No-Code App Building, AGI, Solo Founders, and the Future of Software

Welcome to the Next Wave podcast. I'm Matt Wolf and today's guest is building what he calls the last piece of software the world will ever need. Anton OCK is the CEO of Lovable, a platform that lets anyone build fully working software just by describing what they want. No code, no dev team, just your ideas and AI. In this episode, we talk about what that means for the future of software development, entrepreneurship, and even AGI. Anton shows off a live demo of Lovable, shares how kids and solo founders are launching real businesses in just hours, and breaks down how this tech could level the playing field for creators everywhere. If you're building anything, tools, products, businesses, this conversation might just change how you think about the whole process. So, let's dive in with Anton from Lovable. Hey, Anton. Thanks for joining us on the show. How you doing today? It's great to see you, Matt. I'm good. Yeah. Awesome. Well, I don't want to waste your time. Let's just jump right into it and let's talk about Lovable. So, I'm curious a little bit about your backstory. What were you doing before Lovable? How did Lovable come about? Why did you decide to build it? Let's let's get the background a little bit. Yeah. Yeah, sure. I can I can go back to my childhood. No, but I decided to study at university with like as my everyone did then. Um and then I thought computer science. No, I'm going to go into physics because that was where what all the people who became the most generalist both in academia and in the industry studied here in Stockholm, Sweden. What I realized I was at this place where you know they discovered the Higs boss on the particle accelerator in CERN for 3 months. I was like amazing. There's 10,000 super smart people here, but they're trying to solve a problem that is very hard to solve. It's like very inelastic. You don't have any real world impact. And I'm obsessed about impact and making things happen. So I then I understood I'm not going to be staying in this on this track in academia. Uh so I went into building things in the industry and for the last 10 years, I've been building AI products and I've been specifically building creative teams that build AI products together and then um a bit more than one and a half years ago I decided to start a new company which is what I'm building now. You said at one point that you're trying to build the last piece of software. What does that mean? Yeah, sure. I can take you to the I like what spurred us starting the company Lovable and look this was like early 2023 that it was clear to me like this next generation of AI can actually start reason and it's specifically good at writing code. Um, if you put it into an advanced system of where like the reasoning engine is used to take this decision on behalf of a human, then you're going to be able to build a completely new type of interface. Mhm. To build software products and this AI is going to help developers become more productive. But the more interesting unlock here is for the 99% who never learned how to code. So uh this new interface that is like you talk to an AI and it builds your product you build it together with AI um it's going to let the 99% go from zero to one and enable anyone to unlock their creativity to build great companies to just create things great software and to build businesses on top of it. So that's what we're set out to do. And the last piece of software is this like it's a platform to build software products and it's going to make sure that you don't need to write humans don't need to write code anymore if they don't want to. Gotcha. So what I mean with that in mind, what do you think the role of like a software developer or a software engineer? What does that look like in the future? I talk to people who ask me this like Anton, what should I do? like I'm I'm an engineer and um I think engineers should just always put on the hat of and seeing themselves as the person who translates a real world problem into a technical solution and that means different things depending on what type of engineer you are and it changes over time what you do and using AI is of course going to be a larger part of that translation now I think what happens when you have AI that makes it faster to create software is that there's going to be just much much more software and there's going to be more iteration cycles to make each piece of software very very good and like the jargon for that in some tech companies is make them lovable. So I think that's the actually the end outcome of this like lowering the barriers through AI AI and new platforms like ours making it very very easy to just take an idea write it in and you get a full working product. Right. Right. I almost see it as like, you know, like with a symphony, right? It's almost like taking the people that are playing the instruments and moving them to the conductor. Right? Now, everybody can be the conductor and you're all sort of you're telling the various instruments what to do now. Instead of actually being the player of the instrument, you get to become the conductor. Yeah. It's very fun. It hooks people. Yeah. And and one thing that I've loved about this sort of new era of like AI software development is I don't need to necessarily build a SAS product that I'm going out and trying to like raise capital on or um you know sell it on a monthly fee or anything like that. I can find like little tiny bottlenecks in my business, little like holes of things that I'm like all right that is kind of a pain in the butt. I don't like doing that every day. Let me build a little software for myself that just fixes that for me. I can build it for myself and not have to worry about like trying to build a business around it. And I think that is one of the coolest things that software like this enables in my opinion. Yeah. The personal software trend is also very big. It's getting larger. Yeah. Yeah. So, let's go ahead and and like maybe jump into Lovable and give a a quick demo, show people what it's sort of capable of. I know you have a project that you kind of already started working on that we can jump in and tweak with and yeah sure I prepared a project right right ahead of a call what I wanted to have in that call I was going to be asked questions and you can you can use it to ask me questions is like a webinar Q&A app so anyone can answer input a question and I'll just focus first on what happened as I built this out so I basically went to lovable which looks like this and I put in the first prompt which was mockup webinar quest submit question app so that then I didn't want it to be a fully working product where it works across devices just a mockup and then it's lovable went ahead and said like okay hey I'm going to build this for you I'm going to show choose some simple design so so then I what what happened was I got this mock mockup UI you can see here um but it doesn't sync across devices so if someone opens this website in one place and enters a question I don't did. So I asked I just asked AI, "How do I do that?" And that's a big part of like way of working with AI is you, you know, if you don't understand, you can just ask. Okay, let's be honest. Your AI prompts aren't giving you the results you deserve. But with a little coaching, you can transform from basic prompts to engineering conversations that get you exactly what you want from ChatGpt. That's what this playbook delivers. Not just random prompts, but a step-by-step system with the exact techniques top AI professionals use every day. You'll have your own personalized prompt engineering system that delivers consistent results. Get it right now. Scan the QR code or click the link in the description below. Now, let's get back to the show. Yeah. One thing that I like about what you're showing here, too, is that it actually recommended Superbase, right? like if if you're trying to develop a software and you don't need necessarily know much about software, you might not know that you need this to be connected to a back-end database. So it's it's cool to me that it's going, hey, we could build this for you, but you also needed a database. Here's what we recommend. Yeah. No, this is like this is a native integration at at this point because most startups and like simple projects that are successful, they they they start on superbase. So So it's a very popular choice. What they did then is they told me like okay yeah you need to just connect superbase and then what I did I went up here and said connect to a new project now its connected now you can go ahead and add AI functionality login or just store data so I said up add real time sync as it explained and then it says I'm going to create the data table like the place to store the the questions um and I'll change some in the UI to handle it to connect to the database and then I pressed uh I had to approve move like okay run this code I got an error and then I said okay I fixed the error and then it's it worked and so so that's what happened when I built this and if I lost open this application or I can send it to you as well it's um it's Q&A.lovable.app app. If you go in here and you answer, you ask me a question, it should synchronize uh real time across any number of devices, which is which is a really like useful simple tool and you like you don't I don't have to no one has to log in. It's just like you go to the you scan the QR code, I ask ask for a QR code afterwards and then you're my like my I can always spin up I can always pick up this app if I want to have a like a Q&A session with a company or with someone I'm presenting to. Let's see. What is the coolest app you've seen built with lovable? Let's see. Submit question. So, I actually typed that on my screen, but you're seeing it on Anton's screen if you're watching the video. Yeah. Awesome. Uh, so that's a good question. Was the coolest app I've seen built. I think I saw like I saw a better version of SH GPT, which I like. like it had more keyboard shortcuts and a way to like customize each thread. I really like that um because a lot of the innovation right now actually happens on the AI intersection the user interface like the user experience side and that was one of the really cool apps I saw. Um it's also fun to see like people are launching we have we built a level app for people to launch their what what things they build and it's almost got like a Yeah, exactly. Exactly. And you have like for for projects that have been around, there's lots of people getting users through this. And um I haven't seen all of them, but there's so many cool things that people build. Uh that that was just like me demoing how Lovable works. What I what I could do next is to just show you like how how the AI handles a change. If if I want something to happen instantly because the AI is not instant. I I could could show you that like something you can do here is that you can edit uh text and style by just selecting it similar to website editor. Um so you don't need to wait for make little sort of Yeah. For small changes. Exactly. Exactly. Yeah. Uh but but let me first um ask you something. Do you have any style you you really like to see this in? Um, I mean I typically like my websites in dark mode and I always like to use like blues and purples like like my background like those are the colors I like to use. Um, so we'll say what can spell uh and we'll say something about like and cool hacker font and just make it look better. What I sometimes do is I just attach a screenshot to the AI pasted it in here to for it to see like how does it look now? Let's let's make it look cool. More beautiful. But but this is just how how Lovable works. Um you can um do some more things like you can customize knowledge if you have if you wanted to remember like always use this way of uh connecting to one to an API that you wanted to use or on some type of integration. Um if you want to edit the code as an if you want your engineer colleague to edit it you it's it's all kept two ways synchronized for them to go to GitHub which is like how engineers build software to date um and you can invite collaborators so I could I could send you this link and then you'll be able to edit this project if you want to. That's most of it. We we're waiting for the AI to run the change here. And the the last part that a lot of people are wishing to do here now is to add custom domain which is you might want to host you you might want to buy just a website domain for your project and I think you can even do that inside of you can even do that inside of lovable now. So does lovable host the whole thing and then you're just sort of pointing your your domain name to lovable. Yeah I can I can just select the domain here. I don't think this domain is free. I'll pay inside of this flow and then um it's all hosted like we're using state-of-the-art hosting infrastructure all with with one click. Yeah. But I imagine if somebody did want to export all the code and bring it over to their own host or whatever, they could do that as well. Yeah. So it's all super flexible. You can do anything that a human engineer would like to do with this setup. Here's our new style. I I'm not sure I love it, but uh Hey, but it did what you asked. Yes, that's true. So, one of the one of the things like when it comes to, you know, using AI for code that I've ran into a few times is like I'll have it build something and then there'll be a bug and then I'll say, "Hey, this bug is popping up. Can you fix it?" It'll fix that bug and then maybe introduce a new bug or it'll keep on like having that same bug over and over and over again. I know a lot of the LLMs have gotten better and better and better over time. We've now got Cloud 4, we've got Gemini 2.5 Pro. A lot of these LLMs have gotten a lot better at coding. Um, but I'm curious like how does how does Lovable specifically like maybe help overcome some of that kind of frustration that like the VI coders might have? Yeah. So, lovable is not just a call to cloud uh like the new cloud model. It does like an agentic chain which is that it tries to understand okay what's the context here like exactly what information is most relevant to solve this specific uh problem that you're having. If you're seeing a repeated bug like that's some one type of situation and then we're applying uh best practices that we've going to iterated ourselves towards to solve that specific type of context that you're in. you're stuck with the same type of bug. And we feed in some of those best practices for that are like adopted to work for the specific technology stack that lovable applications are built on. Uh so that that's what we do to date. But um um and and another important thing is that we have we give access to uh like what human engineers use to debug which is the AI is able to um read all the error messages or like all the cons the logs or that uh are um created as the user interacts with the website. So so that's fed into the AI system. So then you can see it kind of like if there's a bug it can really get much more of a picture of like what actually happened here and then use that in terms of figuring out the error that's that's what takes most time for most software engineers to understand what is it exactly that goes wrong. Yeah it doesn't work is not like isn't is not sufficient information to fix it. So so those are some of the things we're doing and we're working on a lot more. Very cool. So this is this is sort of getting more into the like theoretical philosophical kind of area, but I'm curious like when everybody has access to be able to develop any software, how do companies like actually build a mode? Like have you thought about that at all? Like how how would a software company actually build something if like anybody else can just go and use a tool to build the same thing? Like how how do how do businesses get formed around this kind of thing? I mean, you don't need a moat to build a great business. I I don't think so. Uh and um I mean, most of the modes are the same. I think there is a bit of a mode in terms of just trust and knowing who's behind this that this is like these people who built this have my best intentions at heart, right? And and that that will always remain. Um then there are like network effects modes. There are there's an like maybe at some point there's an economy of scale as well like that you can make it give a better value proposition because you're larger that hasn't shown to be so useful in in software businesses but maybe that's going to be the case in the future as well. Yeah. Yeah. Yeah. Um so I want to I want to shift the conversation quickly to the topic of AGI because I know in the past you've also mentioned that you want to help contribute to getting to AGI. A, how would you define AGI? Because I feel like obviously everybody kind of has a slightly different definition. And then the followup to that is like how does lovable or what you're trying to build sort of get us closer to that? Yeah, my favorite definition is that at the point when um anything you could hire a human remote worker for can be done with AI, um that's when you have AGI and it hedges you against that humans can do some things that only humans can because we humans don't want to talk to a machine. We want to talk to a human. But if it's a remote like a remote and over Slack, I think that then it's only cognitive labor. It's a pure cognitive labor. But um how do we get there? How building uh systems that execute write and execute code is a huge part of this. And increasingly what I'm thinking is that we are not going to work on the like foundation model layer. We're we're creating the most delightful and intuitive interface to interface with this type with the technology. And now it's for you know spinning up software and that's hosted available for anyone in the in the future. We're naturally adding a lot more types of interactions for in into our interface like oh browse the web for me and find all this these interface these things and then put them into a website or even like okay can you check for all the feature requests from my email and then implement some of them like that's the type of direction we would go for on the very long term and uh that type of interface is going to be one of the like import most important things in how humans perceive the level that AI has reached, right? Like what it's a a good friend of mine told me way way back 10 years ago, the AI is never better than the UI. So if you can't as a human get value from it, then it's worth worthless, right? Yeah. Do you do you think that like the the future of UI, the future of user interface is going to be what it is now where people are sort of typing prompts and we have these sort of visual user interfaces or do you think it's going to switch to some alternate modality? We don't know what what it's going to be yet. Uh it's going to be a combination of things like us humans are really good at um getting a lot of information visually. Uh so that's going to be a continue to be a big part and like we're not as good as getting a lot of information in by reading text I think as as like just looking at a picture and boom you get a lot of information really fast, right? So that's that's going to be a part of it. And then in how like how you as a human communicate as much information as possible to an AI which is going to be important as well. I mean at some point I do think we're going to see more and more adoption of like brain computer interfaces but just speaking or lip lip reading or might be like a emerging pattern UX wise with AI. Yeah. Yeah. and and and I you know I had some chats with people over at like Microsoft and Google and and that sort of thing. And sort of their position is that they want AI to be much more predictive, right? Like it knows what you want to do before you ask it to do the thing, right? It's going to it's going to get to this point where it starts to understand you. It understands your patterns. It understands what you do on a daily basis, anticipates it, and then just gets ahead of you on it. So, I don't know. To me, that's like a real sort of fascinating future that that we're heading into with AI. 100%. Yeah, that's huge. We spoke to one of our users recently, Philippe. I think it was a very fun story where where he had built large companies before. He raised $50 million, hired 130 engineers, and he's now he's past that, and he's just building a business himself with uh with using Lovable. And then he can like take all the ideas instantly and it moves so much faster which is a bit paradoxical to having this large organization where there are many changes of communication and it sounds very productive to have 130 engineers right but uh he's making tens of thousands of dollars on this like this small new business that he's growing organically um this is like the stereotypical AI native founder that I think is going to be we're going to hear a lot more from in how one people one person can build much faster than larger companies and you doing where they let AI do more and more on just the building part but also the marketing side and all of that is going to be like one human and a lot of AI systems so that that's what we're seeing but what also inspires me a lot is that kids you love to use lovable because they are super creative right they love to create things and then I've seen many like 14-year-olds and even younger who post their uh like they're selling something online or they're uh offer their services to walk the dogs with a website built on lovable and increasingly now since we launch we launched the teams plan recently lovable is getting huge in larger companies like fortune 500 companies where individuals in teams they accelerate how they come the team and the entire company takes the takes decisions and both in the terms of like okay we should really build this thing look I've already built it it's working like and then then you look in engineering and for building tools that just accelerates like finance the marketing building landing pages and all of that. So it's found that it's a tool that's being used for so many different different things and we're just keeping up on our side. Yeah, I think um you know last year at some point Sam Alman from OpenAI mentioned that he thinks within the next couple years we're going to see the first billion dollar company built by one person, right? And then Daario just from Anthropic just said it again like two weeks ago that he thinks within the next uh within 2026 we'll probably see the first one-person billion dollar company which just absolutely wild. Also mentioning the you mentioned kids are building apps. I actually had a conversation with uh Kevin Scott, the CTO of Microsoft, and he told this whole story about how his daughter um built an entire app for her school, and he was frustrated because she didn't even consult with him, and he's a software developer. She just went and built it herself. I mean, what what you're saying, we're definitely seeing more and more of. Yeah, it's fun that kids are going to be of course better than like older people generally to use AI and it's going to be such a big difference in how productive you are if you're good at using AI. Yeah. Yeah. So, I have I have one like small rabbit hole I want to go down with you really really quickly. I'm curious how the developer community as a whole has received something like lovable because I know you know some developers probably absolutely love it. it speeds up their time, but then there's also that sort of existential fear that their skill that they've been building is no longer needed. So, how has it been? How has the reception been so far? Um if you just look at the product uh and the features a a lot of developers love that you just um create a fully working application and then if you want to go in and customize use your your normal ID you you just sync sync it with like sync your codebase and then you're getting more done and you're shipping more value to your to your customer your employer um and that's like a very positive reception generally if you if you do zoom out and you're like oh but wait where is this actually headed um then I it is the case that people are like wait um what's my role in this in all of these but I think it's not so different from u anyone working in as a like white collar jobs that right this everything is going to be easier and easier to automate if you if you're on top and master these tools you're going to be much much more valuable in the in the workplace but otherwise like you're not going to maybe have as an as cushy job as a software engineer has been, but you're going to have to combine that with like doing sales or doing um like maybe doing something more manual as well that uh there there's going to be a potentially in the long term uh reduction in how many people sit and uh build software. Yeah. No, I couldn't agree more. And you know, like you mentioned to the a lot of the white collar work is in that same boat, right? Like if if your job is sitting around looking at Excel spreadsheets all day or bookkeeping or doing research for a law firm, a lot of that work is also going to probably get automated away through AI uh fairly quickly too. and and just but just one one thing on the software the the demand for software um doesn't end like there's seems to be a lot of things that can improve be improved with software and hence there's going to be more people building software maybe fewer people write writing code that that's how I see it yeah oh yeah that that makes a lot of sense to me well Anton this has been absolutely amazing I know you have to get off to another meeting so I don't want to waste uh any more of your time so the the app is over at lovable app. That's the best place to go to get it. Yeah. Or levelable. Lovable.dev is the normal. Oh, lovable.dev. Okay. So, head over to lovable.dev. Is there any place that you maybe want people to to follow you on social media or anything like that after listening to this interview? I share fun takes on building from Europe and on the AI space and what's happening at our with our advancements at my Twitter. That's my first and last name combined. Awesome. Well, thank you so much for hanging out with me and having this conversation. It's been really fun and um you know, really excited to see how Lovable evolves over time. So, really appreciate it. Thank you. And likewise, it was a pleasure. I'm looking forward to another chat in the future.