Cybersecurity and Future Risks

Hello and welcome to the Times Tech podcast where every week we unpack how technology is reshaping business, culture and everyday life. I am Danny Forton outs here in Silicon Valley where he's trying to keep his children occupied during the summer holidays. And I am Katie Prescott in the city of London . And Danny and I are on YouTube. On the YouTubes , find us . Isn't this a glorious time to be alive? So if you want to, you know , watch, instead of listen, head to the Times Business page on YouTube. There's a website called YouTube. com . They show short form and long form videos there. Goodness me. I must check it out. Yeah, it's a new thing. It's a new thing . It's echemology . Well, good to good to know we're keeping up. So what have you been up to? Oh, I have been hobnobbing. Oh yes . I've been hobnobbing with friends of the pod in West London because it is London Te ch Week here , which brings not only the great and the good of British tech, but also lots and lots of your fellow San Franciscoans to London to talk about technology . I always feel like these things are like there's all this hype, right? And then you go to the big conference or the big do and then you just find yourself like walking around from one venue to the other tent drinking bad cop like lukewarm coffee. You're like, Oh, I've got to go see that session and then you go sit in the session and after like three minutes you're scrolling on your phone and you're like, what am I doing with my life? Actually, it has been fantastic. There's one main stage at Tech Week. The Prime Minister launched the event. It's been really notable that over the past few years the Prime Ministers have started turning up for it and I think it says a lot about how tech's been rising up the agenda. Lisa Sue from AMD, Friend of the Pod flew over just as her cousin, Jensen Huang did last year . So lots of big announcements from the Government about investment in tech, but also from the US . It was a bit strange this year, I have to say, because the theme on the stage all the time was sovereignty . It was British tech, how can we promote British AI? How can we invest in British businesses? How can we make sure that British companies are winning British government contracts? And so you had this almost slightly awkward tension between the sort of geopolitical split , push against the US almost like, you know, talking behind the back of the most popular person in class or something. It's like, Oh look, they're the big successful one, but we want to be like that too , but then at the same time welcoming lots of American companies to the UK who were talking about investment here. That created a tension, I would say at the event . Because there's the sovereignty bit and then there's the social media bit and it's all swirling . Exactly. And the Prime Minister on stage is opening the event to an audience of tech business leaders and politic y people issued an ultimatum to US tech about look, you've got to stop children being able to exchange nude images on their phones or we're gonna, you pass know, legislation in the next week. That's reasonable. It feels reasonable, but it was a slightly unusual message at a conference to promote Raw Ra too. Yeah, there wasn't a t shirt cannon? No this wasn't even I'm not, I'm so glad I wasn't there. If I'm not getting hit in the face with a balled up t shirt at high speed , I'm not going Is that the bar? If everyone if you want Danny Fordson at a conference, you need to shoot him in the face with a t shirt. Exactly. Otherwise, what is the point ? So that landed I think slightly awkwardly obviously, it wasn't, meant for that audience, right? It was the rest of the UK that was listening to that. And it played really big in the news here . But I think the fact he was there was also a bit of a damp squib for the tech community because he's very much seen as on his way out . His position is very shaky as Prime Minister . So there wasn't, I would say the House view, there wasn't overwhelming delight with his message . Because on one hand , you have like this whole social media band kind of movement that's happening around the world . And you have all these lawsuits which we've talked about on the pod here. Like there definitely seems to be like this huge shift. It feels like the public is very much like yeah , ban social media for kids. It seems like politically a really easy choice to make . But at the same time, to your point, you're talking about tech sovereignty. How can we convince these big U. S. tech companies to invest on our island ? I think that's the tension, right? Of like, this is the thing that everybody wants, but I'm sure in the back rooms you have, tech companies being like , yeah , that's that's a nice social media law you have there. It'd be a shame if we stopped investing in your country five billion dollars that we were going to put into your data centers over the next three years. And it was really notable that the event headlin ed with British Prime Minister followed by Lisa Su . And so it still feels just as it did with Jansen Huang last year. So it feels like we're bringing in the big US tech titans. But I should say the British heavy hitters were out in force as well . So Matty, the boss of eleven labs was on stage , Alex Kendall from Wave . There were some really big British names too. So it was good. I certainly wasn't wandering around checking my phone. But we're only halfway through the week, Danny. I mean, there's still Founders Forum to come tomorrow. There's Founders Forum. They've got there's this company they fire rockets into space . Can't remember what they're called? Anyway, I think they're having an IPO kind of big That's happening . Yeah. Good, I'm so glad I'm not a Sunday journalist happensing on Friday night in London. Thank you . So yeah, so it's such a big week out here. And we should say before we move on to what else is happening , we have a great guest this week.s Do we not? We have got an amazing guest actually who is helping us with the Times Tech summit this year. Yes, it is G two Patel. He is the president and chief product officer at Cisco . They're building the hardware and software that kind of enables this whole AI boom. He's an expert in all things egentic. And you know, don't sleep on Cisco . They were the dot com darling then they were the dot com kind of dog and they're now wor th half a trillion dollars and they've just come roaring back because they made up all these big bets on AI and G two's right at the heart of that. So we're going to have him to talk all things AI later on in the pod. While we're on the subject of AI . Yeah . There was another small development this week, wasn't there? This blog that Anthropic put out yet another bombshell post from anthropic when AI builds itself . So this was all about recursive AI, which you know, I guess you've been saying is a big theme in Silicon Valley at the moment. Yeah, that's right. I feel like if you want to understand what Silicon Valley is thinking , take half an hour and you will understand. One is there's a video on YouTube that YouTube. com website for videos. Yeah . By Tom Blomfield who founded Monzo. He's now he's a partner at Ycominator. And gave like this thirteen minute talk to the current batch of startups at Ycominator called How to Build a Self Improving Company . And it's all about this notion of like basically your job in twenty twenty six is to build a different kind of company where you spend all your time shoveling your job is to give as much quote unquote context to your favorite AI, whatever it may be. So that means all your emails, all your text messages , put microphones in your recording in your meeting rooms, transcribe those. Just feed as much data in as you can . And then this brain has control of an army of agents that kind of improve the company while you sleep, you wake up, there's a new prototype, there's a new product. There's like a bunch of sales calls that have happened, whatever it may be. So you take that and then like overlaid on that is this other piece from Anthropic , which as you say, is called When AI builds itself, it's this blog post they just put out that explains what they're seeing . And it's kind of crazy. They were saying like, if you go back like a year and a half , low single digits of their code was written by their coding agent. Now it's over eighty percent . And they're like , the role of humans is changing pretty rapid ly and basically it's shrinking . And they don't really have any answers as to like how this goes or where this goes. But it's another kind of scary example of what this all means for us humans . I should say looking at some of the startups that have raised the most money in the UK over the past six months . Recursive AI has raised five hundred million dollars and ineffable intelligence both working in this space and they're only a few months old . That's the astonishing thing. So it feels like this is really the next focus for everyone 's attention. Yeah, and that's what Blomfield said at YCominary. He said like companies that are basically leaning into this idea of a company brain commanding an army of agents are getting to demo day, like each batch has three months to work on a product all leading up to demo day where they get up on stage, say, this is who we are, this is what we make . Please give us money, venture capitalists. They're getting to demo day with five times more revenue per employee than historically over the past twenty years . It's worth just reading a bit of the anthropic blog because it ends on this very stark warning , which is basically everyone should probably stop. So it says if it were possible to effectively slow the development of this technology to give ourselves more time to deal with its immense implications , we think that would likely be a good thing . But basically it said you know if you slow down , it lets the least cautious actors catch up which could leave everyone less safe . So they say without a global coordination mechanism, companies and governments will have to make difficult decisions about safety while under competitive and geopolitical pressures , which is quite a big statement for a company developing this . And you know, here in the States, kind of the background music to all of this is China . And every time someone says something like what Anthropic just said there's like, you know, we should really probably slow down because this is getting this is escalating quickly. You have somebody else who said, look, if we slow down, China's not . And this is like nuclear weapons and we cannot afford to quote unquote, lose to China . So full steam ahead . And then overlaid on that is the IPO piece. So we have SpaceX this week. Open AI just filed . Anthropic has filed. So three trillion dollars IPOs . These companies have raised in open AI and Anthropic, I believe more money than any company has ever raised ever . I think opening eyes raised one hundred seventy five billion dollars . They have investors who want the money back. You know, so the financial incentives are immense. So I just think between the financial incentives and that whole we're in a race with China , no one's going to slow down. It's really noticeable that that was all anyone was talking about at London Tech Week. You know , the stuff about technological advances wasn't really there. It was all about geopolitics. I mean, you spoke about China there , but there was a huge amount of hand wringing here, as you can imagine , with the background of open AI's IPO being announced, but also SpaceX is coming up about where the money was going to come from in Europe to keep up with the US and to make sure that there is some semblance of tech independence here. I interviewed the boss of the new boss of Deliveroo on stage, who's a Finn, responsible for slush, which is like the big tech conference in Europe . And he was saying, It's really worrying that Spain is one of the fastest growing economies, and that's because of tourism . He said, because of people going there to sun themselves and I've heard also comparisons between the tech scene, unfavorable ones and fashion here . So there's a real worry about how on earth when you're talking about those huge numbers and that speed of change that anyone can catch up here. One of the investors on stage with George Osborne who and that whole panel with George Osborne from now open AI was about sovereignty . There's a German investor called Judith Dada, who said if we don't drastically change course on AI in Europe, we'll end up politically sidelined with an economy that's stagnating, tech we can't govern and without the ability to continue with social welfare systems . So it's a serious topic of conversation here and you can feel everybody getting worried about it. And you know, I've covered the sector for a long time and it's been an issue. I mean since even before I was a journalist, I remember working for an entrepreneur in London the sort of issue of where's the funding going to come from? Why is the US the best place to scale a business The sale of Google Deep Mind was constantly brought up as a sort of source of pain Why did we sell this British gem to America . So that's always been there as a thread, but it just feels really loud at the moment. And it's what's so interesting is like me being here. Nobody is even stopping for a moment to even think about what's happening over there . And then over there, that is the rest of the world where hundreds of millions of people live in like the largest one of the largest kind of economies in the world . And people are like freaking out rightfully about how they kind of fit into all of this and how they're going to leverage it. And I think it's really important that US companies do wake up to this because it is going to affect them. Things are changing here about how money is spent, how big contracts are awarded, you know, big government contracts to U. S. firms. People are really, really mindful of that. So if people aren't thinking about it, I think they soon will be. So self building AI was not on the agenda. I think people were talking about a bit, but on the main stage , it was really, you know, these sort of geopolitical concerns, talk of job losses and worries about the idea that you'll be able to start a company with many ag ents and you won't need to hire staff at all. Ivan Zao, the co founder and CEO of Notion based between New York and San Francisco, spoke about this. It feels like there's a lot of talk about we're going to live in the future that one person can create a billion dollar business and the world will be run by one person startups or zero person startups . Personally, I prefer not to live in that world. We care a lot at the notion we care a lot about human values, the values such as human teamwork and camaraderie. I think those are good values that we're bringing to the next generation of knowledge work and therefore it',s meaningful not just for in terms of productivity, but in terms of preserving human value to bring teamwork how agent works. And usually at least at the moment, a company will be run by a CEO and CIFO has multiple people have to make decisions. So a company will be an entity of multiple people. Therefore, it's quite valuable to make sure we'll figure out how to get an AI agent to work with team of humans. What do you think? So are you going to work? I mean, you and Norman, you guys have a special relationship, as previously discussed . Well, seeing as you're not, you know, you're not here next week, maybe we'll get Norman to co present the program. There's an idea. Everybody 's actually scary. Don't do that because he's very funny I order you to not do that yeah because that might be crossed the tech agenda he sends me a little briefing every day . Yeah no that's that's scary that's scary. This is a micro example of the great job fears we're talking about all the time . Our guest today will certainly have some thoughts on all of the above. It is Jitu Patel. He's the president and chief product officer at Cisco and we're going to talk to him about this new era of self building AI, self improving AI , and why it could have great consequences and also disastrous consequen ces for cybersecurity and everything else. That's in just a moment Welcome back to the Times Tech podcast. Our guest today is GT Patel, the president and chief product officer at Cisco . Cisco essentially built the foundation of the modern internet. It was founded in the eighties by a group of Stanford scientists and pioneered the first commercially successful router , which is basically a specialized computer that allows completely different networks to talk to one another. And during the dot com boom , they briefly, very briefly became the most valuable company in the world. And then of course we all know what happened next. There was the crash and ever since then they've been working their way back up but, a few years ago they made a big bet on AI. And now they build their own silicon, design the systems that it powers, write the software, and provide the security that watches over it all . Their share price is now higher than it was in two thousand , but it's been a long journey and it only got there again earlier this year. Yeah . So while you might not have bought a Cisco product, today if you use the internet, connect to the office wifi or stre,am a video d.ata Y isour almost certainly passing through their hardware and the company is now valued at a cool four hundred seventy billion dollars. Half a trillion dollars, not bad. Welcome to the podcast, GTU. I think I want to start with something really basic, which is , you know, I grew up in the Bay Area, so I know all about the kind of Cisco lore , but a lot of people don't. Can you talk about what Cisco is today? Sure. I've been here for six years. I wasn't here for the ups and downs that happened in the past, but I will tell you that the reason the ups and downs happen is when you have a massive platform shift and a mega trend that's occurring in the mark et. If you actually catch the wave , you tend to enjoy the fruits of the momentum of the market. And if you don't catch the wave, you don't enjoy those fruits . We missed the cloud wave , but we are, you know, squarely extremely well positioned for the AI wave . We were very careful not to miss the AI wave, but it was also kind of lucky for us that the capabilities that we had actually formed the critical infrastructure for the AI era. Like we have high performance, low latency, energy efficient networking, we've got optics, we make our own silicon, we actually have security for kind of protecting agents from the world. We also have security for protecting the world from agents and the way that they behave. We've got observability, we've got a data platform. We've got applications . So like there was this vertically integrated platform full stack that was fully co designed that we were unique enough to be able to really capitalize on. And by the way, the AI wave requi res that. Without GPUs, without a network are not that useful . And so the way that you should think about Cisco right now is it's the critical infrastructure for the AI arrow. Could you sum up very simply and briefly then what your customers are buying from you and who those customers are? Yes, basically the way you should think about Cisco is we are an infrastructure comp.an Soies if you happen to be there's three big outcomes that we help drive for our customers. First one is because of the AI move right now, there's a tremendous amount of investment that's happening in data centers. And these data centers need to be AI ready. So we provide them all of the apparatus that's needed for data centers being AI ready on the networking security observability side of the house. And the way we think about it is that's where the digital workers live. The second outcome then is all of the workplaces where people work, whether it be a campus, a branch, an office , a factory floor, home office a restaurant, that needs to be future proof for AI as well. There's a big movement that we should talk about right now, but inferencing is going to happen everywhere . And as inferencing happens everywhere, you'll see people buying like little Mac minis that are sitting by their laptop that are running agents . And so inferencing is happening on everyone's desk side, not just in the cloud. That's the second big area of our comb is future proof people's workplaces . And then the third one is this notion of have resilient infrastructure. We call it digital resilience where if you do happen to have an outage or a breach of some sort because of security , how quickly can you actually get back up on your feet as a company so that you don't have massive amounts of downtime which can be a business ending event. So those are the three big outcomes we solve for. Before we kind of get into the nitty gritty, I just want to talk about this moment in which we find ourselves in the market . We're recording on Wednesday before the big SpaceXIPO. You've had open AI and anthropic both file . So the idea is that in like the next six months or nine months , we're going to have three trillion dollar IPOs . And it feels like , you know, we're both in California. Like the gold rush is happening again . And I'm just wondering from where you guys sit , how sustainable does this feel? Does this feel to you because you're the one selling this kit to the companies that are building the systems and running the system . Where it's your take from where you guys sit ? You know, the very common question that gets asked is Are we in a bubble? And it's a multi layered answer. Let me actually first start with two things can hold true at the same time. We are in the midst of the most seismic platform shift that has ever been experienced by humanity . And there also might be some companies that have frothy evaluations. None of the three that you talked about are in that category, by the way. So that's none of those We'll see what happens on Friday. No, I'll tell you why because I think there's a very sustained demand for what they're doing. And we should talk about this in detail. The second big thing to keep in mind is we are right now constrained in three major areas . If you think about like what would make us not be able to harness the full potential of AI ? There's three major kind of impediments. The first one is there's an infrastructure constraint. We just don't have enough power, compute, network bandwid th and memory in the world to go out and satiate the needs of AI. That's number one. Number two , there's a trust deficit. If we don't trust these AI agents, we're not going to use them . And so you have to make sure that you establish trust in the delegation that you do. The difference between simple delegation versus trusted delegation is the difference between market leadership and you know the extinction on one side if it's just simple delegation and market leadership, if it's trusted delegation. So I think there's a huge , huge kind of opportunity over there. And then the third one is these agents make decisions on our behalf. The difference between an agent and a chatbot as we move to this agentic era is that these agents are like digital coworkers. They get augmented to our teams and they're going to conduct tasks and jobs for us almost fully autonomously . And as that happens , you have to feed it context because that's how they make decisions. You and I ingest so much context every single day and that's how we live our lives and we kind of connect things together. If you don't give that context to agents they'll still make decisions. They just won't be great decisions. So there's a context gap that needs to be in where the agents have to be enriched with context. All of these three things are constraints right now, but when you say in a bubble , you have to compare it to what was happening in the nineties with people equated to the late nineties and dot com bubble . The infrastructure was built out anticipating demand . Right now, the infrastructure is being built out and the moment the infrastructure is made available, the demand is consumed and the demand is not consumed just for vanity purposes. People are using these products on a daily basis and it's become a habit and that is how they get worked done. Like we have thirty thousand engineers and these thirty thousand engineers use these tools to code and we don't actually we now have a product that is hundred percent written with AI, no human lines of code . By the end of the year, we'll have six products one hundred percent written with AI. By the end of next year , we will have seventy percent of our entire portfolio that'll be one hundred percent written with AI. That's crazy to imagine, but that's happening. So like there's an intrinsic demand signal and a value creation that's occurring right now. And so we have to so I think there's there's going to be this is a secular shift, not just a blip that's going to happen or like a brick spike and then everything's going to go back. That's that's at least my take on it. But that doesn't mean that you're not going to have a few frothy companies that get overvalued, but that's the definition of a secular shift is you have a lot of experiments that happen during that course in time. How long before you have an agent on your board? I'll give you an example of what we're thinking about right now. Wouldn't it be great if every single person had a digital chief of staff that attended every meeting that they were in at the end of the day gave you a summary that said, Hey, here's all the meetings that you attended. Only for you, this is the context that you need to have . And boy, wouldn't it be nice if nothing slipped through the cracks? Because right now what ends up happening is like I'm I'm not the most kind of, you know, I'm more left brain than right brain. I'm, you know, and so when you when you start thinking about what happens in a day if I sit in ten meetings , invariably I'll forget something that someone said that I'm not going to follow up on. And then that was a huge waste of time for people because a lot of people prepared. They were in the meeting . Ideally, wouldn't it be great if you had a digital chief of staff in that meeting? By the way, that does not change for a board of director as well. Like, wouldn't it be great if you had someone that actually was providing the right level of governance input? They were providing the right level of strategic input, they were able to concatenate information and provide original insights as we were having strategic discussions . It's super valuable to have an abort. But how long? Until it happens . I think right now if there's a credible idea , no idea is passed a couple years. The biggest thing that happened in this industry was which was actually which was actually a genius move and frankly anthropic gets a lot of credit for it . The genius move was they decided that coding was the app that they were going to go after first. And what that did was it made this entire like clawed code now builds claw code and codex builds codex. And because of that, that exponential curve has become a vertical line. And then these models are becoming, you know, self recursive and they're learning all the time and they're teaching themselves as they're actually going out and serving queries. So what I tell my team is if you test something out and it doesn't work, the normal human tendency I' tom say going to put it aside for about six month to twelves, I'll come back to it later maybe or I'll just give it up and say this never worked. Now what you have to assume is that if it doesn't work today, you have to assume it's going to work in the next two to three months and the amount of time it's going to take you to plan is going to be far greater than the amount of time it's going to take for them to fix the product to make it work. Should we pick up on this idea of recursive AI? It's something we've been discussing in the program today. Clearly we had that warning from Anthropic this week on it . What are you seeing there and what are your views on it and its uses? It's actually not that different from humans. Humans are self recursive in nature, right? Like we continue to keep learning and every single time. Like even if you ask a professor or teacher, when you're teaching, are you learning? You are because as you're teaching people there's different ways to articulate the idea that comes to your mind and you're making connections within your mind. And so you 're constantly learning and training your mind even when you're going out and importing knowledge . I think morals are starting to get to be in that kind of domain. It's pretty exciting because the class of problems we'll be able to solve with this is I don't think it's actually sunk into people. People kind of listen to it. It sounds good, but think about every disease that we'll be able to solve at some point in time and they're not too dist ant future. Think about poverty being eradicated in massive scale. Think about how education would be disseminated to every human on the planet . And there's equal rights to participation that anyone has because knowledge no longer becomes this kind of scarce commodity that only a few people have. Those are exciting parts of the future. Now, by the way, you have to be super cognizant of not having rose colored glasses and understanding the risks of these technology as well, which we should talk about. But the upside is undeniable at this point. And what will happen in these things is you'll always have these waves where people will be like, this is going to replace all humans and humans won't have any value to add to society. And then literally the next week, people will be like, this thing doesn't work and I'm just gonna it was just a passing fad . But the reality is there's a sustained level of imp rovement that's happening. And when you look back five years from now and say, imagine how we lived our lives back before Chat GPT came about , it'll seem very foreign , just like it seems very foreign right now and you live your life without a cell phone or constantly being connected. Right. Right. And then so shall we end on some scary stuff? Cybersecurity . Okay, so on safety and security I think there are three things that need to happen in safety and security. Firstly , these agents are going to be attacked by the world. You're going to have prompt injection attacks, you're going to have data poisoning efforts so that you can alter the behavior of agents . So what you'll need to do is you'll need to make sure that you protect that agent from the world . So there's a whole bunch of safety and security capabilities that need to be in place so that you can protect the agent from the world. Number two , these agents themselves are non deterministic because they run in non deterministic models. They're kind of unpredictable. And the way I think about agents is they're like teenagers. You know, they're supremely intelligent , they have no fear of consequence and as a result, they can sometimes do really stupid stuff . And so what you have to do is you have to be able to intercept it and you have to be able to assess the behavior of an agent algorithmically and know that when this behavior is not being exhibited , you know that something is wrong and you have to keep an eye to be able to intercept it. So the second thing you have to do is be able to protect the world from agents , protect the agents of the world, protect the world from agents , and assuming that every bad actor, every adversary, every threat actor that you have in the world is going to use AI to exponentially increase the amount of attacks that come our way on critical infrastructure or on any infrastructure . We cannot go out and combat those attacks at human scale. We're going to have to do that at machine scale. We are, of course, spending hundreds of millions of dollars, billions of dollars towards this problem right now. And I'm pretty optimistic about the kind of things that we're doing where I think for the first time in the bull case scenario you might actually have the defender have the advantage over the adversary, which has never happened in the past. In the past, the adversary has always had the advantage because they have to be right once, defender has to be right every single time. This is the first time that the tables might turn. On the other hand, if we don't effectively operate at machine scale and try to do things in the traditional way . You know, catastrophic effects in society because think about a power plant going down, a hospital going down, a human lives will be cost if that happens. So I think that's the area that we're spending an enormous amount of energy . That's where I don't have my rose coloured glasses. I actually think about the downside pessimistic scenarios all day long . Well, so what do you so I have thoughts about I have so many thoughts about G two and what he was saying he said right at the start, you know, he's obviously hugely optimistic about the power of AI and there are many upsides but it was quite clear to me talking to everyone at the conference this week that there's huge concern about he talked about eradicating poverty, but also what it means for people who are left behind, who aren't sitting in a very cushy, well paid tech job. Actually, it does feel like to me, particularly here in Europe, there's a real two speed economy at the moment and things while the future with AI might look very rosy, I think we've got quite a rocky road together . Indeed, indeed, but wait, but wait, wait, what's that? What's that sound? Just coming on no I don't That's the listener email of Claxon Cool . Oh wow, look at this. Yeah, we have had a few emails kindly sending their regards to Norman , my AI agent. And this week we've had one from John Dudley. I must tell Norman he'll be completely delighted. John says Hi folks. I enjoyed last Friday's pod. One thing this is important. It's not confidential that Anthropic have filed something with the SEC as they are free to announce that, it's the content of the filing that's confidential. This is standard practice as it gives the company and the SEC the opportunity for a private review before the public filing of the S one listing document keep up the good work and give my regards to Norman and John. He is right because we were joking last week because Anthropic pinged around an email saying Confidential filing to the SEC, which made it sound like it was like, We've got a sheet but we're going to announce announcement. Here's our secret plan . Well, actually they weren't saying it was confidential. They were saying it's a filing which we've made and it's of a certain type of conf aidential ty pe. Fair, fair point. Thank you, John. Thank you, John for being such an assiduous listener. And Norman will be so pleased. I mean, he always likes recognized. I don't want to I don't want to kind of keep bagging on my guy Norm an, but like he doesn't have feelings, okay? Just a guy. And he's not gonna replace me . He is not gonna replace me. We'll see in Mexico. All sorts of cats away Anyhow, well that is it for this week's episode of the Times Tech podcast. If you have any other questions, topics, ideas, opinions . You want to get in touch with us about, you can email us at techpod at the times dot co dot UK techpod at the times.co. uk I am not here next week. Have a good day . I refuse to be replaced by a machine . You can't do anything about it, Danny. I'll put it on the right. , well we'll let you know how we got on. Well, hopefully I'll be back. Hopefully I'll be back everybody. This might be you