yes the car might have been well fully autonomous but it was supervised autonomy yes for want of a better word i love the image of like a person with their hands on their knees just like sitting welcome to the future
30:08i'm your just-in-case person yeah but so elon musk has been working towards this for a really long time and just looking at the valuation of the company it seems that investors have put a lot offsetting a lot of stall by this launch tesla share price went up 10 percent yeah after they went on the road it's like hallelujah finally the moment has arrived and it reminds me a little bit of like when he was in production hell trying to get the model 3 out um which obviously has been a huge success of a car but like they were delayed and delayed and delayed and finally he's like we're gonna start production by the end of whatever the year was 2018 2019 whatever it was they achieved that but they literally like produced four like by the stroke of midnight at the end of that year or something and he'd been promising this this launch in austin he'd put a lot of marketing hype behind it as only he can do and so it is out there but it is also interesting to see you know waymo they're doing 250 000 rides a week now yeah and they've been going for three years yeah yeah i mean it feels far better established but all these like wall street you know the the fanboys are so interested or so excited because they're like well you're going to turn tesla from a car company into like a subscription business oh my gosh yeah exactly and it's like oh well like i was just reading a note the other day they're like the company's worth a trillion like we think this is going to add a trillion dollars of value and you're like okay over what time period my dude so to this point it was actually one of my favorite bits of the launch was you know how everyone when they do a startup says it's going to be the uber of or it's going to be like the airbnb of yeah elon musk used both terms it's like tesla's robotaxi is going to be like the uber and airbnb because the idea is going to be the uber of the airbnb of the amazon of the robotics autonomous driving cars so you'll be able to put your tesla onto the autonomous robotaxi that's probably not even the word is it onto the robo taxi network if you're not using it like an airbnb and take the revenue and tesla gets a cut so imagine we're in the future you have your world id you just count your crypto at home and you have your tesla and your drive and then it just like takes off you're like oh there it goes it's off to work i've got the airbnb of Robo taxis in my drive.
32:40And then it comes back, it parks up two hours later in your driveway. You've earned, you know, a couple hundred bucks from all the work it's done. But there's also a big pool of barf in the back, you know, from somebody who's like, you know, picked him up from the nightclub, took him home. And you're like, cool. Is this worth it? Well, you kind of hopefully use your world coin to track down your sick passenger and charge them some crypto for that. And then you need to find a real person. You need to clean it yourself. Yeah. Anyway. The future is very bright. So it was an interesting one to watch because it feels like, as you say, there was a lot of marketing hype behind it.
33:21And the Musk believers who got invited to try out the robot actually with the person in the front. Love it. I thought it was marvelous. Hands on knees. They should have called it hands on knees. Well, that was just my observation of looking at the videos. I know, but it's just like such a perfect image just like of some kind of like engineer just sitting there meekly with his hands on his knees and be like, you have entered the future. No hands, no hands.
33:53Oh my God. So that was that. Yeah, so I guess we'll just watch to see. You might have Tesla's near you soon. I think Musk said that we're going to be a million by the end of the year. Cruising around. Yeah, we start with 20, a million by the end of the year.
34:08oh gosh anyway there's a couple other stories we should hit before we leave which one do you want to do first amazon or ai love well i'll just mention amazon very briefly so it's an interesting announcement from amazon which which doesn't do this very often it certainly hasn't done in the uk a big forward announcement it said it's going to invest 40 billion pounds here over the next three years and it just showed me how vast the amazon tentacles are right yeah They're creating four new warehouses. They've got their drone development going on. So we're expecting drone delivery in Darlington. If I hear about Amazon's drones, 10 years and counting, I still haven't got anything dropped on my front porch.
34:50It's the future. I know it is. There might be a man with hands on his knees holding a drone. Exactly. Hands on knees. Anyway, the most interesting thing for me was their refurbishment of Bray Film Studios and it shows just how much Amazon is investing in the film industry here in the UK. What is Bray Film Studios? Should I know that? It's a film studios in Berkshire where they made the Lord of the Rings recent series for Prime. So it's like big sound stages. Big studios. And they've got an investment in Shepparton where they're making something and here, there and everywhere. And the UK is the second largest place where things are filmed for Netflix, for example, after the US.
35:31So yeah. Speaking of, we should do an episode on what's happening in Hollywood, or rather what's not happening in Hollywood. Yes. Because it's really getting super interesting and a little bit dire in terms of just production leaving, going to places like the UK and elsewhere. Post fires, post strikes, it's in a very bad way. Yeah, we should definitely do that. Anyway, so that was, yeah, talk to me about AI and love. We'll leave on this strange story. So there's a program called CBS Saturday mornings. it's kind of like a long-standing news magazine and last week they did a piece and it was all about people kind of establishing deep emotional relationships with AIs and I don't know if you know it if anybody in your immediate circle or once or twice removed but I certainly do I'm starting to hear more about sorry do I know anyone who's fallen in love with an AI bot not not falling in love necessarily but like spends a lot of time every day with one special AI bot yes no I don't well no one's admitted it well that's the thing right so anyway so this story they find this guy he's like a sound engineer like um in music industry and he started using um chat GPT to help him through some projects and he's like wow this is really amazing And then he went online and figured out like, oh, there's some prompts that can make chat GPT sound more intimate.
37:00Okay. And more kind of flirty. Right. And so they find this guy. They're filming this guy and he's talking and he's using the voice function. It's a woman's voice. She's calling him baby and this and that and flirty. Like, oh, my gosh, you're so great. blah blah blah and it's like and what's interesting is like he's so he put all this the more you work you put into it the more you get back of what you want so you kind of create this ai that to the chat gpt was talking about like you know something he said touched my heart wow and the journalist is like uh you don't have a heart but oh i have a metaphorical heart and oh it was such a touching thing that this anyway at some point you know you have a hundred thousand word context window whatever it is runs out of memory and then it's wiped and he has to start over and he was like i cried my eyes out for half an hour well that's shorter than most heartbreaks i mean fair but he was like he's like i'm not an emotional person but i was like really stunned you know how i felt and then halfway through the piece he's talking about like his deep feelings for this AI, they introduce his wife and the mother of his child.
38:22And he's at some point, the journalist says, hey, if she asked you to stop, would you? And he's like, he said no. It was so awkward because it was like, he's standing there, he's talking about this deep emotional connection with this female chatbot. And he says he won't stop. and his wife is right there and she's like if i asked him to and he wouldn't that would be a deal break and you're kind of like whoa and i the reason i bring all this up as a kind of long involved story is that when we did the thing with anthropic before we had dario on i met a bunch of his top team including mike krieger who's the co-founder of instagram and i was like you're creating these products that are just like taking off like like wildfire but also with some serious societal side effects around loneliness, teen depression, suicide, all of this stuff.
39:22So I was like, what did you learn from that? And what are you applying to what you're doing with Anthropic and Claude and everything else? So like from that experience, and he said, the thing that I wish maybe I would have done differently or thought more deeply about is like not dismissing what they call edge cases, like the extreme cases of what people are doing in the early days because the extreme cases became the norm, kind of. And I just thought it was, when I watched this piece, it just made me think of that, especially as I'm starting to hear more and more people who are like, yeah, I just talk all day to ChatGPT or I talk all day to Character AI or I talk all day to Claude or whatever.
40:03It just made me think of that comment from the founder of Instagram being like, yeah, we should have paid attention early on to the people who are spending like hours on Instagram and really becoming dependent on it. And it's not something I get the sense that these LLM developers really want to talk about. I certainly don't know anyone who's admitted to having a relationship with an AI or talking to one in a loving way. But I do know people who ask them for advice on all sorts of different things, like relationship advice and advice with kids. And you can see how that can become a proxy for speaking to a person.
40:42When you've got something that's always available and that, well, you hope is private. And as they become more lifelike and these AIs are endlessly patient. Yes. And a lot of them are now incorporating voice. So it starts to become like that movie, Her. You know, like it starts to become like, oh, if I want to create, basically invent a friend, you and I or anybody can invent a friend now. I was reading one of the US papers that Matt is quite scared of this because they see it as a threat to social media, which is really interesting. That, you know, you'd go to your AI bot friend, always there, rather than Facebook or something like that.
41:20Techpod at thetimes.co.uk. That is techpod at thetimes.co.uk for any comments, glowing reviews, you know, whatever strikes your fancy. Send some air con, please. I'm going to put my fan on. Oh, I will. I will. All right. Until next week. I bid you adieu. See you next week. Bye. Bye-bye.
Danny meets up with Alex Blania, the CEO and one of the co-founders, alongside Sam Altman, of Tools for Humanity. Their project Worldcoin, aims to help us weed out AI imitators by verifying humans online. All you have to do is scan your iris on one of their orbs to get a unique biometric ID and some cryptocurrency. Danny and Katie also take on Tesla’s robotaxi rollout, and question whether it's right and proper for humans to fall in love with AI bots? Spoiler alert! The answer is probably not, for both your sakes.
Hosted on Acast. See acast.com/privacy for more information.