Lovable CEO, Anton Osika: The State of Foundation Models, Grok vs OpenAI, and Replit vs Bolt

Anton Osika 0:00 I think university is not the best place to learn. Doesn't matter what you're studying. I'd invest in Groq, and I would probably short Anthropic. No, I would, I would short OpenAI, let's say.

Harry Stebbings 0:11 Why?

Anton Osika 0:11 I think it's more the slope on the Groq team. They're doing something which I respect a lot, which is to hire missionaries for the data curation part. Morale is super high. OpenAI has gone through all this mess, right?

Harry Stebbings 0:24 Do you think there will be a leading model that has not been created yet?

Anton Osika 0:28 Yes, from China.

Harry Stebbings 0:29 Do you worry about China?

Anton Osika 0:31 I do think there's, like, a 50/50 chance they will have the best model, that we'll be using a Chinese model at some point. Because...

Harry Stebbings 0:37 Ready to go? Anton, dude, I'm so excited to be here with you in person. Thank you so much for joining me on the show.

Anton Osika 0:54 It's great to see you, Ethan. Thanks for coming to Stockholm.

Harry Stebbings 0:57 Dude, I, it's great to be in Stockholm. Um, I want to start, you just recently

Harry Stebbings 1:01 raised a great round, and I want to start with that. We're seeing a lot of money go into this space, and I just wanted to start with, is it a capital arms race, and a case of who has the most money wins? Or is it something else?

Anton Osika 1:16 Uh, I think it's an arms race to build the best team, and then it's an arms race to build the, the best brand and trust from your users. And, I mean, capital can help. For us, it's not a constraint at, at all. Um, if you're building something like the best foundation model, it, it might be a constraint, just because the compute for training and so on is, is so large. But for, for us, it's all about moving extremely fast and collecting the best talent.

Harry Stebbings 1:39 So if we think about talent as the number one there, we've seen Zuck pay NFL-style contracts. I mean, like, mega, mega sums for the best people. How do you think about and analyze that, and how difficult it will be to, uh, get the best talent moving forwards?

Anton Osika 1:58 Hmm. Uh, I think for, for me, it's actually more difficult than for Zuck to know who, which engineers are going to really thrive, push the culture forward, push the ways that we're working in the products forward. Uh, for Zuck, it's like, there's these 10 people that, uh, know everything about how to train mo- foundation models, and he's more paying for that knowledge than for, like, these people. The talent itself is so good. It's probably, it's pretty good as well.

Harry Stebbings 2:21 Totally.

Anton Osika 2:21 Uh, so it's very different.

Harry Stebbings 2:22 Do you think you do not need the same caliber of engineering talent if you're working in the application layer?

Anton Osika 2:29 Mm, you just need very different, uh... I, I think, like, I w- at first, one of those people Zuck is hiring, they wouldn't p- perform as well as my, as the engineers in my team, doing what we're doing. So it's, it's very different type of talent. Um, and, like, I... If I knew who was, like, the perfect engineers to hire, uh, I could maybe step up our, our, like, our compensation bands, uh, to get exactly those. But, but I don't know who are the best people. So I, um, I need to just, like, figure out, are these really, really good people to work with? Are they moldable? Are they going to work well together in this team? Um, and then... And give, like, the compensation that you give on the top-of-market compensation rates for that.

Harry Stebbings 3:10 You've built an incredible team, also, like, of less obvious talent in the early days, and then you hire amazing rock stars like Alena Verna. When you look at your hiring process, your talent assessment process, is there anything that's non-obvious? So like for us, I look for people who have either extreme trauma or extreme masochism.

Anton Osika 3:29 (laughs) Yes.

Harry Stebbings 3:30 No, I'm, I'm being serious. I think not enough people are opinionated. You said brand is important. Br- great brands are opinionated. People love them or hate them.

Anton Osika 3:36 Hmm.

Harry Stebbings 3:37 Lovable, good, good example. (laughs) Um, but what is yours that is non-obvious about hiring or talent assessment?

Anton Osika 3:44 Hmm, so I like to think a lot about slope. And that, like, if I talk to someone and I learn a lot of things talking from them, and I, I notice that my conversation is, like, very dynamic and, and exciting, that, that is usually feels like a very good indicator that they're going to, uh, adapt to the org- organization, and, um, their slope will be very high. Um, I, otherwise, I think there are good ways to just understand how did they perform. Like, if I could be there with a video camera when they worked in the past, uh, that gives me a lot of signals. So that's, that's usually what I spend a lot of time, uh, when I'm talking to new candidates.

Harry Stebbings 4:21 When you think about slope, wha- it's noticeable with you two. Like, we haven't known each other for a huge amount of time. But when I compare when I first met you to when I meet you today, it is still very different in terms of your leadership.

Anton Osika 4:33 Hmm.

Harry Stebbings 4:34 Where have you not progressed where you would like to still?

Anton Osika 4:40 I think I still operate lovable in a very scrappy startup-y way, even though we're, like, at a later growth stage right now. Uh, so being... Adding a bit more structure in, like, in a few key areas is, um, somewhere I'm looking to progress or to, to start bi- like, being an excellent operator in.

Harry Stebbings 4:59 The joys of this show and being friends is we can have a discussion, not like a one back-and-forth interview.

Anton Osika 5:04 Hmm.

Harry Stebbings 5:04 Do you think you actually need that? We've had founder mode be so, uh, propagated and praised, and being close to the metal, Jansen having 52 direct reports. I would say structure and that middle layer is where slowness and apathy comes.

Anton Osika 5:19 True. Yeah.

Harry Stebbings 5:20 Do you think you need that?

Anton Osika 5:21 That's a good question. I, I think I'm going to always operate with this, uh, with most of my impact coming from, like, founder mode. But I do need, uh... Given that there's so many things thrown at me and coming in from all the different directions, to have a, kind of a protective layer that introduces a lot of order in, like, how to pri- prioritize all these incoming things. And, and that comes down to a well-running organization. Um, if I'm... And then, and then you... For a well-running organization, you need a very organized manager, uh, uh, somewhere at the top. And I'm not planning to be that, like top percentage manager-

Harry Stebbings 5:57 You don't have to have that today.

Anton Osika 5:57 ... myself, but surround myself with great leaders who, who do more of the, like, organization.

Harry Stebbings 6:03 Do, do you have a protective layer say 'cause I get probably-... 25 intro requests for you a week, and I probably make one a month.

Anton Osika 6:11 Yes.

Harry Stebbings 6:11 Do you have someone who does filter?

Anton Osika 6:13 I do, yeah.

Harry Stebbings 6:13 Yeah.

Anton Osika 6:13 And, and they say it's a kind of (laughs) um, wonderful, uh, chaotic protective layer that works together as, as a close team. And I don't really have like a name for it, it's just the, the people working closely with me. Uh, I'm not sure who you've interacted with in the team, but it's, it's... So, so the team is made up of, like, previous founder type generalists, uh, that work closely with me, and I just work in terms of, like, quick feedback, "This is not what you should be doing. This is what you should be doing. Um, works okay now, I think we can do even better."

Harry Stebbings 6:44 So you said talent was number one and brand was number two.

Anton Osika 6:48 Mm.

Harry Stebbings 6:48 If we think about great brand, what does great brand mean to you?

Anton Osika 6:53 I think, uh, super concrete example is the Apple ecosystem, where they obsess about details maybe too much so they move slowly, but that's what builds up trust and, and is a very strong brand. And that's sort of, that's what, what we're aiming for as well in every interaction. Like, every time we update the product, how do we make sure we roll it out so that we really understand users, uh, and how they react to all the things we're changing very rapidly in the product, in the

Anton Osika 7:19 company.

Harry Stebbings 7:19 There's a couple of questions which everyone has, where they will kind of throw them as a critique, at Lovable or at anyone in the space.

Anton Osika 7:28 Yeah.

Harry Stebbings 7:28 And I think one is protection, um, defensibility.

Anton Osika 7:32 Mm.

Harry Stebbings 7:33 When you think about defensibility today, is brand the core element of defensibility, or is there something that people do not see?

Anton Osika 7:42 No, I think you need to build a product, if you want to be, like, maximally defend- be defensive, where if you are on this product, in the, like, the platform that the product is, you don't wanna leave because you have so much value that's, you've created on the platform that you're getting automatically every day. So, I, that's what Lovable is becoming, and this product-building platform where you... Lovable today is your technical co-founder. We want it to be your co-founder in general that handles all the admin, setting up your finance operations. Uh, and if you're on a platform like that, you, you probably don't want, want to leave.

Harry Stebbings 8:15 Would you say to all founders building in AI today from day one, "Don't worry about defensibility, it comes over time"?

Anton Osika 8:22 Hmm. Yes, great question. I, I, I have, I have friends who ha- has this fun analog in terms of an AI startup, which is that AI startups are like chickens shot out of a cannon up in the sky, if you, if you kind of start getting traction. Uh, and then it's all about flapping fast as a chicken, because there are new chickens shot out (laughs) from cannons-

Harry Stebbings 8:41 (laughs) .

Anton Osika 8:41 ... uh, every day. And if you keep flapping faster than the other chickens, (laughs) then you're going to do great. And I think that's a good, like, first level of analysis in how you should operate.

Harry Stebbings 8:51 I'm just gonna say, for any vegans that are listening, no chickens were shot out of cannons.

Anton Osika 8:55 (laughs)

Harry Stebbings 8:55 And that is the most-

Anton Osika 8:56 Not yet.

Harry Stebbings 8:56 ... extremely Swedish way of (laughs) ... You know, it's the Reid, um, Hoffman who says, you know, "It's about kind of running off the cliff or whatever with a paraglider-"

Anton Osika 9:03 Mm.

Harry Stebbings 9:03 "... and just kind of flapping." That, that works too.

Anton Osika 9:06 Yeah.

Harry Stebbings 9:06 Um-

Anton Osika 9:06 I think, I think that's my recommendation, to just be, like, execute fast, grow faster. Um, and then when, when you're starting to get up there, you can start, maybe start thinking a bit ab- about defensibility.

Harry Stebbings 9:17 Totally get you. That's the one criticism. Another is, like, actually when you look at these businesses, and a lot of people are criticizing this with your Replets, your Bolts, your Lovables, they're not actually very good businesses in terms of unit economics, and so much is passed through. So, like, bluntly, if I give you a dollar, how much is passed straight through to Anthropic and OpenAI?

Anton Osika 9:39 Hmm. I don't give you the exact numbers, but if you look at the paid usage, it's majority. I- it's not everything.

Harry Stebbings 9:47 Okay. How does that change over time?

Anton Osika 9:50 Uh, so as our business develops, w- we're looking to get most of our revenue once you as a user are, like, "I love this platform, I'm never leaving." Uh, and, but today it's only, like, in the beginning and you're paying to build pr- pretty much. So over time, uh, we just want to create so much value you stay on the subscription, and the, the small, uh, part of the cost is, goes to the, to the AI g- compute.

Harry Stebbings 10:16 Will you be able to make money through not optimizing models? And what I mean by that is-

Anton Osika 10:21 Mm.

Harry Stebbings 10:21 ... in the future, you may not need the very best, very latest model to do the simple About Me website. And so you can route users-

Anton Osika 10:30 Yep. Yeah. Yeah, the, I think, um, as all applications develop, the AI is going to be adapted to those applications. And for most things, it's like, super simple to do it. It's like you're driving a car and you don't, you're not thinking about what you're doing. So when you're in a new situation driving a car, then you're like, your brain really goes on fire. And we're not there, we're not close to being there yet. I think for us it's too early to optimize for that, because the AI is, every month it's, like, doing new completely different things. So we just want to be able to iterate really fast on what the AI is able to do, and not optimize the models for what they, what it's doing.

Harry Stebbings 11:07 That's really interesting. So you build for what tomorrow's model can do, not what we have today?

Anton Osika 11:12 Yeah, to a, to a quite large extent, yes. And, um, like, what... Generally, in, when I think about models, uh, they're the models that are, like, very thoughtful and deep thinking, um, and now we put as much work on those models. In the future, it's going to be a mix, y- in terms of, uh, when it's obvious what it should do, then it, it just cost- doesn't cost any money, it's super fast. But when it's a new situation, um, which building a software product often results in, then it has to think much more.

Harry Stebbings 11:44 One other area where you can see real, like, margin expansion-

Anton Osika 11:48 Mm.

Harry Stebbings 11:48 ... is also in token selling. Like, when you think about how you price tokens, given prosumers and consumers don't fundamentally often know the price of tokens, you can actually have quite a considerable markup on token usage. Do you think that is a place of real elas- elasticity to gain margin, or not?

Anton Osika 12:04 Yeah, so w- we looked at, like, the number... This was a few months ago, but we looked at how much revenue is flowing through the AI from Lovable applications. Okay. And it was more than 10 million.... dollars in ARR. Um, and now all of that revenue needs the user to go through this, like, uh, bit complex process of setting up their connection to the, uh, eh, model providers. So, that's something we're just simplifying, uh, we're looking at simplifying. Uh, so, stay tu- to stay tuned for how we enable more simplicity, first of all, for our users with that, um, and then, uh, if we can reduce the underlying cost, may- maybe we can add- take a margin there as well.

Harry Stebbings 12:45 How do you think about mental plasticity to delay margin optimization?

Anton Osika 12:52 Mental plasticity, what does that mean?

Harry Stebbings 12:53 So like, the willingness to wait-

Anton Osika 12:56 Yeah.

Harry Stebbings 12:56 ... for margins to come.

Anton Osika 12:57 Mm-hmm.

Harry Stebbings 12:58 And what I mean by that is, like, if you look at Deliveroo, yeah? Its margins are shit in the early days, and over time they get better and better as you have more and more people use it and more density, more orders in small areas.

Anton Osika 13:11 Mm.

Harry Stebbings 13:11 You've got to be patient, so to speak.

Anton Osika 13:13 Yeah.

Harry Stebbings 13:13 Same with OpenAI.

Anton Osika 13:14 Yeah.

Harry Stebbings 13:15 Same with Lovable. How long does one think before you're thinking margin optimization?

Anton Osika 13:20 Hmm. Um, I, I have these two conflicting pieces of perspectives on it, uh, and one is, I speak to Nick, who built Revolut-

Harry Stebbings 13:28 (laughs)

Anton Osika 13:28 ... and he just tells me, like, "Anton, you need to compute the payback time, and then you need to do super hard performance optimization on acquiring new users." Uh, and then you, of course, need to have a good, like, payback, payback times in terms of profits for- per user, uh, which makes sense. And then, if you can do small changes in- on margins, it, it actually affects a lot on how m- how fast you can grow, uh, but, but the other perspective, which I index a bit more on right no- right now is, you just want to have as much mindshare and as many users who just love the brand as possible right now, and then you can think about that, that later. So, exactly how I trade those two perspectives off is, um, I mean, you d- you need to look at the weights in my neural network, but, uh, it's some combination of the two.

Harry Stebbings 14:11 It's so funny. I think it's absolutely number two.

Anton Osika 14:13 Mm-hmm.

Harry Stebbings 14:14 And then it's number three.

Anton Osika 14:15 What is that?

Harry Stebbings 14:16 And th- Which is Nick's, which is like incredible performance optimization-

Anton Osika 14:20 Yeah.

Harry Stebbings 14:20 ... really understanding funnel metrics from CAC to LTV-

Anton Osika 14:24 Exactly.

Harry Stebbings 14:24 ... how you drive efficiency through the channels, and then do you know what's ironic? You kind of then go back after that stage to an art, which is where Nick is now, which is, we've done that so well-

Anton Osika 14:34 Yeah.

Harry Stebbings 14:34 ... we have to sponsor race cars, because brand, again, becomes the most important thing.

Anton Osika 14:40 Mm-hmm.

Harry Stebbings 14:40 It's so interesting. You know that kind of bell curve where you see it up and then it comes down again? It's, it's kind of like that on brand, where you have it at both ends of the spectrum.

Anton Osika 14:47 Yeah. I think you, like, if you can do everything at once, the company benefits a lot, but you usually can't, like you should, you should be focused.

Harry Stebbings 14:55 What would you most like to do now that you're not doing or can't do?

Anton Osika 14:58 Mm-hmm. So I would like to rethink how the applications are built, like what is the best way to build an application, right? What Lovable does now is it takes all the best practices from decades of, like, how a great software prod- product we built, but that's not the, how the future is going to look like. They're going, all software applications are going to have some ty- type of AI. They're going e- they're going to have extremely seamless payment and checkout flows, and that's something I'd love to, uh, for us to spend time on, of, figuring out and making possible for our users not t- to just have, like, a superhuman AI engineer, but to have a AI engineer that builds, like, the future of applications.

Harry Stebbings 15:41 When we look at ... We mentioned, um, kind of margin optimization there. When we look at kind of the model providers-

Anton Osika 15:47 Mm-hmm.

Harry Stebbings 15:47 ... we said that it's one way you also have to have that mental plasticity. We saw OpenAI kind of suggest or proffer Lovable-style competitors. To what extent do you feel OpenAI and Anthropic will come after Lovable in the way that Claude code comes after Cursor?

Anton Osika 16:04 I think, uh, in the long term, it just comes down to execution of a team, um, and I think many people are going to offer what we're offering today. We just need to offer much more, eh, when the, when that time comes, and have a, give a better user experience, give a better value proposition to our customers.

Harry Stebbings 16:23 Who do you worry more about OpenAI doing or Anthropic doing it?

Anton Osika 16:28 I think, um, w- what we're betting on is to be, um, like the gateway for, for humans, right? And be the best user experience for, for AI, um, and so far, OpenAI is doing that better than Anthropic. So I see them (laughs) as a more serious competitor in

Anton Osika 16:45 tha- in, in 12 months.

Harry Stebbings 16:46 Mm-hmm. How did you analyze GPT-5? And when you look at performance post, are you more or less bullish on OpenAI?

Anton Osika 16:54 So w- we looked a lot about, at how does Op- how d- sorry, we looked a lot at how, uh, GPT-5 would impact our users before we decided, "Okay, let's put this into the product." And we looked at the, how long time it took to get responses, we looked at our quali- quantitative evals, and, and then we just vibe-checked it in many different ways, and what we concluded was that it's oftentimes too, like, ambitious for our users-

Harry Stebbings 17:21 Hmm.

Anton Osika 17:21 ... so n- and that's why we decided, "Hey, um, this is very smart, so let's give it to all our users and see what they, what they tell us w- in terms of how, like, what's good and what's bad." Um, what we found was that, um, it's, like, for the use cases when you had to solve a really, really hard problem, it's great, and all, in terms of is OpenAI doing great job, I think th- this was a really smart, uh, obvious choice for them to say like, "We have these five different models that you have to select in ChatGPT. Let's just bring it down into one model, GPT-5." Um, but, so they sh- they should, definitely should have done that, but it, but it comes with a lot of trade-offs, and so far, I'd say they executed pretty well on it. The model is still too ambitious.

Harry Stebbings 18:10 There's also a question of when you set the bar at AGI-

Anton Osika 18:13 Yes.

Harry Stebbings 18:13 ... and then you get model optimization-... and model routing, really essentially-

Anton Osika 18:17 Mm.

Harry Stebbings 18:17 ... which is what it is.

Anton Osika 18:18 GPT-5, yeah.

Harry Stebbings 18:19 That's it.

Anton Osika 18:20 Yeah.

Harry Stebbings 18:20 Like, capability wise, it hasn't been a step function improvement than what we had before.

Anton Osika 18:25 No. No. I, I don't think, uh, like, it, it, the biggest part of GPT-5 that's disappointing is that n- now they have to optimize all these different things into one model. Before, it was like different models, and they had to do it really fast. So it's inevitably going to fall short in some dimensions. Um, and I mean, it, it just is a disappointment that you can't improve in all the directions at the same time.

Harry Stebbings 18:47 How do you use OpenAI versus Anthropic within Lovable today?

Anton Osika 18:51 Mm. So we have a, this very complex agentic chain, where we pass the user's response, uh, the application information in, through many different models. And then can, we take the really fast and small ones, and then we use, for code writing, we usually use Anthropic. Um, and, and right now, you can say, like, "I want to use GPT-5." Um, and that's better when you're solving a really hard debugging problem.

Harry Stebbings 19:16 Super inter- And you've seen it be better than Anthropic when it comes to a hard debugging problem?

Anton Osika 19:20 Mm-hmm. Yeah.

Harry Stebbings 19:22 What do models not do today that would be a step function change in what Lovable can do?

Anton Osika 19:28 I think, well, like something I'm, I'm super excited about is that the AI has more context about who they're talking to and how they should be answering to, you know, guide them through our specific application.

Harry Stebbings 19:41 Mm-hmm.

Anton Osika 19:41 And, uh, they, solving that problem is something that we have to do, and we have to do it, uh, both with, like, how we build this agentic chain, uh, and over time, in building, uh, absolutely world-class, paying $100 million for getting the people that train the models. So that, so that's on the, on the horizon for us to, to get it to, like, be hyper-personalized for you specifically.

Harry Stebbings 20:02 When you think about hyper-personalized for you specifically-

Anton Osika 20:04 Yeah.

Harry Stebbings 20:05 ... and the users that you have, you know, you recently announced the $100 million, or amazing milestone to hit in seven months.

Anton Osika 20:11 Mm.

Harry Stebbings 20:12 That's fucking nuts. For years, dude, it was like zero to ten million in two years, was like the gold standard.

Anton Osika 20:18 Yeah.

Harry Stebbings 20:18 That's what I was brought up on, which makes me feel really old. Uh... (laughs)

Anton Osika 20:22 (laughs)

Harry Stebbings 20:23 Um, which is, uh, amazing, um, to see. My question to you is, when you look at revenue breakdown, of the $100 million, just kind of guesstimate, what is split between hobbyists, pro devs, kind of normal people? How does it fit-

Anton Osika 20:39 Hm.

Harry Stebbings 20:39 ... between the different segments?

Anton Osika 20:40 You're right. So, uh, people do everything with Lovable. They come with their idea to build a software business and product, and then there's a lot of people in large companies, they, that use it as like, "Okay, now I can prove, show what I actually think we should build in, in the business." And then they buil- they build a working, uh, product that then they can like decide, "Are we going to put this, this into our, uh, give it to their, our, our engineering team?" And they actually implement it. Um, and then it's everyone else who build, like, their personal website, their small business websites, where, you know, like in a few minutes. And 80% of people are, are in the first category, that they're building real, uh, complex applications, in terms of revenue.

Harry Stebbings 21:19 80% are building complex applications?

Anton Osika 21:19 Yeah, in terms of revenue, 80%, 80%.

Harry Stebbings 21:21 Wow.

Anton Osika 21:21 And then, yeah. But and then the second segment is actually, uh, growing very fast, because, uh, like, enterprises are slowing to wake up, uh, slower to wake up. But you might have seen this product leader from Google who says, like, "We're never again writing a document about a product. We have to use Lovable or something to build out a fully working demo." Uh, so that, so that's, that use case is also growing very fast. And, and in terms of the third use case, I mean, uh, a lot of people have been burned trying to build nice websites in this, like, uh, no code website builders, Wix, Squarespace, and so on. And if you can just always do everything in Lovable and with a UX that I think is more sophisticated in moving fast, um, that's, I, that's also growing. But I think the, the first two are the ones who are really, like, game changers.

Harry Stebbings 22:05 Okay. So we go back. So 80%, sorry, is like actually building complex apps.

Anton Osika 22:10 Yeah.

Harry Stebbings 22:10 And then 10% is enterprise and 10% is hobbyists?

Anton Osika 22:12 Something like, something like that. Uh, th- yeah. Something like that.

Harry Stebbings 22:15 Is that what you want it to be?

Anton Osika 22:17 So, uh, we want to build for the new generation of AI-native founders that, um, build, like, maybe one person unicorns soon. And, and then the, the, the funny thing is that those people also have jobs maybe in, in large, successful companies. Uh, and they wanna help their friends and their family to build, uh, simple websites. So I think this is a, yeah, I think this is a good split. But-

Harry Stebbings 22:41 Is that an optimal market to go after if you're thinking about... God, I sound like such a VC, but like value extraction, which is like an AI founder building a mega business on Lovable. Great. You want to have to have a lot of mechanisms for value extraction-

Anton Osika 22:55 Mm.

Harry Stebbings 22:55 ... and then, be it payment solutions or you name it. But if they're single seat, uh, and p- it's g- just, like, tough to get true value extraction from that, is it not much better to be hobbyist for everyone, for mom and pop to build the About Me website, where it's seven billion people?

Anton Osika 23:09 Hm. Uh, like our mission is to enable a lot of people that have the opportunity to build businesses, but they have been held back by not being able to write code and like not have access to capital to hire engineers. Um, so it's obvious to start with the people who are going to build businesses. And then, uh, it naturally trickles down to everyone else as well, as a function of that. If, like, those, I think those are the best people to start building for. And, uh, where you can extract value, I think less about that. I think about our mission.

Harry Stebbings 23:39 We should have like a Lovable holiday fund, which is like-

Anton Osika 23:43 Uh-

Harry Stebbings 23:43 ... every year, we pay the most talented people within large enterprises for a week's holiday. (laughs)

Anton Osika 23:51 (laughs) So that they build their business.

Harry Stebbings 23:52 Yeah.

Anton Osika 23:52 So they, they can pay off the other people, yeah.

Harry Stebbings 23:53 And, and, and then they just use Lovable for the week.

Anton Osika 23:55 Yeah.

Harry Stebbings 23:55 And after the week, they can see if they want-

Anton Osika 23:58 They leave, they quit their job. (laughs)

Harry Stebbings 23:59 Yeah. I think it'd be the funnest thing ever.

Anton Osika 24:00 I think it sounds good. Yeah.

Harry Stebbings 24:01 Yeah.

Anton Osika 24:01 Yeah.

Harry Stebbings 24:01 Um-

Anton Osika 24:02 But I could just, uh, expand a bit on like the, on the business thinking there as well.

Harry Stebbings 24:05 Yeah.

Anton Osika 24:05 So, um, like the future, I, I think many of the largest businesses, they haven't been create- started yet today. And now with AI, you can move, move so fast. You can iterate much faster, close to your customers. You can drive prices down. And we want to be the enabling tool, tool for that movement.... and, and then, um, I think over time that's going to result in a lot of revenue. But, like, w- if we can do every, all of these use cases as, that you're asking about at the same time, what, uh, our, like, use case percentage of adoption or revenue is going to converge towards is just, what's the, what's the spend from enterprises? What's the, like, uh, the, the end game spend from enterprises? What's the end game spend for consumers on tools like this if we continue to dominate this completely new category?

Harry Stebbings 24:55 Obviously, I'm super grateful for you for taking my money, but the reason, the number one reason why I would invest in Lovable is because the market TAM or the TAM expansion is actually incomprehensible, which is, like, very much like Uber. You could never have foreseen the market expansion that would take place.

Anton Osika 25:11 Yep.

Harry Stebbings 25:12 Very much like Lovable, saying website builders is an ex-market is completely the wrong analogy to understand how big Lovable could be moving forwards. And that's a common thing of the best venture investments ever made.

Anton Osika 25:23 Hm. Yeah.

Harry Stebbings 25:25 It's really interesting.

Anton Osika 25:26 Yeah. Yeah. I, I think also, just want to say on the enterprise use case there, so if I was a CEO or, like, a C- CTO of a large enterprise company, I, I wouldn't think in terms of, "Oh, how can we make our engineers more productive?" I would think in terms of, "How can we get the most information about what we should build as quickly as possible, uh, into new products, into- or into our ex- existing products?" And that requires everyone in the company to be able to work in one place to change and edit their products and propose new changes to it. And it's hard for us and- or for any- or for anyone in that, uh, matter to build a product that does that, like, tomorrow. It's better for us to start with the founders who are building it from the ground up and then move it in- into the enterprise, the same experience into the enterprise.

Harry Stebbings 26:14 It also allows for this incredible democratization of ideas within companies.

Anton Osika 26:18 Mm-hmm.

Harry Stebbings 26:18 I interviewed the CP at Duolingo for 20 products, and he actually said how two designers, not traditionally ones who actually come up with obviously building products-

Anton Osika 26:27 Yeah.

Harry Stebbings 26:27 ... from day one, um, created chess in Duolingo, and they did it with... I can't remember what the tool was.

Anton Osika 26:33 Mm-hmm.

Harry Stebbings 26:34 I hope it was Lovable. Please say it was lovable. Um, and that was actually their first iteration of it, which I thought was an amazing instantiation of this. So, I totally get you there. Does that mean then that we lose the design process and the brainstorming process? Do we skip that and go straight to prototyping?

Anton Osika 26:52 Yeah. Uh, l- look (laughs) . So, uh, to, to date, what you've done is that you've taken an idea and then you went through many, many, many steps until it's, like, a fully fast-growing product. And all of the... I call that a product life cycle. One part of it is writing the code, which is, like, where AI now has made it much faster. Um, there are many steps after that. There are steps before that, which is to, like, mock it up, validate it internally, validate it with your users. And, uh, what we've done so far is to take all the first step until, like, this is validated, this is what we need to ship. And, and even we have e- even external users on it into one few minutes or a few hours of building. Uh, so that's where we've seen, like, the most, there's the most maturity on our product. The steps that come after is something we have to build out as fast as possible so that you don't need, like, a products, design, engineering organization. It's one, all one tool where anyone with the best ideas spend the most time.

Harry Stebbings 27:50 The steps after or the steps before? And what I mean by the steps before-

Anton Osika 27:53 Yeah.

Harry Stebbings 27:53 ... is you see at the other end of the spectrum Figma doing Figma Make, where they're like, "Hey, we'll always have the design process and then we'll move with you into the second phase of that product life cycle-"

Anton Osika 28:03 Yep.

Harry Stebbings 28:03 "... into the build or prototype phase from design to prototype." To what extent do you worry about entry from that earlier standpoint?

Anton Osika 28:13 Uh, entry from F- Figma or-

Harry Stebbings 28:15 Yeah. Entry from Figma, given the fact that they own the design part of the, uh, phase to just then move into prototyping.

Anton Osika 28:23 So, I think humans, uh, are sometimes too obsessed with small details being perfect and you, th- which makes you s- move much slower. Uh, and, and the way of doing it right now with the designing where, like, one person does all the design very slowly, uh, very detailed, um, is going to be replaced by AI doing... You, you talk much more high level and you, you talk about your design philosophy and then the AI does the implementation of design. Um, and then you as a human go out and get all the context from th- "Do other people think this is, looks good?" And give, give it all as feedback into the AI. Um, and, and then there's a very seamless opinionated way of taking it all the way to a product, uh, with all the, like, marketing and growth functions in, built in the AI behind it, as well as all the, uh, tooling you need to evolve, uh, high quality software product which comes like testing, quality assurance, and so on. And that's the, that's a new way of doing it. And, and, uh, thinking you should be doing design with Make and like the Figma tools is, um, m- for ma- most people I, I would, uh, my th- is that it will slow you down too much. Does that make sense?

Harry Stebbings 29:30 It does.

Anton Osika 29:31 Mm-hmm.

Harry Stebbings 29:31 What happens to Figma then?

Anton Osika 29:33 I think, uh, for some pixel perfect things it's going to be per- amazing, amazing to continue to use Figma. Um, I don't know how the, like, the distribution will look like in terms of doing it with a more opinionated way, which is what our platform is becoming versus, uh, how many companies want to continue to do it like you do it now in, in a tool like Figma.

Harry Stebbings 29:52 Do you think you're opinionated enough?

Anton Osika 29:55 Mm... I think that our tool enables a lot of flexibility at the cost of some velocity on o- our product development.

Harry Stebbings 30:06 Mm-hmm.

Anton Osika 30:06 But I think it's a pretty good sweet spot. You can build with Lovable and then any engineer can come in and edit and, uh, take over if they want to. So, I think, um, yes, we would, we would be moving faster if we were more, even more opinionated about how things should be done.

Harry Stebbings 30:21 What are you not opinionated on that you would like to be in a dream world?

Anton Osika 30:25 Like, in a dream world, I-... we would be even more opinionated about how you build an application, and we would know what's the future of building p- applications, with AI being such a core part of it. I don't think it's possible, because how AI works and, like, what's the best UX with AI for the products that are built with Lovable changes so rapidly. So, uh, at some point in the future, I'd love to be there. And when, what happens if we can be more opinionated is that you get the, um, like, the right level of detailed adjustments on how the AI works for your product, how the, uh, back end flows and workflows and automations work for your product. And now, we have, we support two, a lot of different things, so it's more, um, you need to be really good at prompting right now.

Harry Stebbings 31:15 Do you think we will prompt in five years time?

Anton Osika 31:18 Mm, yes, I think so. But maybe it will evolve in terms of, um, how you do it. It will... Like, hyper-personalization takes care of a lot of detailed prompting that we have to do today.

Harry Stebbings 31:31 What does that mean?

Anton Osika 31:32 So prompting is basically providing context to an AI of what your goals are and how you want it to do something. If it's, it's, it's like when you have great, uh, employees for you, right? They, they know everything about how you, how you want it to work. So you just have to say, "Let's go to Stockholm and do a hackathon," and then it just magically becomes what you want it to be.

Harry Stebbings 31:53 It pretty much does (laughs) -

Anton Osika 31:54 (laughs)

Harry Stebbings 31:54 ... honestly. I had z- I sent a picture to my mother beforehand and she's like, "You had nothing to do with that." And I'm like, "No, I did not." (laughs)

Anton Osika 32:03 Mm.

Harry Stebbings 32:03 She's like, "I know." (laughs)

Anton Osika 32:04 Right. Hmm. And you can't just say, tell, uh, ChatGPT or something, "Just go" (laughs) ... "Let's go to Stockholm and do a hackathon." He's gonna come up with something else than what you had in mind, right?

Harry Stebbings 32:12 Mm-hmm.

Anton Osika 32:12 And so you can either hype, like, prompt it very, very detailed or you can make sure it's, it knows how you think, and that's what we'll, uh, be evolving towards.

Harry Stebbings 32:22 We said the word opinionated-

Anton Osika 32:23 Yeah.

Harry Stebbings 32:23 ... and we spoke about it with regards to the company. Uh, dude, I love your social media presence because you're also opinionated in your social media presence. And I think it's, respectfully, also just quite blunt. And you've been opinionated in how you talk about competition-

Anton Osika 32:38 Mm-hmm.

Harry Stebbings 32:39 ... and specifically, like, a Rapplit of the world. How do you think about whether or not to engage in an opinionated stance against competition or not? (laughs)

Anton Osika 32:49 Um, look, I don't think so much about competition. Uh, I, what, the, the only thing that matters is that we make our product the best product, and we continue d- deliver, uh, on, like, our value promises to our customers, uh, if there is something that happens. So, there, there was a competitor that, like, found a lot of apps that have been poorly made, um, and they, they said, "Oh, this is a vulnerability." And, and I, I spoke to a lot of security c- professionals. That wasn't really, like, how you would normally, uh, announce vul- vulnerability. So then I went on and bashed them, (laughs) as a, as an outcome of that. I, I think that was a very reactive thing, and I think it was, like, uh, something I'd be happy to share in person to that competitor, like,

Anton Osika 33:33 f- face-to-face.

Harry Stebbings 33:34 Well, I mean, it was, it was interesting 'cause, uh, uh, Jason Lampkin, who's a friend of mine-

Anton Osika 33:38 Yeah.

Harry Stebbings 33:38 ... I don't know if you saw it, but he was using R- Rapplit. Um, and then I'm, I can't remember exactly what happened, but they basically had a massive security breach or they deleted all his database or something bad happened, and it was like code red for them. And the takeaway for him was like, just security on all of them-

Anton Osika 33:56 Yup.

Harry Stebbings 33:56 ... is just nowhere n- near where it needs to be. And like, it's wrong for Rapplit to bash Lovable, it's wrong for Lovable to bash Rapplit, all of you guys suck at security. Is that true?

Anton Osika 34:06 Uh, yes, I think they are. (laughs)

Harry Stebbings 34:08 (laughs)

Anton Osika 34:08 They, they are, uh, like, so let me say it different way. First of all, we talk about security, like, company-wide every week almo- like, every day I, I hear something about security because we take it so seriously. And, um, there's so many different fronts to make it much more secure than if a human would do the, the application development. And that's why it's so important for us to be the best in the world at security. And-

Harry Stebbings 34:31 So you're saying it's more secure than humans.

Anton Osika 34:33 Not yet. Uh, for, I mean, I s- I said this at some point, which is if, um, you take your average developer who, like, normally works in a large team where they have a lot of support, and then they, they, that human goes out and builds an application, they are going to create software that has security holes on average.

Harry Stebbings 34:54 Mm-hmm.

Anton Osika 34:55 Um, and l- when you build a application with Lovable, it's going to tell you to go through a bunch of security reviews, and the AI is going to do a bunch of security reviews and, finally, it's going to give you green light, like, "We haven't found any security vulnerabilities." So if you compare those two, the, like, average, truly average developer with Lovable, Lovable is going to, uh, have a lower chance of having a vulnerability. And so we want them that, put that to 0%. We need to put that at 0% chance of vu- vulnerability.

Harry Stebbings 35:24 It reminds me of self-driving-

Anton Osika 35:26 Yeah.

Harry Stebbings 35:26 ... which is like, you know, for the world's best driver, I'm sure you are better than self-driving. But for the majority, and for the majority who are tired-

Anton Osika 35:34 Yes.

Harry Stebbings 35:35 ... humans have the potential-

Anton Osika 35:36 Yeah.

Harry Stebbings 35:36 ... to be hungover-

Anton Osika 35:37 Yeah.

Harry Stebbings 35:37 ... high, malfunctioning in some way, actually wildly dangerous.

Anton Osika 35:40 Absolutely.

Harry Stebbings 35:40 It's much, much better.

Anton Osika 35:41 Very much so. Yup. So, yeah, and I, I, I would say I'm very proud of what the team done so far with the security and, and, but there's more to come. Yeah.

Harry Stebbings 35:51 So when we look at... We, we've spoken about m- many different kind of, uh, competitors in different ways, from your Figmas to your Rapplets. If you move forward three years, what does the space look like then?

Anton Osika 36:03 Again, I, I think I focus on what our product does and how we serve our customers best. And, like, I don't really predict that. If, if we get all the, the majority of the profit share in this market, that's amazing. Um, if there's, if this is spread out across different companies, that's, uh, that's also fine as long as we build a product that lasts for generations. And that, and I do that by building the best products for our customers. Um, if, yeah-

Harry Stebbings 36:27 And this is why brand is so important for you.

Anton Osika 36:29 Yeah.... that, that's how I think about it.

Harry Stebbings 36:31 Do you mind if devs go to Lovable, get 60% of the code from there, and then fine tune it?

Anton Osika 36:38 Pro-

Harry Stebbings 36:38 How do you feel about that?

Anton Osika 36:39 I don't. Like, from the get go, two years ago, I, I decided that we're gonna build Lovable bec- uh, w- with a, for a world where h- humans don't write code anymore, and they, we're quickly moving there. We're very, very quickly moving there. Uh, today, I don't mind at all. Like, it should be flexible. You, sh- some humans have their way of doing things, and, um, it's, I think it's good if you have, like, a, a ecosystem where you can use, uh, many, many, many different tools on, on the product. Over time, I, I think it will converge towards the very opinionated platform, like the ones we're building towards, and everyone's just s- going to look at what's the cost benefit of doing it with some other tool as well. And only using Lovable is going to be the o- o- obvious, uh, m- most, like, velocity, high velocity and high-quality driving, uh, choice. That's-

Harry Stebbings 37:24 T-

Anton Osika 37:24 ... that's the future.

Harry Stebbings 37:25 Today, does AI make 1X engineers 10X, or does it make the 10X engineers 100X?

Anton Osika 37:32 Hm. It does both. The, it really does both. But in different-

Harry Stebbings 37:36 I have to force you to one.

Anton Osika 37:37 It... so... (laughs)

Harry Stebbings 37:39 (laughs)

Anton Osika 37:40 Uh, for the junior engineers, the 1X engineers, um, they often are bad at something, right? If the AI can bridge that gap, it takes them from zero to one, and to be able to do something they, they were not able to do before, it's like a 10X, they go to 10X or even more. Um, so if, so if that's the case, then it, they definitely got... Then it's more valuable for them. If it's, um, a 10X engineer that is working on something where you need a lot of, like, w- s- uh, many years of experience to work on a system that a new junior engineer completely doesn't understand, the using eng- the 1X engineer doesn't understand the system, then the 1X engineer is useless, so no way AI is going to (laughs) increase their velocity. Whereas the 10X engineer is going to maybe go to 100X e- engineer.

Harry Stebbings 38:27 How will the size of engineering teams change in the next five years?

Anton Osika 38:32 I think the b- for the best companies, engineers will really act as this translation layer. They will, uh, need to be more, thinking more in terms of product, being a product manager. Um, and I think there's, like, m- a higher elasticity on such engineers. So, you might see s- many companies being like, "Oh, more engineers. We can do even more, more, even faster." They are out talking to the customers and changing things with AI super fast. Um, but over-

Harry Stebbings 39:01 Do the skills required-

Anton Osika 39:02 Yeah.

Harry Stebbings 39:02 ... to be a good engineer, then, change with time?

Anton Osika 39:05 Yes, definitely. Again, I think y- being a generalist becomes more and more important with everything, in everything with a- with AI, um, so that you can understand how things g- come together in, as a larger whole, and then you use AI for the, like, deep expertise that you don't need in the future as much.

Harry Stebbings 39:24 When you think about skills required too-

Anton Osika 39:27 Mm.

Harry Stebbings 39:28 ... there's a lot of people who are asking today, "Should I bother studying computer science if we're gonna see Lovable be the last, you know, software that we ever need?" How would you advise your little brother questioning whether to do CS at university?

Anton Osika 39:45 I think university is not the best place to learn. Uh, that com- th- it doesn't matter what you study. Uh, you should be out there and s- really understand how the world works in terms of how work translates to value creation, and you don't learn that at university. Um, so university is, like, a way to m- train your brain to learn new things and, uh, meet a lot of interesting people. Uh, so, uh, computer sh-

Harry Stebbings 40:08 Would you encourage your children to go to university?

Anton Osika 40:12 So, this is now 20 years in the future pretty much.

Harry Stebbings 40:15 Oh.

Anton Osika 40:16 So, it's hard to say, (laughs) to say something about 20 years in the future. I think it's a great experience to have had in life, uh, so why not? But it depends on what outcome you want to reach. If you wanna have a, a job that's where you make the most money, no, you shouldn't, they shouldn't go to university.

Harry Stebbings 40:34 I think the opportunity cost of those years is very high.

Anton Osika 40:37 Mm. True.

Harry Stebbings 40:37 Yeah, in the, in the UK in particular, we just get very drunk for three years.

Anton Osika 40:41 Yeah.

Harry Stebbings 40:41 And that's generally how it is, and generally study generalist subjects like geography and history. And honestly, it is a little bit of a waste of time, in which case you can utilize those years so much better-

Anton Osika 40:52 Mm.

Harry Stebbings 40:52 ... given your stamina, your energy, the plasticity of your brain at that age-

Anton Osika 40:57 Sure.

Harry Stebbings 40:57 ... which is why I highly advocate against it.

Anton Osika 40:59 Agree. I mean, if, if you just do a v- a very, very specialized job for, for those years, maybe you, you'll become less of a generalist, so the, there's, like, there's a trade-off there as well, of course. I think, whereas university, you're exposed to many different concepts, which can be useful.

Harry Stebbings 41:15 We mentioned earlier AI and enterprise. When you look at the biggest enterprises today, they're not able for data, for permissioning, for security, to use AI. Are we going to see the biggest shift in incumbent power in the next 10 years?

Anton Osika 41:33 So, you mean if they're not enabled-

Harry Stebbings 41:37 (laughs)

Anton Osika 41:37 ... uh, there's someone else that c- will come in and, and be enabled?

Harry Stebbings 41:39 Exactly.

Anton Osika 41:40 I think you see this in banking, like for example, where bank is a software company, right? It's all about software systems. And the old banks are moving much slower. I, I think, yes, I, there is going to be some companies that are, like, built ground up for an AI to change their s- their sys- their systems. Uh, and anyone who's exposed to, like, customers understanding the legal requirements and so on can move much, much faster in the creating a good customer experience. Um, so yes, I, I g- I imagine there are also some, um, benefits of being, having been around for a long time in the enterprise. In, in, like, banking, there's a certain element of trust and so on. Uh, so I don't know how large the shift is going to be in, in the, in, like, across different segments of the enterprise market. But, um, many companies will be, get disrupted by cheaper, much, much better alternatives.

Harry Stebbings 42:32 It's interesting you said there about trust. How loyal do you think Lovable users and customers are? Do you think there's a high propensity to switch and an ease to switch, or do you think people fundamentally are loyal?

Anton Osika 42:42 It's 50/50. Like, some people are just super, super loyal to a brand. Uh, many people-... I mean, if you do something that hurts your brand, they will, they will switch, and they're just out there looking for, uh, maximizing some kind of cost versus, uh, capability subjective. So, so there's, there is, like, you can do, um, you can think about both of those groups simultaneously, um, and, but it, like, if you have the best product with the best value, you are g- going to get everyone.

Harry Stebbings 43:12 What question... We, we spoke about kind of people being, um, threatened-

Anton Osika 43:16 Yeah.

Harry Stebbings 43:16 ... large incumbents. What question should large CEOs, business leaders be asking today about the future of AI in their companies and how they use it, that they're not asking, do you think?

Anton Osika 43:28 Hm. I think, um, one of the biggest bottlenecks for these companies is going to be s- some kind of change management for the humans in the organization.

Harry Stebbings 43:39 Mm-hmm.

Anton Osika 43:39 Um, and I think they should be asking how have similar companies to ours done change management very, very, very rapidly? Uh, and get that conversation into, like, the leadership room, and then maybe across the entire organization to start, study those examples of where change management has been very, uh, successful and then look at specifically, um, which AI tools should we be using. Should we be adopting, like, building a product on top of Lovable 100%? Because then everyone can collaborate. Or should we doing, uh, some, should we hire some new type of people to come in and upskill everyone?

Harry Stebbings 44:13 You said about speed there, you said about talent earlier.

Anton Osika 44:16 Mm-hmm.

Harry Stebbings 44:17 Bluntly, dude, I get really fed up with everyone saying that Europeans are about espresso and taking the summer, and, "It's August and July, we're not gonna work." And I advocate for a very aggressive work culture, which, uh, you know, is 996-

Anton Osika 44:33 Yep.

Harry Stebbings 44:33 ... and how I say that. Um, how do you feel about the importance of unwavering hard work over balance in the desire to win?

Anton Osika 44:48 I think over a 10-year period, I, I would advocate for some balance.

Harry Stebbings 44:53 Yeah.

Anton Osika 44:53 But if over a two-year p- period, uh, if you really care about something, then you should make sure that you have, like, you get your exercise, your sleep in really, really well, and maybe something that you know relaxes you, uh, and then just work your ass off (laughs) . That's what you should be doing.

Harry Stebbings 45:10 Do you agree then with Scott from Cognition, who clearly said to windsurf employees-

Anton Osika 45:16 Yeah.

Harry Stebbings 45:16 ... after hiring them or buying the company, it's six days a week, it's unwaveringly relentless, and if you don't want to sign up for that, you can leave?

Anton Osika 45:27 I would say m- in how we think about it, is that you are here to have 10x impact over the other people at other companies. And if you don't have 10x impact, uh, so for, you do that by being very talented, being good at your job, and, uh, being very focused. Uh, and for some people, you need to put in a shit ton of hours, and, and for th- but not for everyone. So I, I would prefer... I, I'd speak about are you s- am I seeing the impact? Are you s- am I seeing that if you would tell me you're leaving tomorrow, I would be like, "No, you are, like, such an important part of this company, you have to stay." Uh, and that's how I push, uh, performance, uh, and impact.

Harry Stebbings 46:05 Do you do the keepers test?

Anton Osika 46:06 Yeah. I do the keepers test. Yeah.

Harry Stebbings 46:09 Has it made you change how you construct teams?

Anton Osika 46:12 The keepers test?

Harry Stebbings 46:13 Yeah.

Anton Osika 46:14 Yeah, I, I think, um, a, it's, it always under- makes it clear to people that, um, I need to always figure out am I... Like, how can I have more impact? Uh, so that's one part. And then, um, I mean, I, I think in terms of what does this organization look like if it's in the, um, uh, is it optimally set up to suc- to succeed right now? And, um, I don't, I don't do... Like, culture is such an important part. If you're just, like, throwing people around too much, it's, it hurts the culture and the ways of working. But doing this business exercise of saying, "Is this c- organization set up perfectly to win?" uh, definitely shapes how I build an organization.

Harry Stebbings 46:53 You mentioned Nick earlier-

Anton Osika 46:54 Yeah.

Harry Stebbings 46:54 ... at Revolut. He gave me the best answer, I think, ever on culture. You know, I've done 3,000 shows, and when culture comes up, it's, like, first principles thinking, I'm like, "Ugh, fuck it, we'll edit this bit out." Um-

Anton Osika 47:04 (laughs)

Harry Stebbings 47:05 Always. B- but he said the best thing ever. He said, "I don't think about culture, I think about winning."

Anton Osika 47:08 Mm-hmm.

Harry Stebbings 47:08 "The single biggest determinant of human happiness is growth and development. And when you are winning, you are most optimally positioned to grow and develop. And so if I create the conditions to win, you will grow and develop." And then supporting that, the other thing that people like to do is accumulate wealth (laughs) as well as development.

Anton Osika 47:26 Yep.

Harry Stebbings 47:26 And you will accumulate that by winning because of your share price increase. It's the most-

Anton Osika 47:32 That's a good quote.

Harry Stebbings 47:33 It's a really-

Anton Osika 47:34 I'll, I'll talk to you soon.

Harry Stebbings 47:34 ... good way to think about it.

Anton Osika 47:35 Yeah.

Harry Stebbings 47:35 Which, and I think it's the same, which is like, if we win, everyone will be happy.

Anton Osika 47:40 Mm-hmm.

Harry Stebbings 47:40 There's very few places where they're losing every day, day in, day out, and blissfully happy.

Anton Osika 47:45 Yes. (laughs)

Harry Stebbings 47:46 It doesn't happen.

Anton Osika 47:47 Okay, okay.

Harry Stebbings 47:47 What's not great about your culture today, if you could change it?

Anton Osika 47:51 So, th- there's a certain, uh, personality type that has, uh, takes a lot of, like, initiative, they're ex- very, very excited about new ideas and doing, like, novel things. And, um, as your company matures, you, like, that is an, is still an important ingredient, but you need the first priority to make what you have high quality, to continue to be high quality, and improve the quality across everything you're doing. Uh, and I want, um, us to be even more, like, memefying, "Let's improve the quality, let's improve how we do things, th- move slow, so that we can move really, really fast." And that that we put-

Harry Stebbings 48:32 Do you want to be more thoughtful around h- where you spend time and where you don't?

Anton Osika 48:35 Yeah, so there's this cowboy versus farmer, uh, analog, where a farmer is, like, optimizing things for the long term, and I think we can do a bit more of that optimizing things for the long term, uh, but, but we always strike the balance of doing, uh, innovating.

Harry Stebbings 48:49 Do you think you're in that phase of company build yet? Like, I, I actually...... prefer the optimize for the short term. I don't know if you spent much time, uh, in China or with, like, Chinese development teams, but they are unbelievable in their psychology around build. They optimize for the short term incessantly, and then just, like, sticky tape the shit, sticky tape the shit, sticky tape the shit.

Anton Osika 49:09 Yeah, yeah.

Harry Stebbings 49:09 And that is often how they're able to do so much so fast.

Anton Osika 49:12 Hm. Uh, I think if you have a super clear product market fit, you have a brand to defend, uh, y- yes, you cannot, you cannot sticky ta- (laughs)

Harry Stebbings 49:22 (laughs)

Anton Osika 49:22 ... everything. You can do that in, like, sprints to move f- to innovate, but you (laughs) really need to focus on, like, are these pieces put too well together? Spend a lot of time on moving things around in the organization, in your, in your, in your product so that it has high quality, maintains high quality, um, and you can build faster upon that foundation.

Harry Stebbings 49:42 Can you imagine if Apple were like, "Oh, fuck, we deleted your cloud, sorry"?

Anton Osika 49:46 Yeah, sounds very good. (laughs)

Harry Stebbings 49:47 (laughs)

Anton Osika 49:48 How bad.

Harry Stebbings 49:49 Uh, I said earlier about, like, Europe and summer and, you know, not moving fast enough. You've said before that it is better to build in Europe.

Anton Osika 49:55 Mm-hmm.

Harry Stebbings 49:56 And I don't want this to be, like, an advert for Europe, um, but why do you think it's better to build in Europe?

Anton Osika 50:03 I think there are very many, uh, good things about Europe. (laughs) There are also good things, uh, things that are better in, in the US, for example. Uh, I mostly think, uh, about I want to prove that you can build a generational product, a generational company team from Europe, and part of it is on, is on hard mode, uh, and I think-

Harry Stebbings 50:25 What parts are on hard mode?

Anton Osika 50:27 Yeah. O- So hard mode is the, the, where the network isn't as great-

Harry Stebbings 50:31 (laughs)

Anton Osika 50:31 ... in how many individuals and companies that have worked on and have context for, like, all the different stages of building an amazing multinational company.

Harry Stebbings 50:42 Completely agree. There are no Elena Vernas in Europe.

Anton Osika 50:44 Yeah.

Harry Stebbings 50:46 Or very few.

Anton Osika 50:46 Maybe we'll get her here soon.

Harry Stebbings 50:47 Yeah.

Anton Osika 50:47 But, um, I think, uh, that's hard mode. I think the, um, like, access to capital, people that will quickly give you a lot of distribution, help you with dis- distribution and brand, part of it is-

Harry Stebbings 51:00 Do you think access to capital is a genuine problem? I think there's so much money in Europe that actually it's a problem.

Anton Osika 51:05 No, for, as I said, it's not a bottleneck for us, no.

Harry Stebbings 51:06 Yeah.

Anton Osika 51:07 It's not. Um-

Harry Stebbings 51:07 And, and you're gonna start to see very soon, I'm sure you're probably already seeing it, but Lovable spinouts, where anyone who leaves Lovable will get a term sheet straight away.

Anton Osika 51:14 100% true. Um, I think, yeah, why, why it's better to build in Europe?

Harry Stebbings 51:21 Yeah. What-

Anton Osika 51:22 There's, so there's-

Harry Stebbings 51:23 Well, so the, the, so, so the hard-

Anton Osika 51:24 We're the best, we're the best talent, okay, yeah.

Harry Stebbings 51:25 So you said the hard things there is, like, number one, what was number one again, sorry? Wasn't access to-

Anton Osika 51:31 There's the network.

Harry Stebbings 51:31 And network.

Anton Osika 51:32 And then the individuals that have done it before.

Harry Stebbings 51:34 Yeah.

Anton Osika 51:34 Yeah.

Harry Stebbings 51:34 That's one thing.

Anton Osika 51:35 Yeah.

Harry Stebbings 51:35 Is that it?

Anton Osika 51:37 I, I think it's easier to get distribution to be on, like, the center world stage in San Francisco and New York, um, but we've been, been able to pull that off from Stockholm, which is, uh, good proof that you can do it from here. Um, and then in the US-

Harry Stebbings 51:50 Why do you think you've been able to do it? I, I have my theory on why I think you've been successful.

Anton Osika 51:56 I mean, I, I think it's about storytelling and, and sharing everything we're doing at the company, um, and empowering, uh, other people who are using Lovable, telling their stories, a bit of that. Um, and that's how we broke th- been breaking through. I think we understand that you should be building in public and share what you're doing.

Harry Stebbings 52:16 I think transparency is everything, and people like to follow people.

Anton Osika 52:20 Mm-hmm.

Harry Stebbings 52:21 And you've combined the two very well, which is you're incredibly transparent around your ARR growth. It's easy to be when it's as good as it is.

Anton Osika 52:28 (laughs)

Harry Stebbings 52:28 But you're incredibly transparent in a way that most people aren't, and then it's led by you and your voice. And the two combinations of, like, cult of personality, you and Anton, uh, you being Anton, not you in the third person, uh, and then you and ARR growth is what really drives the success-

Anton Osika 52:44 Mm.

Harry Stebbings 52:44 ... in that way. So that's kind of what's harder. What's better? Like, why, why should everyone build their company in Europe?

Anton Osika 52:51 Uh, I mean, we, we can be... We are the biggest talent magnet in Stockholm right now, which is amazing. Uh, you can't, it's much, much more difficult to be that in San Francisco or New York. So we can really pick up all the, um, underutilized talent and, uh, shape their, like, 10X their performance by being in a 10X, uh, better culture or ways of working and, like, with, with amazing colleagues. So being able to be that, like, top one is, I, I think the biggest one. Um, there's, uh, a culture of, like, humility and low ego and working really, really well together as a team that I think is, uh, stronger for in Europe and, like, this, uh, way of thinking in terms of s- efficiency and doing much, much more with less. Um, that's also stronger here.

Harry Stebbings 53:37 You have inherently higher churn in the Valley. When you have a bad day, OpenAI offer you a bigger package, and it's like, "Nah, I'll even do OpenAI." And that prevents the compounding of knowledge within teams, which I think is so valuable.

Anton Osika 53:49 Yeah.

Harry Stebbings 53:50 Totally agree with you there. Would Lovable be less successful if it were in the Valley?

Anton Osika 53:55 I honestly don't know. I think we'd be, (laughs) we'd be very successful regardless.

Harry Stebbings 53:59 Did you ever think about moving?

Anton Osika 54:00 Yes, uh, when I was about to start the company, um, I, e- everyone was of course telling me I should go to, uh, to SF, um, but we just kept building, and we got, found some great people in, in Stockholm, so we, we kept building it from, uh, from here. Uh, and I'm happy how it turned out.

Harry Stebbings 54:20 Dude, it's been amazing. That was a decision which you, you made and it worked out very well. What did you do in the Lovable journey that, with the benefit of hindsight and some experience, you wish you hadn't done?

Anton Osika 54:34 When we started, we had this, um, this i- The vision was very clear. The sequencing was not so clear in what, what we should be doing. We had, we had this open source community that was kind of excited about a tool I made a few months ago, before we started the company, GPT Engineer. I think we shouldn't have just, like, scrapped that completely and be 100% focused on what's, uh, what's the future look like in terms of building, opening your browser, just building your product there, which is Lovable, um-

Harry Stebbings 55:02 Why should you have scrapped that?

Anton Osika 55:04 You should be very, very focused on doing one thing.

Harry Stebbings 55:05 Was it not crucial for customer development, customer feedback?

Anton Osika 55:08 No, I don't, I don't think so. There, there was, of course, a, uh, plan for how to incorporate the two and get more value together, like, though, um, open source can be very succe- very useful for many businesses. But, um, in terms, in, in the, with the perspective of maximal focus, it, it was just a bad idea to do two things that were, like-... a bit, bit too tangentially, uh, related. So, so that's definitely one thing. Fo- and this, we talk a lot about, it's doing one thing, finding the bottleneck for the company, and solving for that bottleneck is the, the best way to move re- really, really fast.

Harry Stebbings 55:41 Wha- what is the bottleneck that you will be discussing in the board tomorrow?

Anton Osika 55:45 Um, I think the bottleneck for our long-term future is how we identify the, uh, technical product, like, so engineers that l- will take the product to its next phase and, um, innovate on many, many fronts at the same time. Um, that's in the long term. If you think about the product today, um, it's giving our AI more capabilities that are really p- like, make it a really polished user experience and give it more of those capabilities so that you can build out your full company, uh, grow your business on, on top of Lovable. And, um, then I think a bo- a bottleneck is for how we serve all this extreme amount of enterprise customers love and pull from them, uh, at the same time as we f- we focus, like, probably one is for the four founders building on Lovable.

Harry Stebbings 56:36 Will Lovable have an enterprise sales team?

Anton Osika 56:39 Yes.

Harry Stebbings 56:40 And become an enterprise company?

Anton Osika 56:42 It will not become an enterprise company, but it, but it will have a enterprise sales team.

Harry Stebbings 56:47 (laughs) Are you nervous about it?

Anton Osika 56:47 And, and what is that? Yeah, I'm not so nervous about it. Well, that is just about, uh, talking and understanding your customers, making sure they have the tools to get value from the products. That, that is what I, how I see that, that part. And there are, of course, many enterprises that are like top-down enterprise sales team that hustle themselves to wine and dine CEOs. I, I don't, that's not, uh, what we're going to do.

Harry Stebbings 57:08 What is the hardest role to hire for, for you?

Anton Osika 57:13 Um, so I think hiring engineering leaders is very difficult because it's so hard to predict h- how their past performance will translate to our organization.

Harry Stebbings 57:23 Have you made mistakes on hiring?

Anton Osika 57:25 Um, yes, I've made some mistakes, yeah.

Harry Stebbings 57:28 What did you do that you wish you hadn't done on hiring?

Anton Osika 57:31 I wish I was in the details when I... and not delegate, um, too much. And, and I, I wish I was more proactive about, um, like, does this person want to do, want to reach the outcomes? Are they excited and inherently motivated about the outcomes that this role is going to be, uh, where is it going to be most important for that role to reach those outcomes?

Harry Stebbings 57:52 I think something that's really interesting when you say about kind of leaders there is, I thought of, you know, your co-founder, actually, who was in the video for the OpenAI, uh, release.

Anton Osika 58:00 Yeah.

Harry Stebbings 58:00 And what struck me with that is, respectfully, it was one of the first times I've seen him front and center, not you.

Anton Osika 58:06 Mm.

Harry Stebbings 58:07 How do you think about exposure between the two of you, given you are very much the face of Lovable?

Anton Osika 58:14 I think, um, I'd love for Fabian to have more exposure, but I also want him to be focused on building the products (laughs) -

Harry Stebbings 58:22 (laughs)

Anton Osika 58:22 ... and that's what he's focused on. Uh, and it's probably, um, much easier for people to relate to Lovable if they see one person and keep seeing that one person, and to, to date that is me. And I think it will continue (laughs) to be me.

Harry Stebbings 58:37 What do you think is the biggest secret to a successful co-founding pair-

Anton Osika 58:42 Mm.

Harry Stebbings 58:42 ... scaling at the speed of Lovable scaling?

Anton Osika 58:45 I think the most important thing is just the, the real horsepower and adaptability of the founders. Um, and then, like, if those are maxed out or those are high, um, I mean, you must be able to, um, work together with... Like, if you have sufficiently low ego, it's going to work. But if, if you really want to work extremely well together, what's... I'll take an example, which is Fabian and me. He's, like, uh, not very, um, g- big on doing some weird new way of doing things. He's just, like, simplify it as much as possible. Um, he's quite introverted and quiet, uh, un- until he's, like, has really shaped an opinion about what's the most important thing. And I'm on the s- polarop- polar (laughs) side of the spectrum and saying, "Fabi, we should use this new crazy thing." And, and that, like, polarity is actually very productive for

Anton Osika 59:44 both of us.

Harry Stebbings 59:44 When you think about that and then you think about f- you're married and very happily married. When you think about successful marriages, what makes the marriage so successful?

Anton Osika 59:56 I mean, I, me and Fabian, we can talk about anything, uh, and, uh, that's extremely productive. And we can talk about turning things, uh, turning every stone and challenging each, each other, um, around anything. And we have a lot of, like, humility. I think that's very valuable and very important. And the same is true in my marriage. (laughs) .

Harry Stebbings 1:00:18 Humility.

Anton Osika 1:00:18 I think humility. Yes, a lot of humility. I see, I, I talk about my faults a lot in my marriage, for example. And, um-

Harry Stebbings 1:00:25 Does success-

Anton Osika 1:00:26 ... yeah, I try to work around them.

Harry Stebbings 1:00:26 ... make marriage harder or easier?

Anton Osika 1:00:29 I mean, if you have zero hours to spend time with your partner, it makes it more difficult. Um, I don't see...

Harry Stebbings 1:00:36 I also look at n-

Anton Osika 1:00:37 Yeah.

Harry Stebbings 1:00:37 ... 90% of relationships often struggle, and a lot of arguments is based on money.

Anton Osika 1:00:41 Mm.

Harry Stebbings 1:00:41 Which is a inevitable thing that's very hard, especially as cost of living goes up and for a lot of people.

Anton Osika 1:00:46 Mm.

Harry Stebbings 1:00:46 Um, that then doesn't become a problem.

Anton Osika 1:00:49 True. Yeah, I don't think... For us, it was never a problem.

Harry Stebbings 1:00:52 Has it changed your marriage?

Anton Osika 1:00:53 Um, not so much, no.

Harry Stebbings 1:00:57 Very humble Swedes, aren't you?

Anton Osika 1:00:58 Yes, we... (laughs)

Harry Stebbings 1:00:59 (laughs)

Anton Osika 1:01:00 I mean, I haven't changed my lifestyle since Lovable was be- successful.

Harry Stebbings 1:01:03 Have you not?

Anton Osika 1:01:05 Um, maybe I think less about m- monetary decisions. But no, lifestyle is pretty much the same.

Harry Stebbings 1:01:11 What does the Lovable product look like at the end of 2026?

Anton Osika 1:01:15 Uh, I mean, it's your perfect co-founder that you go to with your idea from the idea stage, but also all, all the way up to...... growing your business once you have customers, and, um, taking care of, like, what Elena is doing, optimizing the products for growth, optimizing the product and optimizing your, uh, communication with your customers, be it in- through email or through, um, different marketing channels.

Harry Stebbings 1:01:40 So you eat the whole stack, then?

Anton Osika 1:01:42 Yeah, it's one opinionated way to do the entire product lifecycle.

Harry Stebbings 1:01:45 So you do everything from email marketing, to SMS marketing-

Anton Osika 1:01:48 Yeah.

Harry Stebbings 1:01:48 ... and everything in between?

Anton Osika 1:01:49 Yes. And, I mean, obviously this is what an enterprise also wants to build their products on. But, uh, in the interim they're using it for, uh, individuals, like people in teams, building out ideas that the com- the enterprise companies should be doing. And they're doing that very, very productively.

Harry Stebbings 1:02:05 Is benchmarking for models bullshit? And evaluations bullshit? I had Edwin from Surge on the show.

Anton Osika 1:02:13 (laughs)

Harry Stebbings 1:02:13 Do you know Surge? It's like the Scale AI competitor, but it's-

Anton Osika 1:02:15 Mm-hmm.

Harry Stebbings 1:02:15 ... actually phenomenally successful. It's never raised a dollar, and it's a billion two in revenue.

Anton Osika 1:02:20 Whoa.

Harry Stebbings 1:02:20 It's f- unbelievable.

Anton Osika 1:02:22 Whoa.

Harry Stebbings 1:02:22 Um, and he was like, benchmark evaluations are bullshit.

Anton Osika 1:02:26 Mm-hmm. I mean, they, they turn, uh, more and more bullshit over time. There's this, something called Goodhart's law, so when you start optimizing for a number, that number stops being a good measure for success, m- even if it was a great number for measure, measure of success previously. So that obvious- that happens, like, with all benchmarks, over time, in some sense.

Harry Stebbings 1:02:45 What metric within Lovable means less over time?

Anton Osika 1:02:49 So, it, it means less if we start optimizing for it, right? Uh, and, like, I guess one example where it means less, I guess, is how many people click, uh, the thumbs up button on home messages, um, because then we can, like, I don't know, say, say fun jokes or whatever that for some reason just triggers people to be more, uh, literally just asking the human, "Click the button." (laughs) "Click the (laughs) button, button if you want to do it." And, and that's... Then we're hacking the metric, right? So, that's

Anton Osika 1:03:18 just one example.

Harry Stebbings 1:03:19 Joakim, we're gonna do a quick fire round.

Anton Osika 1:03:20 Okay.

Harry Stebbings 1:03:20 I'm gonna hit you with some incredibly unfair questions and you can give me your thoughts, okay? What wildly held belief about AI do you think is just very wrong?

Anton Osika 1:03:30 I think AI is smarter than humans, and most people don't agree.

Harry Stebbings 1:03:35 Do you not think they do now?

Anton Osika 1:03:37 I think most people don't agree. And, and the reason is that it's of- oftentimes it's very, very stupid. Um, but it's actually very stupid. But if you give it all the context, um, or you have like... you build a purposeful system for that, th- what they are stupid at, it's smarter than humans.

Harry Stebbings 1:03:52 Do you think we will see a plateauing or do you think we will see a continuous exponential progression curve?

Anton Osika 1:04:01 I think we'll see a plateauing, uh, on the things that we, uh, we care about, which is a lot of nuance and, like, being good at all the different things, uh, at once in a- in, in the same model. Um, there's probably some ways-

Harry Stebbings 1:04:15 But people are looking at GPT-5 now-

Anton Osika 1:04:17 Yeah.

Harry Stebbings 1:04:17 ... and saying, "Ah, we're hitting a stage where-"

Anton Osika 1:04:19 Yeah.

Harry Stebbings 1:04:19 "... actually improvements are much more incremental."

Anton Osika 1:04:21 Mm. I mean, I think we... Like, i- we, we... What we're s- what we've seen so, so far is, like, the sigmoid curves across many different dimensions at the same time, um, and, yeah, we're going to s- we're going to see a, a plateauing. There's some sigmoid curves where I, I think we're still in this, like, exponential phase of the sigmoid curve. So it, um... And those could be something like, um, science and engineering, and, like, bioengineering, where AI is just going to continue to, like, exponentially become extremely powerful and, uh, g- generate a lot of new medicines and new ways of treating health.

Harry Stebbings 1:05:00 Grok, Anthropic, OpenAI.

Anton Osika 1:05:03 Mm-hmm.

Harry Stebbings 1:05:04 You can invest in OpenAI at 380, Anthropic at 180, and Grok at, I think it's 100. Which one do you invest in and which one do you short?

Anton Osika 1:05:14 I'd invest in Grok and I would, um, short, uh... S- what was the numbers again? 380?

Harry Stebbings 1:05:22 380 and 180.

Anton Osika 1:05:23 Okay. And 100. I, I would probably short Anthropic because... No, I would, I would short OpenAI, let's say.

Harry Stebbings 1:05:33 Why? Why would you buy the, uh, Grok and short, um, OpenAI?

Anton Osika 1:05:37 I think it's more the slope on the Grok team. They have... They're doing something which I respect a lot, which is to hire missionaries for the data curation part, and they call it AI tutoring. Um, and they, um... I think the morale is much, much better in that team than both of the other teams. The morale is super high. Uh, OpenAI has gone through all this mess, right? Anthropic's pr- has good mor- morale as well. Um, and they're growing faster on the enterprise side from what I'm hearing.

Harry Stebbings 1:06:05 Do you think OpenAI wins the consumer, in terms like next generation Google, and Anthropic wins the developer and the enterprise?

Anton Osika 1:06:13 No, I think it's gonna be v- unknown. Uh, there's gonna be something else happening that we don't know what it is though.

Harry Stebbings 1:06:17 Do you think the, there will be a leading model that has not been created yet?

Anton Osika 1:06:22 Yes. From China.

Harry Stebbings 1:06:25 Are you... Do you worry about China?

Anton Osika 1:06:28 I think Chinese companies are not as good as either really understanding your users, so not very more worried. I do think there's, like, a 50/50 chance they will have the best model, we'll be using a Chinese model at some point, and that makes me a bit concerned be- because I, because we do-

Harry Stebbings 1:06:43 Would you use Chinese models at Lovable?

Anton Osika 1:06:45 If we would-

Harry Stebbings 1:06:46 Yeah.

Anton Osika 1:06:47 ... then I would looking, have to look into the details and see, like, what, what's bad about that? Do we give them data we don't want to give them? Um, but I, I mean, we just wanna do what's best for our customers. If that's going (laughs) for a Chinese model, then there's no negatives, yes.

Harry Stebbings 1:06:59 It's... I, I completely agree. I think also just, like, the multitude of models coming out of China is just terrifying. When you look at every week, there's, like, four new ones, and they're all as good as the last one.

Anton Osika 1:07:09 Yep.

Harry Stebbings 1:07:09 And the speed of distillation is just fucking insane. Are the models of the future open or closed? Like, which model wins?

Anton Osika 1:07:18 I think the best ones will always be closed. Um, but if you want maximal flexibility and r- some kind of open ecosystem around it, it mi- might be that the open ones are the ones that most people choose.

Harry Stebbings 1:07:30 You can have dinner with anyone, dead or alive.... who do you have dinner with and what do you ask them?

Anton Osika 1:07:37 I think I would have dinner with Newton, because he was, like, religious and super smart, and just, like, talk about how he was in his age and why he's religious.

Harry Stebbings 1:07:47 (laughs)

Anton Osika 1:07:47 He's super, like, he in- he invented so many different things, uh, and he's a, he's bit of a role model. And he's dead, so I can't meet him, un- e- unless I see him now.

Harry Stebbings 1:07:55 (laughs) Uh, yeah, sorry, I c- I can't help with that one. There's no intro there that would work.

Anton Osika 1:07:59 No? Fuck.

Harry Stebbings 1:07:59 Um, that's amazing. What AI company do not enough people pay attention to? Like I said, Surge, for me, is one, which is like Scale AI, but fundamentally a much better business. Barely anyone knows it, and it's ridiculous. Which company does no one pay attention to that everyone should pay attention to?

Anton Osika 1:08:17 I think the browser companies are interesting. So there is Strawberry, who's dotcom, there's the Dia and the Perplexity now. Uh, I'm, I'm very excited to see what happens for all of those companies.

Harry Stebbings 1:08:27 What do you think happens to Perplexity?

Anton Osika 1:08:29 What happens to them?

Harry Stebbings 1:08:30 Yeah.

Anton Osika 1:08:30 So, they're, they want to create their phone, I think, and I think that's a good bet.

Harry Stebbings 1:08:34 You think it's a good bet?

Anton Osika 1:08:34 Mm.

Harry Stebbings 1:08:36 Would you invest in them at 18 billion?

Anton Osika 1:08:38 18 billion? Um, (laughs) it depends on what options I have for ...

Harry Stebbings 1:08:45 (laughs)

Anton Osika 1:08:45 ... now.

Harry Stebbings 1:08:45 That's amazing. Your laugh there just kind of said it all.

Anton Osika 1:08:47 I, yeah. No.

Harry Stebbings 1:08:48 That's very funny. Who's been the single most instrumental person for Lovable not in the company?

Anton Osika 1:08:56 Oh, it's, it's Jenya at Accel. Uh, Je- Jenya, he ran s- uh, sales and like CEO at, uh, Miro, and he was on Dropbox at, at Ant Segment, and I get a lot of, uh, input that I have from him.

Harry Stebbings 1:09:09 What's been the biggest?

Anton Osika 1:09:11 I think, mm, just talking about, uh, like, he's just my coach, and I, I talk about like how I think about things, and then he asks me questions and, and tells me, "Uh, Anton, you have to step up in this, uh, i- i- like, build structure in this, the area."

Harry Stebbings 1:09:26 I love him. I had him on 20 Sales, and he was fantastic.

Anton Osika 1:09:29 Mm.

Harry Stebbings 1:09:29 What have you changed your mind on most? Penultimate one. So, I, I, I thought that you'd see the commoditization of model performance and models, and actually value accrual would be very difficult and there'd be a race to the bottom. I think that's clearly very wrong and very stupid of me (laughs) to have ever thought that, and there'll be very valuable model providers. What did you believe that you're like, "That was wrong."

Anton Osika 1:09:50 In the cont- in the, in the context of Lovable, um, I thought we should be building an agent, um, before, like, the models were ready for it, and because the models were s- starting to get optimized for an agentic system. And what I realized is that, no, no, no, you need to have a product that as many people as possible are using today so that you can optimize not necessarily the AI, but necessary optimize the entire user experience for those users and get ... That's, that's your data flywheel that you want to use.

Harry Stebbings 1:10:18 Do you worry about job displacement at scale in a 10-year time period?

Anton Osika 1:10:23 Y- so I worry about us humans globally not n- even understanding what we want to achieve on this planet. (laughs) And if, um, there's a lot of the rapid change with like white-collar workers being out of a job, and w- uh, humans, we get super worried and concerned and scared we're gonna ... All hell is going to break loose. (laughs) So that, that's what I'm worried about. But if we're a bit more thoughtful in terms of like, "Okay, if there would be insane amount of job displacement, this is kind of what we think we should do, and this is what we want to achieve, um, this is how we make sure people, like, have, can make some made-up job in the interim," um, then, then we would a 100% solve that.

Harry Stebbings 1:11:04 Eight out of the top 10 paying jobs today did not exist 15 years ago. I always think that's an interesting stat.

Anton Osika 1:11:10 Mm.

Harry Stebbings 1:11:11 And we always overestimate job displacement with new technologies.

Anton Osika 1:11:14 Yeah.

Harry Stebbings 1:11:15 Interesting one.

Anton Osika 1:11:16 I th- I think we're going to have, like, maybe a shift away from some very glamorous jobs, which people will be get depressed by. Like, similarly to how being like an artist was like so fucking cool, but clearly you can't make any money (laughs) as an artist.

Harry Stebbings 1:11:30 (laughs)

Anton Osika 1:11:30 And so we're gonna see that again now for a lot of knowledge work, and I think that's gonna be fine.

Harry Stebbings 1:11:33 I, my, my, my kids? Brain surgeons. Brain surgeons. I'm n- no AI is gonna come for brain surgeons for years. Come on.

Anton Osika 1:11:39 Maybe.

Harry Stebbings 1:11:40 No, yeah.

Anton Osika 1:11:40 Have you seen the robots, though, the b- surgeon robots? They're pretty good.

Harry Stebbings 1:11:43 I'm gonna be honest. You're not having people be like, "I'm gonna choose the robot version of that."

Anton Osika 1:11:48 Right. Yeah.

Harry Stebbings 1:11:48 Uh, not for, for a while. Um-

Anton Osika 1:11:50 They do a good job.

Harry Stebbings 1:11:51 Is there anything else that concerns you with AI when you look forward?

Anton Osika 1:11:55 I think a- a- again, us humans, we're very good at competing. (laughs) And in some cas- I mean, in many cases, that's amazing. Like, that's how we get all the best (laughs) companies. That's how we get great technology. Uh, but i- in some cases, like, we're b- we're competing, and then we go to war with each other, or, like, we, um, like, start preparing for wars. And I think if we can be better at, uh, thinking big picture across, across, uh, superpowers, uh, that would prevent the scenario where you have, like, a AI that can kill all people in the others, r- like, in the other nation, uh, i- uh, in an instant, um, and, like, that being triggered without us actually wanting that to happen. So, um, yeah, I'm, I'm concerned that u- us being so competitive, um, in a world where things happen much faster is going to lead to some unexpected results that no one really wants.

Harry Stebbings 1:12:52 Which competitor do you most respect?

Anton Osika 1:12:54 I think OpenAI is pretty good at building products. I think there's cons- other foundation model labs that will do even better at building products, and those are the ones we should, uh, think about for the future, um, but mainly, mainly focus on just what do our users want? How do we make a better product for them?

Harry Stebbings 1:13:07 But you must look across your, your Figmas, your Bolts-

Anton Osika 1:13:11 Yeah.

Harry Stebbings 1:13:11 ... your Replets, your ... Who do you respect most?

Anton Osika 1:13:13 I respect Figma, yeah.

Harry Stebbings 1:13:15 Figma?

Anton Osika 1:13:15 Yeah.

Harry Stebbings 1:13:16 Why?

Anton Osika 1:13:17 They, yeah, because they're good at, uh, listening to, to their users and building good products. And if they can translate that to, uh, the full product life cycle, um, I think they're a very formidable competitor.

Harry Stebbings 1:13:29 Everything goes to plan, okay? We hit all of our numbers, and everything works. If that is the case, where then is Lovable in 2030, five years' time?

Anton Osika 1:13:41 W- we're the int- the mostly used interface for humans to AI, and that's a very huge market.

Harry Stebbings 1:13:50 Dude, it's so much better doing it in person. I've so enjoyed this. Thank you so much for agreeing to do it in person, and I've loved it, man.

Anton Osika 1:13:57 It was fun.