Building Lovable With Anton Osika: The Power Of Simplicity, AI As A Technical Co-Founder, And Why 'Vibe Coding' Needs A New Name

0:00Anton Osika:I'm an engineer, of course, so there's a lot that is currently used by everyone, but the focus is that it has to be so simple that anyone can use it.

0:09Tomer Cohen:Do you see Coders also using Lovable in the future?

0:13Anton Osika:I think you should see it as an Apple device. For Lovable, it just works. It starts very simple, which makes sure you can focus on what's most important, people who are engineers. The level makes their work so much faster. Creating new products, creating new websites, trying out different tools.

0:35Tomer Cohen:There's a moment every software builder remembers. It's the first time we made a computer do something. Now, for many of us, that moment was the hello world moment. We wrote one line of code and suddenly the computer talked back. It was a very short moment, but it flipped a switch in our brain. Now we felt we could build. And that first moment is getting dramatically upgraded right now. With one line of text, you can ship a working product, a game, a website, even a full e-commerce service. And here's the real shift. Today, less than 1 % of the world can code. So the next wave is about empowering the other 99 % to build as well.

1:19Tomer Cohen:One of the most prominent companies pushing this future forward is Lavable. And today I'm joined by Anton Osika, the co-founder and CEO of Lavable. Now, people call this vibe coding, but Anton's framing is much more direct. This isn't about vibes. It's about leverage. It's about speed. It's about who gets to build. In this conversation, we get into the principles behind Lavable's product choices and why simplicity is the hardest decision to defend, what the next magic moment could look like as building becomes more accessible to more people. The real trade-offs behind AI-driven software creation, especially the chasm from prototype to production, and why Anton believes the most underrated mode in AI is trust, manifested as brand love.

2:05Tomer Cohen:So let's get into it.

2:12Tomer Cohen:Anton, it's great to have you on the show again. And thank you so much for joining me. I always love our conversations. Fun tidbit, by the way. This is our third Swedish-originated company on the podcast. The other ones were IKEA and Spotify. I don't have a bias for Sweden, but apparently maybe there's a subconscious bias. But those are household names. I'm sure Lovable will be one day a household name. So it's pretty cool. It's cool to have you here.

2:37Anton Osika:Thanks for having me. We are all thinking about how do we make our product as good as possible and create a team that can facilitate creating that product. And then what I think that's, by having that focus, we expect people to be talking about Lovable, our grandchildren to be using Lovable in 40 years into the future.

2:58Tomer Cohen:I love the idea of grandkids. Before we go to Lovable, I wanted to start with you. So your kind of personal journey. You actually started as a researcher. You learned math and physics before you became a builder. and as a builder you focused really in startups. You were a founding engineer in a startup, then you became a CTO and a co-founder. With Lavable you're the CEO and co-founder. What was the journey there, this move from research into building and specifically going into startups in an area like Sweden and Stockholm?

3:29Anton Osika:Yeah, I heard that you were also building computers when you were a kid. I was. And I was the same. I had a bit of that hacker, take things apart, approach to everything. And doing physics is a bit similar. You learn this way of thinking where you break everything down into the fundamental pieces of how things work. And then you have a set of hypotheses, or you create hypotheses to be able to explain observations. And you test those hypotheses in the real world, and you're very open to being wrong all the time. I was at CERN doing particle physics with a particle accelerator there. And one of the realizations for me was that I can have more impact working out in the real world, taking ideas to change real people's lives.

4:19Anton Osika:So I'm very happy about that insight. And the other one that's big as an experience is that I've always been around people who are creative. Like my mom has a lot of great ideas, designers, people in products specifically, who can clearly see problems in the real world and want to make those ideas into reality. They haven't been able to do so because of how difficult it is technically. And I always felt like this feels artificial, it should be easier. And that frustration and wanting to solve that problem for all these people is also a big part of how I got to starting Lovable with the deliberate idea from the get-go to build a product that just works for people who don't know, they don't have a background in coding or building software, which is a very, very rare skill set to be able to do that end-to-end and build full products.

5:17Tomer Cohen:How would you define what Lovable is for people?

5:20Anton Osika:So Lovable lets you take any idea and make that into reality as working software. And it's used by a lot of different type of people. We have a lot of people in product that are now not held back by engineering backlogs. They can just build out full prototypes or even ship in some cases straight to production and be able to move things around. We're also seeing a lot of founders that are building for the first time and they can have Lovable literally as their technical co-founder that does product building and in some cases a lot more things. And we're seeing people in operations, finance, marketing, be able to not be held back by engineering backlogs, but just align what to build faster, validate and ship all the way to the production.

6:08Tomer Cohen:Love it. So instead of companies, you're seeing all those functions use it really well in founders as well. in the industries that you see it gravitate towards people are using it in certain industries?

6:17Anton Osika:I mean it's used across all types of companies. It's Fortune 100 companies, it's just some tiny small companies, mom and pop shop that are creating their websites with Lovable and adding an AI chatbot to the website to be able to answer questions. And I always, every single week, I'm stunned by the problems that people have been solving using our tool.

6:39Tomer Cohen:I'm curious for you when you think about your time as a researcher, your time as an early founding engineer and then multiple kind of founding experiences, what principles you got from that that really made lovable what it is today?

6:56Anton Osika:I think about my role is to create an environment where people who have good judgment and like to think end to end of solving problems can just go out and execute. So that's what I learned from all the startups that I've been a part of building. And then the other thing is about testing ideas very quickly while still being understanding that you should, there's a few core things that you should really, really focus on to get to not just do them 80-20 and then leave it a bit tough. But keep focusing until you really make things really good. and there's an expression in product development that is like building an MVP is not enough.

7:43Anton Osika:You have to build a minimum lovable version of that product to actually really get something that you should be launching to customers for real.

7:53Tomer Cohen:Love it. I think about the early days of, it used to be about, you know, the first hello world kind of thing when you write your first program and the computer responds with hello world and now you can write in a sentence but you want to build and Lovable would create it for you and you have this like, oh my god, that was unbelievable. Where do you see that in three, five years?

8:13Anton Osika:I'm learning what are many different wow moments of using Lovable from our users. I was at the dinner yesterday and people had been building AI applications with their kids and that was like wow moment for both the kids and the parents. I heard about a founder that built that likes their pets and they have been starting to AI generate and deliver physical frames of their dogs to people around the world and we picked up this fact that they were making$100 ,000 per month on this apparently it's$300 ,000 per month so they're all of these founder stories where I'm getting aha moments hearing from from the customers

8:55Tomer Cohen:Maybe to your point, it's not just about like, oh, this thing can build. It's like, I can create a business.

9:01Anton Osika:There's a new economy where there's thousands, tens of thousands of founders that wouldn't be able to start a business before. And now they are using Lovable as their technical co-founder and succeeding as being an entrepreneur. Some of them make it to be financially independent on that. The side hustles, regardless, are helping them learn how to solve problems, the build software, but all the problems that you need to solve to make a business successful.

9:28Tomer Cohen:I love that framing. In many ways, people think of this as like an app development capability, but really it's an economic tool. I mean, it's evolving everyone's role at work now, where anyone can actually be someone that builds high quality software if they put their mind to it.

9:46Anton Osika:And then what that results in is that if you think end to end and you think in terms of how does this fit into everything else, then you can just absolutely shape the trajectory of the company you're working at. I'm hearing people getting promoted because they were able to re-architect their entire marketing website and be able to have their organization move on cycle lands that used to be weeks to just hours.

10:11Tomer Cohen:I've heard you talk about when like you know today less than 1 % of the world's population can code and it's really about enabling the rest 99 % to be able to have to tap into software tools but really it's economic tools that can help them build their business. Is that the audience you're thinking about like top of mind was that from the beginning the hypothesis around like how do I help designer product managers but then people who are completely non-technical to go and build?

10:37Anton Osika:Yeah, that was the idea. So there's 30 million developers globally, and there's even fewer that can kind of take an idea and build it end to end, the full products. We from the get-go decided we're going to build something that has to be so simple that anyone can understand it and that it just works. And that has been a big advantage for Lovable and our users, that it seems extremely simple. We simplify something that's very complex and then if you are a technical developer you can dig into the code and change the code if you want to. Do you see coders also using Lovable

11:13Tomer Cohen:in the future or this is not meant for them? Like you're building directly for the non-technical ones. They don't want to see files, although there is a way to see files.

11:21Anton Osika:I think you should see it as an Apple device where for Lovable it just works. It starts very simple which makes sure you can focus on what's most important, people who are engineers, the level makes their work so much faster in creating new products and creating new websites, trying out different tools.

11:45Tomer Cohen:They gave Apple as an example, and Apple makes, they're very opinionated about how the product should work. You know, versus for example, Android will give you all the options or many of the options and then they kind of push the complexity to the user to decide whether there's more So you can decide where you want your product to be. You can be, I'm going to make it super simple, but I'm very opinionated on the back end. So I'm going to make decisions for you, or I'm going to give you the freedom to choose how you want to use the tool, but then it might be slightly more complex to use it. And then I think Lovable is a little bit more if there's like a middle, more towards the opinionated.

12:21Tomer Cohen:But obviously that made the product very, very simple. But it's going to work really well for non-technical folks. But if you are very technical, maybe more complex versus other tools.

12:29Anton Osika:Yeah. In fact, the reason we've been successful as well is that we have previous CTOs that are thinking a lot about what is the right technology to use for building applications, for Lovable to build applications for a customer. And somehow, sometimes it depends. But there are a set of choices that work really well together, which in turn makes Lovable more reliable than any other way to build web applications. They ensure that the application can scale and we have encoded how do we ensure that there are security checks, security is built in from the start in a way where if you just go to a chat GPT or just talking to an AI, it will take decisions that oftentimes doesn't give the same coherent experience in the end.

13:16Anton Osika:Because the pieces that it evolves to with some technology architecture that hasn't been and as swapped out from the start as if you're using Navable.

13:26Tomer Cohen:Any example that comes to mind to you when there was like a, you know, there was a discussion around the team around should we expose this to the user or not?

13:35Anton Osika:Yes, definitely. One thing is like, should you be able to choose if you want a fast response or a slow response? Should you be able to choose between different AI models and the different capabilities like that? And we fundamentally think that it's on our side to choose what's best for users at large. Because when we simplify it that much, we can run much more experiments where we are switching out the AI, and we're making the instructive to be faster or slower, and we see what users prefer. And we're also making it more adaptive.

14:10Tomer Cohen:There's a lot of Vibe coding tools out there today. and everybody chooses a path of where they want to excel. And it sounds like for Lovable, it's like, no, no, I am trying to make the simplest product possible for the end user.

14:25Anton Osika:Yeah, it's an interesting graph to work on where we are fundamentally adding a lot more capabilities all the time. Like adding ways for the agent to be sending emails, payments, creating agents with Lovable. And we have to make sure that at every step, the amount of complexity for the user preferably remains the same or even reduces. And it's a very interesting balance.

14:51Tomer Cohen:Any kind of one decision you made once that really upset power users to a point and you had to rethink again?

15:00Anton Osika:We've taken some features that were for more power users and put them... Before we had more of a formalized launch process, we decided, No, let's take this video and put it under the research, the labs toggles where you have to toggle it manually. I don't think we've had to make such a hard decision. We are careful to make sure our power users can kind of keep the functionality that we're simplifying if they really want to. And that's also an interesting trade off.

15:32Tomer Cohen:I'm curious if we kind of zoom out to the whole category of kind of vibe coding. Do you like that word, by the way, vibe coding? Does it resonate with you or do you think that's...

15:41Anton Osika:I think vibe coding is polarizing in the sense that it sounds like it's very casual. Yeah. It doesn't create high quality. We're seeing companies like Xperiality has 80 ,000 real estate agents globally. And there are dozens of countries. The marketing leader there, VP Marketing.