Why vibe coding is a slot machine (and what to do instead) | Eric Ries @ Lean Startup Co.

Foreign. Hey there. Welcome to another episode of Sauce Unbound. I'm Anna, and with me here is Eric Reese. I feel like you don't need much in an introduction. I mean, as I told you before, I live in Thailand, and I mean, I've been around Southeast Asia for a while. Every time I go to a bookstore, if they have English books, there will be Lean Startup. Wow. I'm like, okay, everyone knows it. That's great. I love to see that. Yeah. Well, welcome to the show. Thanks for having me. So, you know, you wrote a new book. It's called Incorruptible. So we're going to talk a little bit about that. But, you know, because I knew you were coming to the podcast, I thought, okay, I want to do my research. I want to see, like, how people reacted to Lean Startup. And I mean, I have quite a few people here say, yeah, okay, like, 10 years ago, we actually based our entire startup, our entire product on the Lean Startup principles. And just overall kind of on the Internet, people are very happy. I was very interested to find out why people were unhappy and what was missing for them. And the only thing was basically that there was not enough hand holding. So you were talking about building a product, you were talking about fast iterations. Talk to your customers, learn, go back. And people were like, okay, what do we do next? What happens when we are not the creative genius anymore? We're not just grinding. We actually have a business to run. So is that something that Incorruptible addresses? Oh, no. Thank you for asking the question. And it's true because. Because Lean Startup is so popular today, people have a hard time remembering how unpopular it was at the beginning. It took a while for it to catch on. It wasn't an overnight success. And yes, that complaint was. Now it's relatively rare. But, like, it used to be a one of several that were, like, really, people were quite obnoxious about it. And I understand, like, I came to be very sympathetic about it because part of the reason Lean Startup is so successful is that it's short. You know, it's a relatively easy read. It is. I tried to write it in such a way that it would. Would have endurance. So I didn't. I didn't write about Trello or about, you know, certain, like, the Agile tools that were popular at that time. I didn't write about. I actually took out hundreds of pages of, like, technical details about split testing and unit testing and all this stuff that, like, I, as an engineer really found that stuff really interesting. But I wanted to write a book that was like, really from first principles, people could apply the ideas to their own context, translate them into their own industry. Now, this has been so much more successful than I ever imagined because the ideas have been used in big companies and small companies and nonprofits and governments and every kind of combination of company, industry, sector you can name. People have tried it out. But I think the reason it has had that staying power is the same reason why some readers do find it frustrating. Because if you're looking for more prescriptive guidance on first do this, then do this, then do this, and especially if you want it tailored to a specific stage or function. So, so now, I mean, where now there's like a whole industry of books that are like, how to apply Lean Startup to user experience, how to apply it in an enterprise, how to. Like, there's two different books called Lean Enterprise. There's, you know, there's like, there's Lean UX and then there's UX for Lean Startups. There's like a lot of books in this, in this series. I helped O'Reilly publish, publish a bunch of them, and then a bunch of just have been published on their own. So I'm, I'm pleased that the information is out there for those that want it. But yes, it is true that people find it a little bit frustrating. But on the other hand, I really felt that we can't claim that Lean Startup is a scientific theory if we then ask people to take it on faith. I was like, that's actually contradictory to our purpose. What we need to do is give people enough tools to test for themselves if the ideas make sense and then adapt them to their local conditions. And that's also been a really important part of its success. Like, if you go and read the kind of. You know, I'm always so flattered that so many entrepreneurs all over the world have written these testimonials and they write about, as you say, like, that the. Com that the book really helped them. That, to me, is the ultimate test of a book like this is not. Does it sell copies? Celebrity memoirs sell plenty of copies. That doesn't mean there's anything between the covers at all. And frankly, most business books are never read. They're. They exist to be put on the shelf to display it as like a, you know, like a logo or, you know, like a, like an icon. So, you know, the fact that people have actually found it useful to me is really valuable. I think the reason you have seen such a variety of people make the ideas their Own is precisely because it has that first principles thinking. Now, you're right, though, that the book is not complete. There are topics that are not covered, or covered only inadequately there. So, for example, you know, I've written a lot since the Lean Startup came out about how to apply the ideas as organizations scale. I wrote a whole book called the Startup Way, which corporate innovation people know me for the red book. That book is red. Lean Startup is blue. So there's like the startup people all know me for the blue book, but the corporate people all know me for the red book. And I even wrote another, a book called the Leader's Guide, which is again, about how to apply these ideas in a highly scalable way. So that was one kind of omission, but there's a more subtle one, and that's what the new book is really addressed itself to. And I don't know if you've ever seen on Reddit the meme with Anakin and Padme. And Anakin says, I'm gonna change the world. And Padme says, for the better, right? And then he gives her, like, an evil grin. Oh, yes. And then she's like, for the better, right? Okay. So I used to think that was the funniest joke on the Internet. And then a couple years ago, it stopped being funny to me because we've seen all these, especially technology companies who seem to have forgotten to specify for the better and so that they changed the world. But a lot of the people who made those changes are rich and miserable because it didn't actually live up to the promise of what they should do. And I kind of took it for granted that when we were teaching entrepreneurs to change the world, that obviously we all met for the better. Right? And then I had this really sick realization. Oh, this is a little while ago. And I was really worried. I said, wait a minute, did I write in the Lean Startup that entrepreneurs should change the world? I'm like, I'm pretty sure I said that, you know, like a throwaway line. So I was like, better go check if I specified for the better. And I went and I found my old copy. I don't read that book very often. I, like, pulled it off my shelf and I'm looking through it in the introduction at the very end, it says such and such a thing, Lean startups, so that the next generation of entrepreneurs will have the tools they need to change the world, period. And I was like, oh, I gotta hang my head in shame, because I just thought it was so. I couldn't even imagine what other purpose could it be? But now I see how that's been a pretty major oversight. And what's interesting is people hear that and they're like, oh yeah, Kumbaya, you know, making the world a better place, whatever, like, I just want to make money or whatever. So like, I get that not everybody, like, is as idealistic as, you know, as others, but even among people who are not especially mission driven or social impact oriented, this inability to specify what the change is for has actually destroyed a lot of companies and prevented them from being successful. So for a lot of founders, the, the, this failure, it was really my failure to prepare people for this has caused a lot of problems since. And I, way I put it in the new book, is that I taught people how to build something worth protecting, but not how to protect it. And the sad reality of our world today is that everything that needs protection, like everything that is worth protecting, will eventually need protecting. So we have to learn urgently, I think, as founders, as builders, as product managers, as leaders, how do we encompass encode these new things that we make into a structure strong enough to keep them safe? Okay, so just again, kind of going back to overlap between lean startup and now incorruptible, do you think it's even possible or need it, let's say for, you know, a very early stage team to have a mission to kind of find out why, why they're doing this and you know, double down on culture and governance when, when, let's be frank, there's maybe not much to govern yet. Right, right, exactly. So that's one way of putting it. And, and that's what you've just described. Is debt by far the dominant view. If you ask almost any lawyer, investor, investment banker, like any expert, should I worry about that stuff? They'll always tell you the same thing. Oh, it's too early to worry about that. After you have product market fit, after you have this, after you have that, it'll be the right time. And what's funny is it's never ever the right time. It's always too early. And then one day it's too late. So like, you raise a Series A, is this a good time? No, no, no, do it. When you have more leverage, you raise a growth round. I joke in the book that you're the bold contrarians on your board are like, you know, you don't want to be too different from other people because you might not be able to raise money, you know, later rounds, just deal with it later. You're doing IPO prep And now it's like, oh, well, actually, investors might not like it. It just like, it's never the right time. And then one day you wake up and it's too late. You try to make the change, and you can no longer do it. So my claim in the book is that founders should just set this stuff up right from the beginning, and it's not that difficult. And in fact, it's easier the smaller you are. So the very fact that there's not much to govern makes setting up the governance extremely easy. Set the precedent, set the structure, set the culture, intention and values. Set that stuff up at the beginning. The earlier you do it, the easier. Now, that doesn't mean that if you missed, you know, if you started your company last week, it's already too late. You should have done it last week. It's all. There's always something you can do. It's never too late. But I think in general, doing these things earlier than later is. Is really important. And I'll give you one very concrete example. Yes, please. Because this, I think, is where a lot of founders go wrong. I was meeting with a founder. He actually was already very successful. He had a couple of successful companies that had exited starting a new company. And he was very. How do I put this? He was very gruff in his effect. Okay. So he was like, I don't care about ESG nonsense. I'm not woke. You know, I don't want to hear about st. I hate the word stakeholder. It's like, just. I just trying to make money for my investors. We make, you know, I know that more money we make for my investors, the more strength we'll have to have the leverage to do with the way we want to do. And I was like, oh, well, what is it you want to do? He's like, I want to build a company. He was an engineer. That is the greatest company that engineers have ever worked at. So engineering quality is the only thing we care about. And you're just going on and on and on about his engineers and how great this company was going to be for them. I was like, fun. I was like, oh, so you really love your employees? Like, I would do anything for my employees. But you don't like the word stakeholder. He's like, absolutely not. Okay, cool. And then he said to me something was really interesting. He's like, I don't need a corporate purpose or a mission or any of that stuff, because I just have this very simple strategy to make these great products. And so I asked him this question. I said, well, what if one of your engineers came to you one day after you're successful and said, hey, we've got this product. We make X dollars per user. If we kind of skimped on the quality a little bit, we could make an extra 10%. It would cost us 10% less. We could make more money. Would you do it? He's like, hell no. What are you talking about? No. Quality is the only and most important thing. I said, but I thought you said you wanted to make money by any means necessary. He's like, well, not like that. Oh, what's that? Not like that. What does that mean? And so then like, we started to unpack it and he, he didn't realize how radical his own vision was. He didn't like, he didn't talk about having a mission, vision, stakeholders. He didn't care about any of that, like, fancy sounding language. But he did inside carry what I call the builder's intuition, which is that the best way to make money is by creating something that people love, making their lives better, and then capturing some of that surplus value that you create for them, for yourself. And he didn't realize. Very few founders realize how diametrically opposed our financial system is to the simple idea, and therefore what a revolutionary you are if you say something out loud. As simple as, I think quality should be our North Star. Because as soon as you say that quality is your North Star, you're explicitly saying that shareholder value is no longer your North Star. And that dissonance between the mission that companies claim to have and the legally mandated shareholder primacy, it's called mission that our financial system requires them to have. That dissonance destroys a lot of great companies. All right, so, you know, so we're not blamed for, for not being handholdy enough on this podcast. Sure, sure. What would be, you know, what would be the first steps that you would recommend any founder do if they even have an idea of building something? Yeah, so what, what the. There's, there's kind of like broadly speaking, two halves to this equation. There's what I call the ethos of the organization, its character. I try not to use the word culture because a lot of people don't even know people agree on what culture means. So I use the word ethos, the character. What does the organization do when nobody's watching? The pioneering management theorist Mary Parker Follett a century ago called it the invisible leader. Not the person whose name is on the business card, but the sense of, of common purpose that people follow when there's no one else around. You know, if you have the chance to defraud a customer, will you, if you can get away with screwing them over, will you, if you could lobby the government for an extra, will you? And like most organizations are presented with thousands upon thousands of these temptations every day. And frankly, if we're in bigger organizations, I'm sure many of your listeners have had this experience. If you are one of the people, I call them the torch bearers. Someone who really cares about the product you're building, wants it to be good, wants to take care of customer. Like, have you really internalized the company's mission? Life sucks because every day there's some schmuck with a, with a spreadsheet coming into your office being like, you know what we should do? Like, we should make the product a little more addictive. We should load it up with 12 more ads. We should like, you know, what Cory Doctorow calls and shidification, if you don't mind my language, like, let's make it a little bit worse so we can make more money. And it's exhausting to be like, thank you for your input, but no, I don't think we should do that. Let's do the right thing. And that person's like, well, what's the ROI of doing the right thing? It's like, man, we don't do the right thing for roi. We get tremendous ROI out of our product because we do the right thing. So what we have to do is develop an operating system, a set of management practices and cultural practices that build something worth protecting. And then we have to do the more structural things to protect that pre existing ethos. I, I say it in this order very intentionally because a lot we started talking about corporate structure, people start turning to be like, oh, well, you're saying that if I, you know, if I just, I put this armor on, then I'm automatically a virtuous person. No, the armor can't do anything other than protect a pre existing being. If you, if you build something that sucks, making it permanent just makes it suck more. That's not, I'm not, I'm not into that and that's not helpful. It doesn't do anything. But if you've really, if you've thought about with your co founders, you're like, okay, we really have some kind of mission that we're excited about. We have a set of values that we're prepared to defend. Then there are some very simple structural things you can do. The most simple one is called what Being called what's a public benefit Corporation? A PBC conversion in 44 US states. You can do this in a lot of jurisdictions around the world they have an equivalent filing and it's like super easy. In Delaware it is a two page legal filing in which you establish at the level of your corporate charter this is the purpose of this corporation, this is what it's for, the public benefit. So it can be like very simple. Like if you look at Anthropic's Public Benefit Corp statement, it's like to create safe and beneficial AI for the benefit of all humanity. Some of them are a lot simpler than that. Just to build a great product, just to be high quality, just to serve a certain community. It could be anything, anything that you truly believe in. And what's powerful about that statement is by writing it into the corporate charter, you protect against a lot of situations in our modern world which we can talk about if you're interested, where you might be forced to compromise that mission for the sake of, of shareholder value. So that's like one very simple thing. I view it as an absolute no brainer. I'm shocked that not every founder does this. And if you do, at the beginning it's no, there's no investors to consent to it. Not you literally do nothing. It is a simple filing you do, it takes a day. Very easy. But for founders in particular, since you asked, I also would put other protections in place. So for those that don't know, there's something called founders preferred shares. I'm not going to get into the legal wonery of this, but that's a very valuable form of protection. You can then, once you've established a separate share share class for founders, you can attach additional voting rights, board rights, all kinds of goodies can be attached to that. And it's not that I think that founders should be megalomaniac dictators for life because believe me, I know a few of those people and they're totally miserable, okay? Like it's very, very, very unpleasant. What we want to do is view founder control as a temporary bridge to a more permanent structure. I call it in the book the Architecture of Institutional Longevity, where we start to invest this power not in people, but in structures so that we can have succession, continuity, preservation of values over long periods of time. And I tell a lot of stories in the book of companies that have mastered this architecture. That's really important. But I know a lot of founders are like, I'm just trying to get through product market fit. I'm just trying to get through next week. You know, I don't want to hear about living to 50, but the decisions you make at the beginning. I know so many founders that like their number one regret in life, on their deathbed even. It's like, I wish I had set this up differently from the start. So do your future self a favor and take this seriously. Yeah, it's, it's interesting that you started talking about culture and I mean, I'll call it that because we use that word. Yeah. SaaS Group is an acquirer of SaaS companies. And interestingly enough, culture is something that we look at very closely when we acquire a company. Sometimes you have to. Yeah, sometimes when I have a call with founders, my friend calls it an asshole check. So basically, if we don't share values, if I cannot hold onto a conversation with you for half an hour, then after an acquisition, when you stay with us for a year, that's going to be a nightmare for pretty much everybody involved. Right. We have 25 companies, 400 people. Even if the company is doing so great and everything's looking up, we don't want that company. But at the same time, and we are recording this on February 5th, yesterday, I believe PayPal faced this horrible downturn and I think their stock dropped like 25%. And there was a quote from their former CEO David Marcus, who argued that PayPal's decline was a direct result of a decade long shift from product innovation to financial optimization. And that's basically what you're talking about. But at the same time, I think it's so difficult to not fall for that because let's face it, everywhere you look, everyone's trying to sell you dashboards, more data. Look at this metric and look at that metric and you're not tracking that. Oh my God, what are you doing with your business? So I feel like it's a very natural course for people who don't have that mission in their heads. So how to maybe, maybe you can unveil something that you're writing your book, or maybe something just your opinion on how to look at metrics but still be value driven, not numbers driven. Yeah, that is a really, really important issue. In the academic literature. This is called surrogation, where the metrics become a surrogate for the thing itself. So you're supposed to be focusing on delighting your customers so that you have high customer retention, so that you make money so that your shareholders benefit. But it's so easy to be like, let's skip the delighting customers part and just look at Retention. And now let's make the product addictive or add vendor lock in or whatever. Oh, look, our metrics go up. Oh, you know what, Skip, let's just forget that. Let's just focus on the share price. And if the share price is going up, presumably our customers are happy. And what happens is you can make any metric you want go up for a certain amount of time by cannibalizing your most precious resource, which is your trustworthiness. So that pattern that you were talking about from, from David Marcus is so common where you, you juice the stock. It's like, you know, it's like taking steroids or, you know, think about whatever your favorite. Like, you know, like if you have a metabolic disorder or something in your body where you're like growing out of control, you can do that for a while and then eventually the reckoning comes. And the problem with our increasingly short term financial markets is oftentimes by the time the reckoning comes, the person who made the decision has already exited with their stock options and their fortune. You know, I tell a lot of stories in the book about CEOs who drove their company to the ground and walked away with $100 million or more in compensation. So, you know, it's rational what they're doing. We've built an incentive system that, that rewards that surrogation. So it's actually because missions are so powerful and compelling that they're the only force powerful enough to pull away from surrogation. So this is, this is the, really, the key to resisting this pressure is you can't tell people, don't look at the stock price. You can't tell people, don't look at the numbers. You have to give them something to look at. I say like, this is our goal. This is what we actually really care about. And there's this really funny phenomenon that I've learned being around a lot of mission driven leaders over the years. They, they all have a reputation for being a little bit of a pain in the ass because they have this, this principle that they care about. You know, think about Steve Jobs and his design aesthetic. I tell the story in the book of Jim Senegal, the founder of Costco, and the time that, you know, his COO decided that they should raise the price of the $50 hot dog they sell outside of every Costco. And he was like, I will effing kill you if you try to raise his price. Like he made that guy, he made his life so difficult. He's like, no, we have to find a way to do it. Profitably at this price point, it's just like so much extra work. You know, there's. Everybody knows Patagonia today as a company that's like really into environmental causes. But the environmental stuff came later. Their original obsession was about product quality. And Yvonne Chouinard, the founder of Patagonia, was just absolute nut about quality. He would teach these classes to employees or he would sit down for a quality product quality class and he would like take up. He would pick a random product, like he'd bring a bag of, of Pete's coffee beans and he'd stick them down on the table and says, is this a high quality product? And everyone would look at him like, sir, it depends. Do you like coffee? I guess. And they'd be like, if I said, if you're a coffee drinker and you like the taste. And he would be like, no, quality is an objective quality. So it is objective. It either is or is not quality. Every product is like truly fit for its purpose or not. And give it. They would do like they would do. They would debate this for hours and hours and hours. And yeah, that obsession with quality created all these problems. It was like a constant nightmare that nothing was ever good enough. There's this famous story about a time that a retailer wanted to carry their product and sell it to their members at a 3% discount. 3% discount made this huge bulk order. Patagonia would have gotten the same amount of money per unit. Okay, they were just going to sell it at the end, user is going to get a discount, the retailer is going to take a slightly lower cut. And they went crazy being like, no, no deal. And like, blew up their whole year. Because he was like, no, discounting is not. Doesn't signal quality. It signals something anyway. Just like each of these stories, you hear these stories and you're like, God, what a pain. But what people miss is that the pain of this principle is actually like, these leaders love it. They welcome the difficulty of it. They like that the ethos requires constant challenge from reality. Because every time we do that, we have an opportunity to teach people what the principle really is, what we're really doing here. And that's where the breakthrough solutions come from. There's so many stories in the book about times when a leader encountered this kind of difficulty. Insisted, as Jim Senegal did, that they effing figure it out. And they do. They come up with some breakthrough that's better than either of the alternatives that they were previously considering. And it's only companies that have those principles deeply embedded. They can achieve those breakthroughs because everybody else just compromises at the first sign of difficulty. Give you one, one other example. I saw this. This isn't a. I was watching a video. I forget whose video this was. They're telling this story about going out to dinner with a friend at the friend's favorite restaurant. And they get. Sit down and they kind of. Friend's not looking very comfortable and the service is a little bit rude and whatever. And he sits down there, start the meal, and they bring him the food. He takes a bite of his food and he's like, hold on a second. I look something up on my phone, guys, like, you're being really rude. You know, like, he's just. One second. I just need to know if this restaurant was bought by private equity. And he's like, Googles it. And he's like, yep. Shows it to his phone. He's like, I could taste it. I could taste that it was owned by private equity. So, like, that's the, that's the, the like, inevitable consequence of surrogation. And it's really like, it's, it's. We have to start to see it as genuinely sad. Like, we have plenty of re. Academic research and research out the wazoo that shows that this is value destroying in the long run, but we also have to learn to see it as a, As a tragedy. For this was somebody's dream to make this thing really beautiful. And when it, when it loses that essence, it's a major loss. Yeah. There were quite a few stories here on the podcast, to be honest, where founders admitted that, yeah, okay, maybe taking VC money and, you know, going away from, from their original idea and original mission and into this optimized world was a bit of a mistake, and they were either on a way to buy out the VCs or. Yeah, yeah, it's very challenging to do that once you go down that road. But I wanted to ask you, Obviously, you know, I tell every time. I don't remember the last time I had a podcast which did not include anything about AI. So I'm just going to throw it there because. My pleasure. Yeah, happy to talk about it. Yeah. Again, when I was doing my research about you and what you're writing about and how people still use Lean Startup as a baseline for what they're building, there were, of course, people who were talking about using it for their AI creations. And on some level I thought, okay, people will probably say, okay, this is no longer kind of viable because, okay, we're talking about something minimal that people would Just take a look at, just for a bit of a visual in a way to understand if that's something they can sign up for. And in the age of AI, visual is a very different thing and personalization is a very different thing. So maybe people will be complaining about it yet again. But it turned out people were like, okay, we're still using the fundamental principles, right? You have to talk to customers, you have to give them something. It's that with AI, it's so much faster that instead of, hey, we talk to a customer and then we end up working on it for a couple of months and then we talk again. Oh yeah, you can do it in a matter of days. Just like, okay, so much better. Does this work? Does this work? Does this work? And then instead of wasting time, you're just, you know, you're, you have a much better leverage. So what's your take on this and how, you know, your idea of experimentation and learning you think has a life in AI startups? This episode is sponsored by Rewardful. Looking for new ways to find customers for your SaaS business? Consider building an affiliate program. Rewardful is the easiest affiliate tracking platform to set up, manage and scale. For SaaS companies, building a successful affiliate program can be a little bit intimidating at first. First. And that's why Rewardful has taken what they've observed from their most successful customers, affiliate programs, and distilled that into an exclusive online course. The exciting part, their affiliate marketing course is absolutely free. Start the course at academy.rewardful.com academy.rewardful.com and turn your biggest fans into your best marketers. Oh yeah, I see it. I mean, it's more important than ever for that. You already laid it out very beautifully. I will say though, that there has been a bit, because we're having such a mania about AI right now. Like there's a bit of a back to the future feeling where like the number of companies that have demos but no products like that was like completely eradicated in software for a time. It used to be like all an epidemic. We used to have a lot of it. People raise money on flashy demos and then the product, they would never ship a product ever or the product would never do anything anyway. Like, you know, sas, we, we got rid of that for a long time and now it's like made this like stunning, stunning comeback. And you have companies that like, they release models but no products anyway. It's, it's, I think an unfortunate reality of the fact that people are just, you know, when there's a Mania, everyone's chasing, chasing these returns. I will say the other thing. So you're absolutely right. The, the, the sweet spot for lean startup has always been those industries that combine a relative advantage in cycle time and uncertainty. So if you think about like a two by two matrix consulting style, there's like uncertainty, high or low cycle time, fast or slow. If you have fast cycle time and high uncertainty, then that's always been lean startup sweet spot. But of course fast and slow are all relative. So. Relative to what? Relative to the other competitors in your space. So whenever a technology disruption happens, generally, generally both quadrants, both axes activate at the same time, the industry gets way more uncertain because now you have like all this crazy stuff. There's a bunch of industries that have either disappeared overnight or they're being completely reconfigured right now as we speak. So the uncertainty goes up and you have this, you have new entrants with a ridiculous cycle time advantage. So if people can build in a day what used to take a month or a year, they're going to have a huge, huge advantage. And people see this of course all the time in AI. But I remember, I remember when, when this happened in appliances. I'll tell you a funny story. This is from a few years ago. I was talking to a, I happened to talk to several American manufacturers in a row just you know, because I was talking to companies back in the startup way days and they were all saying, they're all complaining about the same thing because like in these big, big time kitchen appliances, the average cycle time was like a new, a new platform they called it, every 7ish years and then like a new like model update like every two or three years. So like basically like basically on average like a new thing every five years. And Samsung had entered the American market for appliances in the most like bizarre way. The American companies all found it super bizarre because Samsung would release like four versions a year. So they'd be like, they'd be just cranking out refrigerators left and right. And their early entrants were all terrible products, really low end, cheap, crappy stuff. Like people were, the Americans were very dismissive and they had like, every year they would be coming out with these like wacky features that didn't make sense. Like they had like a refrigerator with like an Internet browser on the water dispenser and it's just like, like they were just trying, it was throwing stuff against the wallets and seeing what sticks. They're like, this seems nuts but like after five years of Four products a year. Products were really good. Like they really were figuring out what customers wanted. They were having this like massive cycle time advantage. And the American companies couldn't understand it. So by the time Samsung started eating their lunch, it was way too late. Now part of the story is because Samsung comes from semiconductors and electronics, which are like super slim margin products. And so when they enter, when they entered like luxury appliances, they're like, these margins are amazing. We could, we could sell for half the price of our American counterparts and still make way more money than we make on, on the stuff we're used to mass manufacturing. So of course that was part of their advantage, but the other part of the advantage they had this, this intrinsic cycle time superiority. So that I think whenever you find yourself in a situation where you're on the wrong end of that equation, you're in big trouble. And conversely, as an entrepreneur you should always be seeking such opportunities out. So yeah, so, so AI creates lots of those opportunities right now. But I will give a word of warning. You know, I helped start an AI research lab called Answer AI and I've been around and involved in a bunch of AI companies. And I do think that one of the things that people are getting wrong, especially with all this agent talk, is the question of like, are is AI fundamentally a replacement for human creativity or an amplifier of human creativity? A lot of these products, the vibe coding products especially, they are written in such a way as if the purpose of work is to produce the artifact. So everyone's like super proud of themselves that they can generate a hundred thousand lines of code in a day, as if the code is the value. And talk about back to the future again, especially in a startup situation, the artifact is not what's valuable. What's value is the learning, the human learning about the product and what works and what doesn't work. And so a lot of people are skipping that stuff. And so not only are they not learning, they're also vibe coding products that are not any good and they don't even know. So there's like a very, very important study by meter where they did this large scale study of, you know, the study, it was a large scale of open source developers where they asked them to flip a coin. Nope. Okay, everyone flip a coin and the next time you have to do a feature add to your open source project, flip a coin. If it comes up heads, use AI, comes up tails, don't use AI and report what happened. It was like a beautiful, you know, actual double blind, a single blind experiment. And so they did that, and they had everyone report back, and everyone reported back their productivity. They reported, on average, the people that used AI were 20% more productive than they were without the AI. But here's the cool part of the study. Meter also measured their actual productivity both with and without the AI and found that on average, they were 19% less productive with the AI. That makes sense. So the self, the self Reporting productivity was 40 percentage points off the real productivity. Some people are calling this LLM psychosis. I like the phrase dark flow from the gambling addiction literature that you start to play, the AI becomes like a slot machine that you're playing and it starts to be fun. You feel like you're making all this progress, but you're actually not. You're just creating slope. So I think for entrepreneurs, this is super extra super double dangerous. And you need to make sure that you're using AI tools that amplify your own creativity and maximize your learning. Obviously, we make such products@answer AI so I'm talking my own book here. But whether you use our products or somebody else's, like, I think you got to be extremely careful. You'll be focused on tools that have human in the loop, not, you know, not the, not the slot machine. Oh, yeah. Okay. Thank you. I was battling with, you know, with AI, with this whole idea of using AI for everything because you do something and immediately somebody asks, did you use AI for this? And when you say no, they're like, why? They're like, yeah, because it's 40 times faster if I just did it. Because I did it 40,000 times before. And I know exactly step by step process. And here AI would stop and I would have to go in, and then I would have to launch AI again, but then I would have to go in. So I'm, you know, sometimes I think it slows you down tremendously. And yes, it does create slops because you're just like, okay, AI will. Will know this and I will have an opinion, but it's, you know, not your opinion. So. Yeah, and you don't even know what you made. So. So then like, you can't even, you can't answer questions about it. I fall into this trap many times because I, you know, I got excited about AI too. I've used all these same tools, and I've used them badly many times. I, I was just working on a project where I was playing with Claude Cowork, which is great, great product, but like very slot machine, very dangerous. And I was like, make me the artifact. Just did It. And I like vaguely looked at it. I was like, it's pretty good, make it better, make it better. I just kept telling, make it better, make it better, make it better. And it was like, I was like, this is so fun. I'm being so productive. And anyway, I get to the end, now I have this massive artifact. You know, Claude's decided it's like the greatest artifact that's ever been, you know, ever been created in the history of mankind, if it does say so itself. Because I've been telling it to make it better according to itself. And I was like, God, it's too. I can't. It's so long, I can't even read it. So I was like, okay, I sent it to somebody else. What do you think of this? But like, they're like, this is too long, I'm not going to read this. And they're like, yeah, did you write it? I'm like, yeah, just, it's an AI thing. And they're like, okay, well, you know, maybe I'll get an AI to tell me if it's any good. Yeah, I'm like, like, okay. Anyways, like, no, this is. We've gone totally wrong. So then I was like, you know what, forget that. Delete, delete, delete. I'm just gonna write this thing myself. But I couldn't do it. It's like I already have like 10,000 words of this other thing over here and I don't know, like, so it actually, like it has taken a project that I could have easily made progress on with AI help and totally blocked me from making any progress. And that is like, I hear so many stories like that where it's like, people are not integrating AI into their normal workflow. The other error I see people making is people are extrapolating the productivity benefit of using AI by measuring how well AI does at one of the steps of a multi step workflow and be like, well, since this step is 10 times faster, the whole workflow will be 10 times faster. But quite often that step is not even a very important step. And in fact, in programming, like writing the code, it's not that important of a step. It's not, it's not an unimportant step. It's valuable. But like a lot of times the other things matter a lot more. So one of the things that's happening is we're having a big debate over the relative merits of software engineering, which we're going to have that same debate in every creative field, which is hitting software first, which Is like, is the, is there anybody actually designing this thing? And my experience with these tools is when given proper design and good oversight, just like a human employee would, they're much more productive. And especially because they, you know, they have jagged capabilities, just like a junior employee would. The more scaffolding and structure you give them, the more they're like you. You're basically like using architecture and design to help them overcome their own limitations. And therefore you get a lot of productivity benefit out of it. But that requires a lot of thinking and upfront work and creativity that people are thinking AI is going to get them out of doing. And I would make that bet. Yeah. All right, that's perfect. Honestly, that it confirms what I, what I'm thinking. That's why I'm so happy. But, but yeah, it's just, I remember, I mean, I also started like everyone else, right? We were told, hey, there is this magical saying that you just tell, do that and it does it. And you know, I remember how frustrated I was using ChatGPT and whatnot the first times because it was so sloppy and so generic and not good at all. And it was like, oh, okay, maybe it's not for me. Maybe, you know, I'm the problem. And it turned out I was right because I was like, do this. And it didn't know, it didn't have the context or anything. And now with systems it's absolutely so much easier. But yeah, be the human in the loop. I think that's the hot take. But I wanted to ask you because you mentioned that there will be torch bearers in the company and you need them, right? You do, you absolutely do. You have to still preserve this culture of experimentation and, you know, kind of healthy creativity. But if, you know, if I'm thinking about like new companies, new products, new teams, it's fairly easy with AI. You know, again, we are creative geniuses. AI just empowers us and blah, blah, blah. It's the bigger companies, it's the established companies that I think have the biggest problem with it because they have processes, they have step by steps, they have hierarchies, they, you know, they, they have their own established ventures where everything works the way it works. So what would you Recommend? You know, 20 to 50 people, company that really wants to start with AI. Like where to start, how to, how to convince the stakeholders. I also hate that word, sorry. Why it's become so, so meaningless. Yeah, yeah, it became just a buzz that this is something worth doing. And unless we allow people to, to, to, to have the Torch, then you know, our mode can be gone and you know, nobody wants. I'll give you the answer, the true, tried and true, like, like verified by science, absolutely correct answer that nobody wants to hear, which is that for a transformational technology like this, there's no substitute for actually learning how it works yourself. Nobody wants to do that. Everyone's like, just give me the, give me the 1, 2, 3, step by step, whatever. And it's like, if you don't understand the capabilities of this technology, you're going to have a really hard time. So I really would encourage people to actually take a minute and try to understand how deep learning works. I know for people who are non technical, this sounds very daunting, but it's not. This is not some like, alien technology, okay? But the way that it works is so counterintuitive. Most people even in the field really don't understand why it works. In fact, it just doesn't seem like it should work at all. Like, if you understand, if you can read, if you have enough technical knowledge to read the transformer paper, the original paper, attention is all you need, is the most accessible one. Like, and the good news is AI is a great close reading partner. So like take a thing like that or read a book, or you choose where you want to do it there. Obviously there's plenty of great courses. There's the, I would recommend the fast AI courses on deep learning are still great, but like, sit there with a, with Claude or with, you know, one of your tools and have it help you understand. Be like, ask it the dumb questions you're afraid to ask and have it walk you through the paper step by step and help you understand every element of it. I really think that is underrated because the second thing that's going on is we have a mania, right? A technological mania. People keep asking me, like, students especially, is, is AI like, is it an overhyped, you know, financial bubble about to explode? Or is it a transformational technology? And as the meme says, why not both? Like, there's no reason it can't be both. You think about the telecommunications bubble that was catastrophically bad, and yet telecommunications turned out to be pretty, pretty good. And all that fiber we weighed, that was totally like, we're using it all now. It's all turned out to be fine. So I think a lot of good will come out of the mania, but a lot of crazy stuff is happening too. So in a mania, not only are companies under all this pressure to adopt AI and therefore are doing just the stupidest things with it, they don't give them any. And then they like say it doesn't work because I'm not getting productivity benefit. I'm like, but who told you to do that? They're extremely vulnerable to having snake oil sold to them. Okay? Like these vendors are making ridiculous amounts of money selling garbage products that don't work to, to companies. I know, I know in one of the verticals where I help start a company in that vertical, there's a certain, I won't say, but there's a certain AI powered product that is like making crazy amounts of money with a product that like everyone who actually uses it for a living says does not work at all. But the people that use the product are the people that buy the product. And they've made a fortune convincing C suite people that if your people tell you the spot doesn't work, they're just Luddites who don't like AI, which is easy to do because actually in our world, of course there's all this pressure to use AI. If you say I used AI, if I didn't use AI, people would say to you, how come you didn't use AI? Well, actually in the real world where normal people live, they hate AI. AI has very negative perception because first of all, it's the same people that brought them social media and people hate social media now. But second of all, like the, the product was made by, by theft, by this, the, you know, a non consensual training on all this data, you know, every, every artistic work in the world, including mine. So like, I, I understand, I feel, I'm sympathetic and yet, like it or not, here it comes. So we really have to get past the snake oil nonsense phase of the hype cycle. And so for companies who are being barraged with this crap, it's very hard to tell who should I listen to, who should I trust? And the only solution is to go down a level and develop some technical knowledge so that you could start to see through these salesmen. What I don't understand, people are so reluctant to learn anything, but I don't understand why this is seen as like such an unreasonable ask. Because like, I feel like I don't have the time for this. But like, if you really believe your own rhetoric about how catastrophic it would be to miss the AI wave, and if your board is really telling you this is existential, all the thing, like if the stakes are really that high, like isn't it worth a couple of hours to just make sure you make the right decision? I just, I feel like people have thrown up their hands, like there's nothing they can do, when in fact the, the actual knowledge that is needed to understand this technology is readily available to anybody who wants it for $10 a month. I mean, it's just, it's so cheap. Why not, why not find out how it works? Yeah, I, I feel like it's also a problem. I mean, we face that too. Right? When this whole thing happened, we were just like, okay, now we're all using AI and we had no guidance. No, like, okay, here are the products to explore. Here is the time dedicated for it. And I think a lot of people, especially in bigger organizations, they honestly sometimes don't have those two hours because, hey, you're so busy. The stakeholders want you to learn it, but they don't give you the time to learn it. Yeah. So you're in that limbo, like, what the hell am I doing here? Like, I'll tell you. I'll tell you. A leader I really admire, actually, he's in Southeast Asia. He's, he solved that problem by taking a two week vacation. He just, he moved to San Francisco. He got an Airbnb for two weeks or a month? Like, for a couple weeks, just, it was like, forget it, I'm taking pto. He's like, I'll take it as an unpaid leave if you want. I'm just, I'm leaving. I'm going to say just spent all his time, he spent the mornings hacking on code, trying to learn all these new tools in the afternoons, like going to AI meetups and meeting AI companies and whatever. And he came back from that trip actually knowing what he was talking about. And the next time he had a, you know, he's a very senior executive. The next time one of his teams told him that's impossible, he was like, I have a demo of it. I did yesterday. What are you talking about, impossible? Like, he, he developed a visceral understanding of what the technology is capable of. So again, I get that people are too busy, but I'm always like, if you're too busy to do the thing that would make you less busy and like, what kind of leader are you? Who decides what, how do you, like, who's deciding how you use your time? And I, look, I understand there's junior employees out there who have a tyrannical boss and who have no, you know, I get that there are people who really, really, really do not control their time. I always ask them, are you sure you want to work at a place like that? But anyway, some people do fine. No problem. Maybe that's all you could do. This is not for you, but for people at higher levels in the hierarchy. Give me a break. You're in charge. You're in charge of your time. Be a grownup. Like, this is if you, this is the moment. Seize this moment. And the things that are going to like that you're going to take time away from are going to be more than offset by all the time you will save by being able to call BS on the thousands of hours of nonsense AI crap you have to sit through every day. I mean, we solved it a little bit easier. Like we actually allocated hours for learning that are within your. That's so smart work day. Because. Yeah. Then people feel like, okay, who I have. Like, nobody's going to ask me what I was doing here because I was learning. It's fine. It's great to have as a corporate value. Absolutely. But I wanted to also ask you, AI, you called in mania a couple of times and obviously there was a lot to figure out there. Is there something that you are excited about now, except for AI? Like, it can be technology, it can be just something that that's going on in, I don't know, SaaS or outside of AI? Yes. Yeah. There's so much great stuff going on. Yeah. I mean, the other problem is in the, in the tech press and in financings and in the VC imagination, like, AI has eclipsed everything. And listen, AI is super cool and I, I've been investing a lot of AI companies myself and I do a lot of cool stuff in that. So I get it, I get it. But there's so many other things that are incredible. Like, you know, one of my favorite companies that I've invested in is called Venus Aero and they're, they're building supersonic hypersonic drones that can go, you know, San Francisco to Tokyo in an hour. Like, we're talking about like extremely fast stuff in an ecologically sustainable way anyway. Like just like extremely cool, futuristic technology. There's lots of stuff like that. There's companies working on. Like some of the geoengineering companies I think are incredibly interesting. You know, there's. There's a company called Rainmaker which is trying to change the amount of habitable land on the planet, which is just, you know, an incredible. But they're one of several who are working on, on various forms of weather control, weather modification, and just Jason Crawford from Roots of Progress calls it installing a thermostat on the Earth to say, like, look, climate Change. Like, we don't have to treat it like this runaway externality. We could take control of the climate in this really profound way. Obviously, just huge dangers in all these technologies. But that's what makes them interesting. There's a lot of stuff in the crossover of bioscience and compu. You know, computational stuff. So basically like, like a new kind of bioinformatics for protein folding and for new therapies and just like, just, just so much, so much cool stuff going on. And then like, there's a lot of plain old boring fintech sass. Like, people act like this other stuff stopped happening, but it didn't. Like, it still is still going on. So, like, you know, we were talking about Costco. Like, I know someone's trying to build a Costco for banking. It's just, it's. It's phenomenal. It's super. Without the hot dogs. Yeah, no hot dog, no nothing. Yeah, exactly. Just. Just taking financial services and applying the, the capped margins and the fiduciary trust that Costco is able to make for its members into. Into financial services. So there's just like so many of these things where there's new, new ideas. You know, I tell the story in the book of Cost plus Drugs that is, you know, completely taking on the monopolies that are jacking up the prices of pharmaceuticals to unsustainable levels. Like taking what used to be considered like an intractable problem like climate change, or the cost of healthc care, or you name it, and finding private, you know, innovative solutions that through their technological mastery, change the terms of the debate of what is cap, what we're capable of arguing about, what we're so busy arguing with each other. Those are, those infrastructural innovations are the ones that determine what can be debated in the first place. And I think there's a lot of really cool stuff going on. Interesting. I recently bought a subscription to this, I don't know how to call it, Resource Project. It's called Internet Pipes and what's cool about it, it's basically just a bunch of links and resources that are a bit more, like, weird and things that we don't think about or places that we don't necessarily. Wouldn't necessarily go to on the Internet to look for stuff. And it was so interesting because you were listing all this stuff and I thought about that. Exactly. Because they're listing all kinds of data sets and databases and, you know, some tunnels into very, very old Wikipedia pages and like, you know, stuff that you can really like. You don't think about when you're living in your own LinkedIn bubble. So, yeah, it was really, really interesting. So I think, yeah, I think it's maybe dangerous to not try and get out of that bubble. But anyway, the last question on the podcast is always about a hack. So since we have, you know, let's come full circle. Since we were talking about the book, maybe you have a hack for founders that they can go and apply for that's super easy. How to make their company incorruptible. Hack. Okay, so this is, this is gonna sound, this is gonna sound stupid. Okay, but because, okay, when we talk about corporate purpose and mission and whatever, it gets complicated and sounds fancy and whatever. And so, you know, the book, I. There's a lot of history in the book. I try to really, like, really walk through the history of these ideas. How do we get where we are and how are we going to get out of this mess? At a society level, that's really important too. But from the point of view of an individual founder, like, I think actually most founders are way too galaxy brain, okay? They make things extremely complicated when in fact, this is very simple. Here's the hack. Just give the people what they want. Okay, where do people want to work? They want to work at a trustworthy, like, place where their humanity is respected and where they get to work every day on making the world a better place and accomplishing something that's deeply meaningful to them with people who they have that mutual respect with. Okay, that's not like, that's like, is it a hack or is it just like, duh, hello? What do people want to invest in? They want to invest in companies of transformational potential that are really trying to do something that like, has, you know, has the potential for really huge positive impact. It's one of the most, most easiest to look for indicators as, as an investor, who do people want to partner with is all this data that shows that when people are trustworthy, treat your partners with respect. Are you, you know, don't squeeze them when you don't have to. Well, you do. You just like, don't be an a hole, but actually, like, be nice to people. Like, that's, that's who wants to partner with you. And then like, what do people want to buy? The data is overwhelming that people want to buy from brands they can trust. So like, yes, you can do the growth hacking and you can like, do all this other stuff and you can make your brand and go on social media, whatever. But like the ultimate hack is Just to give the people what they want. Just be a good citizen, stand up for what's right, make a great product and commit yourself to the things that your customers, employees, investors and partners care about. And I know that sounds, I think some people be like that sounds so silly, so obvious, so whatever. But I, the vast majority of companies I meet are breaking one of those simple rules again having talked themselves into some complicated MBA hair brain scheme. And it's like. Or just do the simple thing. Thank you. I mean when you said, you know, people buy business books sometimes just to kind of put them on a shelf and you know, sometimes just too see just to show others that they support the ideas. I also heard a very interesting take that sometimes you buy a book just for the COVID because everything you need to know is on the COVID Right, right. That's true. So it's interesting because I also think it's so true. Most of the business books are just basically common sense that we forget about because we are in that loop in that tunnel. We're like, okay, we're doing that stuff and we fall into this no matter what growth metrics, blah blah blah, let's make our VCs happy narrative. But at the end of the day, sometimes you just need to look at that book on the shelf to remember. Okay, okay. There is this common sense that. Yeah, yeah, I don't. And listen, I don't. If people do that, I don't mind you buy my book for any reason. You know, totally fine. But. But I think it is funny that like business books are a little bit like cookbooks in that way that people like to have them on their shelf, but also for a, for a legitimate reason, which is first of all to remind yourself of like wait, what am I doing? But also in a business context, to signal the people that come into your office what you care about, what you're about. A lot of bosses do that as a way to subtly say these are the things I care about. And so yeah, my aspiration when I write a book is to write a book that you actually want to read. It's enjoyable to read. Yes, it's loaded with information you find super useful. But yeah, I also want it to have that effect. I want you to be able to put it. Buy a copy of the Lean Startup or of Incorruptible and put it on your shelf and have people now know something about you, who you are and what you stand for. So yeah, go for it. Yeah, awesome. Well, I'll be honest with you. I started, I didn't finish yet. But yeah, thanks so much for being here for talking us through the book, everything you stand for. And yeah, looking forward to the launch. Oh thank you so much. I appreciate it. Yeah, awesome. And take care. Great conversation. Take care. Thanks for listening. SaaS Unbound is brought to you by SaaS Group. We're a long term home for a great B2B SaaS. We buy, keep the team and brand DNA and help with the boring stuff like hiring and finance so founders can truly focus on building great products. If you're a founder who'd like to be featured or explore an acquisition, reach out through the form on our website or email me at Anasas Group.