The REAL Reason A Financial Crash Is Coming Next

And the thing, the principle there that allows you to see what the future might look like when it's exponential and not linear is imagining any rate of improvement. And so this is how I think always in my business is when I'm thinking about which bet to make, I imagine if there's just 5% rate of improvement per quarter, eventually this happens. And so with some things we've done on the show, we started two years ago, and for the first 18 months, it didn't work. And then six months ago, it exploded because eventually the rate of improvement got to the point where it was viable. And now it's the single most important disruptive thing we've ever done. And I think about the example I just gave there of agents. I'm like, if you just imagine any rate of improvement, just imagine 1% a year. And obviously it's way more than that, but imagine 1% a month. Eventually we get to a point where a team of agents can launch... If nothing disrupts it. One thing that is very likely to happen in the next three years is the whole thing comes crashing down. It doesn't financially make any sense to build the data centres that we're building. But you're not saying that AI is going to disappear? The technology itself is remarkable, but the financial model just makes no sense whatsoever at the moment. At the moment, which is kind of typical of big bubbles, right? So over the last 180 years, there have been infrastructure build-outs that have bankrupted the economy consistently again and again and again. In fact, there's a pattern. Whenever we spend more than 3% of the overall GDP of the economy on an infrastructure build-out, it bankrupts the economy briefly for 10 years. So this happened when we did railway tracks. In fact, it happened twice in the UK and twice in the USA. Then we bankrupted ourselves putting the electrification grid in place. We bankrupted ourselves putting highways in place. Now there's something that makes this worse than ever before. And that is that when we built train tracks, they lasted for 100 years. When we built roads, they lasted 50 plus years. The telecommunications fibre optics, 30 years. All of these investments were multi-decade investments that we could get benefit from and leverage for decades. Data centres last three to four years before they need to be replaced. So we are building something that has a three to four year life cycle that costs hundreds of billions and it has to be replaced every few years. And there is no financial model attached to this that justifies it at all. So one thing that I'm predicting is in 2029, 100 years after the Great Depression, we will see a massive financial meltdown based off the back of these data centres that are being developed at the moment. And for someone that doesn't know what a data centre is? So it's like a big Walmart or an airport filled with computers and those computers are running AI. So every time you go on AI, your request is going off to a big computer in a Walmart-sized building somewhere. But those GPUs, those computer stacks, those big ginormous computers, they are like iPhones. They last about three years before they get superseded. This year ahead, we're going to spend $650 billion. $650 billion sounds like so abstract. It's the equivalent of giving every single person in America an iPhone Pro with AirPods. But here's the crazy thing. 95% of people who have been given the free iPhone are not willing to pay for it. And the tiny number of people who are willing to pay for it, they're only willing to pay $20 a month, most of them. So there's $20 a month, everyone's getting a free iPhone effectively. And then a small 5% of people are willing to pay $20 a month. The numbers are just astronomically out of balance. The numbers are astronomically out of balance. So what happens next? You're predicting in 2029, there's a financial crash. There's not a 0% chance that the financial model around what we've launched with AI is catastrophic. It's a huge financial problem that very few people are talking about. So let's talk about, in a world of AI, what are the skills that survive? And what are the professions that you believe survive? You've got children, Daniel. When they come to you and they say, Dad, I'm hearing about all this AI stuff, and I've seen these robots in China doing backflips. What should I be learning now to make sure that the next 20, 30, 40 years of my career is prosperous? It's the entrepreneurial skill set that is the most important skill set. It is essentially the skills of identifying an opportunity, prototyping fast and cheap experiments to see if you've got something that people want, taking it to market, making your initial sales, scaling up to your addressable market, and then exiting, and then coming up with a new idea. Because even if you're in a big corporate, they're going to want to prototype things. They're going to want to spin out new products. They're going to want to come up with new initiatives. And we need to step through that process of, is this a good idea or not? Can we validate it? Can we scale it? And you want to go through that process as fast and as cheap as possible the way an entrepreneur would. But on that first point of identifying a good idea, now, is there any way to know if the idea that I have is a good idea or a bad idea? Yeah, this is a step that entrepreneurs call validation. And the rookie entrepreneurs that fail, they don't do this step very well, right? So they don't do market validation. They don't do product validation. They simply get excited about an idea, and then they go all in on the idea. And what really experienced entrepreneurs do is that we come up with ways to do fast, cheap experiments. So for example, last year, I had two ideas, and I actually kind of liked one a lot, and I liked one a little bit. And what I did is I set up a waiting list campaign, and I basically invited people to join the waiting list for idea number one. And about 750 people joined that waiting list, and they filled in a set of questions so that I knew that 750 people were interested in this idea. This was my favorite idea. And then the second idea, I invited people to join that waiting list, and 4,500 people joined that waiting list. So even though this wasn't my favorite idea, it was way more exciting for people, and it was sitting on top of a much bigger need. So off the back of collecting 4,500 people on the waiting list and collecting all the data and the information, we went to some angel investors about a week later after doing that, and we raised a quarter of a million pounds, so about $300,000 or $400,000. And all of that took about a week or two. With that example, how do you know that the business itself and the delivery of that idea actually meets their expectation? Because I can think of multiple examples where someone clicked on an idea that I had, but actually the delivery of that idea fell short of expectation. Because sometimes things can sound good when you see the flyer or the poster or whatever, but the reality of the experience is actually not that great. So let me walk you through the six steps quickly. Step one is called Founder Opportunity Fit, and Founder Opportunity Fit is finding something that you want to do. Step two is Validation. So this is where we actually try and see, is there a market? Could we build something? Could we sell something? So that's two validations. Can we build it? Can we sell it? Step three is called Product Market Fit. And this is the next step where you're actually trying to figure out, can we actually make this live up to people's expectations so that they really like it? Is there a group of people who buy this and they're happy with the purchase? And we want to do that carefully and cheaply. And then finally, the next step, once people say, yes, we can do that, we go to market. And go to market is making sales. And then after going to market, we scale up, right? Step five. And then the final step is we exit. So we go through those six steps, and we call that a value creation loop. And that value creation loop goes through those six steps, and each step has its own little job. And what we're trying to do is just validate, go to the next step, achieve the milestones, go to the next step. And that's how we do it as entrepreneurs. So in a world of AI, are there any particular opportunities that you think are new and exciting that anybody listening right now could pursue to make money, whether it's passive income on the side, or to leave their current role and pursue entrepreneurship before? Well, the starting point is that everything is an opportunity. Because what do entrepreneurs do? We find things that could be done better, faster, cheaper, or with more emotional benefits. But has AI created any particular interesting opportunities in your view that you think that's a huge arbitrage opportunity? I think one of the best opportunities is a small SaaS company. So software as a service was an elite level business opportunity that very, very few people could enter. That was something that only maybe tens of thousands or hundreds of thousands of founders at most globally could have a software company. So it was very, very elite. And the reason it was elite is because it was very, very hard to mobilize the talent required to build a software company. You probably needed maybe 10, 20, or 30 developers to build a software product that would actually be a good software product. You needed to raise millions of dollars to get through the costs of developing software companies. And you probably needed 10,000 customers to have a breakeven software company. Now, because of AI, all of those costs have massively come down. And there are software companies that are wildly profitable that have 500 customers or 1,000 customers. They service a tiny little niche. They attach a community and some media and some training to those little software niches. So one of the most exciting things is that if you essentially said that, I would like to take what I know, turn it into a playbook, take that playbook, ask AI to advise me on what software we could create, and then use AI tools to build out that software, you're able to get very deep into that journey for almost no money. I was thinking about this the other day, because in my company, we're rebuilding our entire ATS ourselves. So an ATS is basically the system you use when you recruit people. Everyone's name goes into the ATS. It's kind of like a search engine, like a database of where candidates who have applied for your company currently are. And over the last couple of years, I've paid tens of thousands of dollars every year for an ATS for my companies. And this year, I thought, you know, actually, I reckon we could build an ATS in a couple of weeks ourselves, and it could be bespoke to us. So we started that project. I saw the first version of it yesterday. We started a week ago. And the first version of it I saw was like, oh, this is significantly better than what we're using because it's bespoke. And this made me think about the opportunity of creating software generally, which is that in a world where it becomes, as you say, cheaper, easier, faster, we're all going to be using more software, but we're probably not going to be paying other people for their software, especially when it's these kind of tools, productivity tools and systems. If it's purely a tool, it's very easy to replicate. But if it's a tool that comes with a community, if it comes with education and training, if it comes with agents... Education and training. Yeah. So like, for example, if you were to do an ATS, right, an applicant tracking system, a hiring tool, and you had the DOAC system and method that I can go and attend the training as to how you onboard people, how you create culture, how you do things at DOAC, I'm now interested in the software because it comes with that exciting training. If there was also a bootcamp that I was attached to that product, if there was also an annual retreat or a dinner series, if there was also some funding opportunities that came with that. So once upon a time, you were just simply going to focus on creating a software tool because that was so labour-intensive and also it had a natural moat or protection around it that you didn't have to get creative. But now we live in a world where the AI lets you to build the product pretty quickly. It helps you to do all these other things that we talked about. And suddenly it's actually quite a fun product that you can spin up. What you've already created would have probably cost $500,000 in six months. Oh my God, six months, I wish. Not that long ago. Yeah. It would have probably taken about 18 months total, end to end. And you've done it in a week. Yeah. And so everyone else can do it in a week. And in fact, if that's anyone's business right now, there's going to be a thousand new ATS systems a day. One's going to trend on Twitter, people are going to download that one, then they're going to jump ship because a new one emerges and it's better, it's more genetic or whatever. And so again, I think, are we seeing software generally become a commodity? What we're going to see is we're going to see all business opportunities be like YouTube content. I mean, if I went back 25, 30 years ago and said there's going to be 10 person unfunded little TV studios that produce content every single week for almost no money, you'd think that's crazy. I mean, Seinfeld used to cost $5 million an episode. The Simpsons and Friends, these were all multi-million dollars per episode. And we've now replicated that that can be done by a tiny little team for free every week. It's interesting because the minute we say that the answer is to do these things that are now easy and free and cheap to do, it's almost like we then don't acknowledge the fact that if everyone's it's no longer a valuable thing to do. It's less valuable. Yeah. As it becomes easier, the other axis is the decline in the value. So we're saying to people, go be a content creator, but the very fact that they can without needing a CNN-sized skyscraper production equipment also means everybody can, which means that that actually is losing value at the same rate that it's becoming easy, theoretically. And so what I continually think about from First Principles is, what is it that I can bet on that is irreplaceably human, which only I can do and a kid in Mumbai cannot do? And so when we say software, I go, kid in Mumbai. And then we say content, I go, kid in Mumbai. Well, physical experiences in the real world. I agree with you. So standing on a stage and telling your story is something, I mean, you already see that thousands of people show up to see this and to be part of, to be in a room with other people who are sharing an experience. So the thing that the AI version of you can't do is turn up to an event. I agree. So I say what I say because I genuinely want to be challenged here. I sit here with these AI experts and CEOs because I genuinely, I hope that they say something that tilts the current belief that I have and like makes me go, oh, actually, you know, that's a good point. I'm in search of good points on the subject matter. So when I said what I just said about content software becoming increasingly commoditised, and because it's becoming commoditised, we're like tempting people to go do more of it without acknowledging the fact that the very fact that people are doing more of it means it's less valuable and it's more of a commodity. And what I'm saying, though, is that these things have to go together. So simply like anyone who's doing real world experiences, I think what's going to happen is you'll go to a conference and that conference will come with custom software that lasts two or three weeks that is really, really interesting. And it relates to that conference. And that software was created as a SaaS product for the attendees of that conference. And it kind of pops into existence very briefly. These things used to be separate and now they all cluster together, right? So it's a buffet of things that you all put together. If we take most successful businesses now, they do social media. That used to be a standalone business. It was crazy to think that a hair salon could also be a media business. I've met with tens of thousands of entrepreneurs. And the number of people who want to go big is tiny. It's a fraction of the entrepreneurial community. What most people want is a business that provides an amazing lifestyle. They want to live and work from anywhere. They want to be able to travel. They want to do three or four days a week. They want to drop their kids off at school. They want to do non-monotonous, interesting work that is a little chaotic, a little creative, a little bit fun, a little bit challenging. They want to work with a small group of people that they get to know and enjoy and travel through life together. And those are the types of things that the majority of people actually want. Very few people actually want big. So one thing that I'm noticing is that it's probably harder than ever to build a big business, but it's easier than ever to build a small, successful business. So if we were to say that a business that does 1 to 5 million of revenue with a team of 10 people, that is totally super possible. But once you go beyond that, it's becoming harder and harder because you get disrupted more frequently. So you have to be an elite athlete of entrepreneurs to withstand the constant disruption and the constant change that's coming. So because your natural mindset is to think really, really big, you're probably missing that for many people who are working in an annoying corporate environment and their job is monotonous and they feel tethered to a desk or they feel tethered to a team that doesn't really care about them. A lot of those people are going to actually end up in small, dynamic little businesses that will never be big, but they'll be great. They'll be fun. And people will have flexibility and they'll have great opportunities as a result. I think you're right. I do have a bias towards trying to find big opportunities. But actually, I think the big blue ocean opportunity, which is, for anyone that doesn't know, I think of Red Oceans from the book called Blue Ocean Strategy as very competitive. Hyper competitive. Horrible markets where it's like a race to the bottom. You're all doing the same thing, like you're saying the same thing, doing the same thing. And you all think that your unique positioning is unique, but it's actually not. The market doesn't value it any differently. Like a wine company saying that their wine is a year older or using a fancier word on the label. It's a red ocean, a blood bath. A blue ocean is an uncontested waters where you have room to grow, much less competition. And there is actually something fundamental to your proposition, your business idea, the value you're giving to the world that is unique. I'm looking for blue oceans. And yes, I will try and build a big boat in the blue ocean because that's my nature. But also, life is easier when you sail in blue oceans. For everybody, even if you want a lifestyle business, I think one of the great examples at the moment is if you started a marketing agency when I started one in 2014, especially a social media agency, you were in the blue ocean. You were? Briefly. And they say, you know, blue oceans often last for 10 years. In 2024, my best friends, I'd say 70% of them run marketing or brand agencies. I'm having the same therapy session with them once a month at the moment. And if you look at the big five, big four, big five agencies in the world, the WPPs, Martin Sorrell's business, they're all… They're all competing to weigh their margins. Struggling. They're struggling, yeah. They're losing their profit. They're losing their market share. And AI and digital technology means that this disruption cycle happens in months, not years anymore. So you said a blue ocean can last for years or decades, not in the world of AI. In the world of AI, as soon as we've identified that you've got a blue ocean, we can replicate that real fast. And we can come and fly out to your blue ocean with lots of competition. Here's something that's interesting. Throughout all of history, whenever you find societies that have low social mobility and hyper competition, there's lots and lots of festivals. So you go into medieval times, they've got a festival for every season. They've got a festival for every life event. There's just constant festivals. You go into rural Africa, rural India, and it's festivals, festivals, festivals, right? Weddings, funerals, seasonal festivals. It's a huge part of human culture that we have lost in big cities. Like we don't do a lot of festivals. It's rare that we do a lot of festivals. These real world experiences. But I don't necessarily think it's a blue ocean because as soon as people realise how much people want to do festivals and want to do real world experiences, a lot of companies can also add this to the mix. I think what is defensible is community experience and embracing the constant chaotic change and coming up with new stuff all the time. One of the bets I'm making, and it goes back to what, again, reasoning up from first principles. If the question is, what can I do that an AI agent or AI could not do? And especially in the context of what humans will continue to need. When you were born, Dan, you were born with a bunch of needs. And those needs like fundamentally haven't changed. One of them is like connection. I think you'll always be able to discern what is real human connection from artificial connection. I remember being 16 years old and in my psychology class at the back, when they told me about the rhesus monkeys experiment, where they got a rhesus monkey, this little small little monkey. They pretended that its mother was a wire mesh. So they made a wire mesh mother. The rhesus monkey grows up to be a psychopath. Because its fundamental human need of connection is not met. They then did it with a cloth and the rhesus monkey grew up to be a bit more stable. And then obviously they showed the examples where the rhesus monkey actually had a physical flesh mother and it was fine. So I say this to illuminate the fact that we have fundamental needs, that if not met, there will be consequences. One of ours is connection. So as a content creator, I can sit here and I can tell you what happened today, right? I can say at six o'clock, this guy broke into this bank and then he left. I think that's a commodity now. I think the thing that I can do as a podcaster, as a behind the scenes content creator, as a streamer, this is why I'm so bullish on streamers, is they're making irreplaceably human content. There's no AI that can... An AI can tell you what happens in menopause. An AI can't tell you its experience of menopause. Yes, we need to find something that only we can say as humans. And every single person has this, by the way. But a lot of us overlook what we can say as a human. I recently came across a financial planner and his name is Matt Pitcher. And he gave a TED Talk. And the TED Talk is about what it was like for him to meet 100 people who won the lottery. So he had a partnership as a financial planner with the British lottery. And every week someone wins the lottery. And then he goes in and meets them a week later after they've discovered they've won the lottery. But what he did is he looked back at his history and he had lots of different things that he was doing. But he looked back at his history and he connected some dots. And he said, hey, wait a second, there's something that only I can talk about. There's something that is my thing that I've experienced, that I've lived through, that I've sat in those living rooms. I've sat eye to eye with those people. And only I can talk about that. And because of that TED Talk, he's had half a million views in the first few weeks. And his business is obviously going to explode and all of those kind of things because he found something that only he could say. I believe that every person has something that's human that only they can say that's highly relatable. You shared with me an interesting philosophy when we first met. I did my first episode on Diary of a CEO and it got millions of views. And I said to you, why is it that my episode got millions of views? But a few weeks ago, there was a billionaire CEO who only got a few hundred thousand views. And a few weeks before that, there was a billionaire CEO who only got a few hundred thousand views. And you said, Daniel, relatable beats impressive. And a lot of people think that they have to cure cancer or launch a rocket to Mars or that they have to win a Nobel Peace Prize or float a company on the NASDAQ. And actually, each individual person, we have to discover what I would call your personal intellectual property or your personal intellectual capital. And your personal story, your personal playbooks, you know, your triumphs and your disasters, the things that you've done that are interesting to a certain group of people that you can quantify, that you can describe, that you can give me what it was like. Those are your personal playbooks. Now, once you have personal playbooks, you have the ability to turn that into all sorts of products and services through AI. But it starts with having personal playbooks.