Former Intel CEO on What Went Wrong, What's Next + Lovable CEO on the Real Promise of Vibe Coding

Key Takeaway

Technical leadership trumps business leadership in technology companies. When business people run tech companies, they promote other business people, creating a dangerous cycle. The greatest tech companies—Microsoft, Google, Nvidia—are led by deeply technical individuals who understand that billion-dollar technology decisions can't be made through spreadsheets alone. Intel's decline began when it shifted from technical to business leadership, missing critical opportunities like iPhone chips and falling behind TSMC's foundry model.

Episode Overview

Pat Gelsinger, former Intel CEO, discusses Intel's rise and fall, the critical mistakes of prioritizing business leadership over technical expertise, and the transformative potential of AI and quantum computing. He explains how Intel missed major opportunities by focusing on shareholder returns instead of innovation, while competitors like Apple, Nvidia, and TSMC capitalized on technical foresight and continuous improvement.

Full Transcript

Transcript of Former Intel CEO on What Went Wrong, What's Next + Lovable CEO on the Real Promise of Vibe Coding from All-In Podcast. Auto-generated from episode audio; may contain minor errors.

Spent a long time at Intel. Yeah. Yeah. Yeah. Only 34 years. 34 years. Yeah. Yeah. Yeah. Probably one of the greatest American companies uh ever and then absolutely went off the rails and got absolutely demolished by Nvidia, TSMC, TSMC, TSMC, uh and I guess Apple to a certain extent. So, you had this incredible Intel Inside moment. We bought our computers based on you know, hey, the Pentium and that sound. sound. sound. Intel Inside, baby. Intel Inside. Intel Inside. Intel Inside. dum dum dum. The dum dum dum.

And so, let's talk about how things went wrong, what went right, and then how did it and you were there for a long time, you took a break, and then you came back. came back. came back. But, there seem to be have been some critical mistakes that we can learn from. So, let's just embrace it and go right into it. Tremendous successes in American company coming back now, I think, think, think, uh reasonably. uh reasonably. uh reasonably. But, when you when we look back on it and we do our postmortem, what were the mistakes mistakes mistakes and what would we change in terms of the direction of that company?

I'm going all in. If you were building a global financial system from first principles today, you wouldn't build it on 50-year-old legacy rails. You'd build Airwallex, one AI-native platform for global accounts, cards, and payments. It's designed to make the entire world feel like a local market. Others are bolting AI onto broken infrastructure, but Airwallex was built for the intelligent era from day one. Stop paying the legacy tax and start building the future at airwallex.com/allin. airwallex.com/allin. airwallex.com/allin. Airwallex, built for the future. I'm going all in. Having spent so much of my life there, you know, I view it I joined when I was 18.

I went through puberty at Intel, right? My joke, right? You know, it's just like you know, I'm so early. Grove, Noyce, Barrett, Barrett, right? Uh and Uh and Uh and you know, they they were the people I grew up at, right? You know, so on. They were my mentors. They were the people I adored for and they were deeply technical. technical. technical. Andy Grove. Andy Grove. Andy Grove. Andy Grove, Gordon Moore, Bob Noyce, you know, co-inventor. You know, these were deeply technical leaders. I remember when I joined the executive staff for the first time.

There was, you know, probably 15 of the 20 people that were in the room were PhDs. PhDs. PhDs. Right? You know, it was just that technical. And you know, I view one of the things that went off the rail was when it started to be run by business people. people. people. Yeah. Yeah. Yeah. to technical people. counters, the finance people. Yeah, and you know, when I became CEO in 2001, that was the first technical leader in essentially 15 years. Mhm. Mhm. Mhm. Right? You know, associated with it.

You know, and if you have a business leader, who does he promote? Business leaders. And you know, right? You know, so I think one of the fundamental things is and you know, as you look at the great technology companies today, you know, they're deeply technical. And founder led typically. You know, and even if they're not, you know, Satya is not a founder. No. No. No. Right? You know, Sundar is not a founder as well, but they're deeply technical individuals. And when you're making these hardcore technical, you know, decisions that affect billions of dollars, you don't do that through a spreadsheet.

spreadsheet. spreadsheet. Right. Right. Right. That's a lousy investment. Right. Right. Right. Unless the technology trends make it the right investment. And I think that's one of the fundamental things. And obviously, you know, in the five years, five, six years before I came back, you know, Intel gave a hundred billion dollars to shareholders. Oh, the dividends. And the And the And the And stock buybacks. Yeah, a hundred What I wouldn't have done for another hundred billion dollars on the on the on the Well, I mean, what what would you have done?

You probably would have made Well, I Well, I Well, I chips for iPhone, [snorts] which Intel passed on, yeah? Yeah, you know, but you know, it hadn't built a new factory in a decade when I got there. It's like, you know, how can you not be building? How could you not buy EUV machines? You know, there's just all of these things, you know, that you would only do as a technologist because the economics behind them by themselves were not good. Mhm. Mhm. Mhm. So, you know, it was getting back to the core of technology to me that was, you know, the fundamental thing.

You know, you make good decisions, you make bad decisions as leaders. Every business does that does that does that as they go along. But, you know, fundamentally, this is a technology business and you need technologists running technology running technology running technology that then hires technologists that are sitting at the staff that then hire the best technologists, you know. And take big swings at you know, categories that could matter in the future, like skating to where the puck's going. If you look at Apple, they did the same thing for the past 15 years, buying back the stock, tremendous amount of dividends, they're the largest holder of capital of any company, I believe, to this date.

And what if what companies do they buy? They buy little tiny acquisitions on the margins. I think the largest one was was Beats cuz they wanted to get inroads into, you know, certain demographic segments like in the Android space that they couldn't get into, but my god, what a colossal waste of time. Like you said, they could have done so many amazing things. Tell me about Steve Jobs in 2008, 2009 deciding I think we're going to make our own silicon and that impact because was that a covert product project?

Did you guys know he was doing that? Did he inform you? inform you? inform you? Well, that seemed to be another one of those forks in the road, yeah? You know, Steve was an incredible leader. You know, he was also a ruthless leader, right? You know, very difficult, you know, read Walter Isaacson's book on him as well. I had many, many conversations with Steve over the years. You know, for um but you know, when they moved to Intel on the Centrino chip, it was a big deal.

Yeah. Yeah. Yeah. Right? And they were putting extraordinary demands on Intel. You know, make the chip smaller, drive lower power. They're a demanding uh customer. And when he was no longer convinced that we could continue to do that, you know, he started the project. Mhm. Mhm. Mhm. Right? You know, and if you remember uh what was it, you know, you know, uh PA Semi, you know, they acquired some small company, started to build some competency. But you know, they did a few little chips internally. It wasn't a big deal.

And then the little chips got a little bit bigger. You know, and Steve was a master of this. You know, just starting, you know, these small efforts to build core competence inside the company. Uh I remember when we had the first conversation with Steve about uh porting the uh operating system to the Intel chip from the power chip that they were running on before they moved to Intel. And we were quite proud of the silicon software competencies that we had in compilers and operating systems. You know, so Steve, we'll help you port the operating system to the x86.

And I remember that Steve said, I've been working on that for the last four releases. four releases. four releases. Mhm. Mhm. Mhm. He had been preparing the core technologies inside of Apple for something that might happen Yeah. Yeah. Yeah. in the future. You know, and he was already You know, to me, I just remember I was just shocked. I you know, I ported the last four releases to the x86. I think we got this. Yeah. Yeah. Yeah. Right? You know, it was that kind of thing.

And that's how they got into the semiconductor, you know, doing their own semiconductor. Hmm, I'm not sure I can rely on Intel to be that much ahead of the industry and I can start optimizing the system design with the silicon design as opposed to relying on one that's been somewhat optimized for a Windows environment versus an iOS environment, you know, and their uh operating system and you know, it was just you know, you know, it was never that kind of thing that they used to say, you know, right, you failed as a supplier.

No, I can supply myself better. myself better. myself better. Yeah, and Jensen decides, hey, he's going to go all in on making these video cards and talk about just incredible just incredible just incredible serendipity that these happened to be also very applicable for cryptocurrency and running these AI jobs. Yeah, you know, Was that luck or skill or combination of both there? both there? both there? Well, you know, when you think about that progression, you know, Jensen he was just building high performance computers, you know, throughput machines, you know, when we were at the heights of our strength on CPUs at Intel, we sort of scoffed at his machines.

machines. machines. Yeah. Yeah. Yeah. Right? You know, so I go, it's a graphic machine, you know, who cares? You know, there's some gamers who want to use that kind of stuff, right? You know, it was always the big CPU and those little GPUs. But when they started to build a real software stack with it, right? You know, it was sort of okay, this CUDA thing and SIMT as a technology, you know, know, know, you know, multi-threading and so on and it just sort of kept getting a little bit better and it was a little bit Jobs-like in that way.

You know, we're just making it better every release and it's becoming a more robust and all of a sudden, you know, the crazy, you know, Japanese HPC guys said, hey, we could take those graphics cards and maybe start using them in HPC. Mhm. Mhm. Mhm. Right? You know, that was sort of defining moment where it wasn't just about doing graphics anymore. This was a more computationally dense platform to start attacking some of the world's most interesting workloads and I think Jensen would agree that was a defining moment and then sort of saying, oh, these aren't just graphics cards anymore.

You know, these are general-purpose computing devices that can start applying to these other uh workloads." And, you know, AI was, you know, had gone through what, its fifth nuclear winter by that point? Or it's like, "Man, you know, you know, this is never going to matter, right? We're never going to, you know, get the breakthroughs." But, the community around it was continuing to develop Yeah. Yeah. Yeah. uh you know, for it. And uh the CUDA software kept getting better uh generation by generation. And uh you know, I had a project at Intel Larrabee, right?

Where we were trying to take the x86 and essentially do the same thing, right? You know, for it. And, you know, in my first departure from Intel, the project was killed a week after I left. Huh. Huh. Huh. And the world would have been so much different, right? different, right? different, right? I mean, it really I think it's illustry of [snorts] illustrative of what continuous innovation, taking some risks, and doing that fundamental research, and the compounding power of technology because I think it was William Gibson who said the street finds its own use for technology.

Like, technology. Like, technology. Like, Mhm. Mhm. Mhm. Nvidia did not predict that this Bitcoin project would take over, and that this would be the best way to do those computations. Nor did they anticipate, I think, you know, that AI would take off. But, because it was the best solution, the hacker community could kind of Yeah. Yeah. Yeah. figure that out. Well, as we wrap on the Intel portion of your uh career, um okay, Apple silicon, that's one. Uh and then you have Nvidia. And then you have this Taiwanese company Taiwanese company Taiwanese company uh that starts making, you know, really great at fabricating the these chips.

chips. chips. Um and Intel missed that as well, yeah? And and maybe you talk a little bit about TSMC and their surging, and we can even get into a little bit of the the politics of it now. And then we'll get into some of these AI chips and venture investing. The you know, thing with TSMC was they started with a vision of foundry. Mhm. Mhm. Mhm. Right? You know, they were going to become the factory for the industry. And again, these factories are so expensive. 20 billion, 30 billion, and uh the engineering and the continuous investment required to do it.

And you know, it was a stunning, you know, a vision uh at that point in time. Intel was IDM as we called it. The integrated design and manufacturing. You know, we never worked to make our process and our factories available for third parties. Mhm. Mhm. Mhm. Right? You know, it was always this thing, "Hey, it's you know, we do enough CPUs ourself. You know, we reuse it for chipsets and some of the other things that we're doing." But it was never standardized in a way that it could be made available for a broad ecosystem, you know, using PDKs and all the design tools.

You know, we did a lot of our own EDA tools ourself. You know, one of the projects that I started early in my career was the foundations of EDA. Right. Right. Right. Uh as well. The first place in route, you know, the first standard cells, the first high-level description language. You know, it was so proprietary. And TSMC basically cut that in half and says, "I don't care whose chip it is. I don't care what you're designing. I'll be your manufacturing partner." Yeah. Yeah. Yeah. And at the time, that was such a trivial piece of the business.

Intel didn't even care. care. care. Mhm. Mhm. Mhm. Right? You know, and so on. And then, over steady progress over a long period of time, and Apple as a customer driving them to be good become really meaningful, you know, obviously the world changed. world changed. world changed. Mhm. Mhm. Mhm. And when I came back uh to uh Intel in 2001, TSMC was producing 5x the wafers of Intel. of Intel. of Intel. Wow. Wow. Wow. Right? Not 10% more, 5x. Yeah. Yeah. Yeah. And all of a sudden, that model of foundry became the model of the semiconductor industry with two exceptions.

Intel and memory. Yeah, memory DIs design and manufacture, right for you know, that is uniquely different and obviously, you know, we're seeing them you know, $3 trillion memory companies just extraordinary. You know, and you know, trillion dollar foundry company in TSMC. You know, the industry has said, I want a lot of wafers. I want a lot of innovation of different designs. I have a layer of standardization and EDA tools and the world changed. And obviously, as I came back to Intel, that was one of the core thesis of the new strategy.

Yeah. Yeah. Yeah. We must become a foundry as well. Five to one and now it's more like seven to one in terms of wafers, you know, to TSMC to Intel. Are we going to be able to onshore that? Obviously, we had the Chips Act and just give us broad strokes what you think's going to happen here in terms of obviously, Taiwan is in play. Some people in the administration believe um [snorts] it's going to happen the year after Trump's out unless he takes his third term.

Other people believe like it was going to happen as early as '27 or maybe going into '28. So, are we going to be able to replicate that here in America in a reasonable amount of time or is this like truly could be a cataclysmic event if you know, God forbid China decides hey, we're going to blockade um Taiwan and then the Taiwanese decide yeah, we're going to burn on the fabs and we're going to fly out all of the engineers and ship them to America. Mhm.

Mhm. Mhm. Well, there's a lot in that question. You know, do we have an hour to talk about this question? Well, I mean, we have six minutes, but Oh, okay. Oh, okay. Oh, okay. Yeah, do the best you can. Okay. Okay. Okay. Uh also I talked also about the AI bubble. So, super You know, three things about this super quick. You know, one is the Chips Act is having benefit. having benefit. having benefit. Yeah. Yeah. Yeah. Right? You know, when we started the Chips Act and you know, when 2001 when I came back, the US was building about 12% of leading edge.

Today that number is more like 18%. Mhm. Mhm. Mhm. Okay. You know, we're making progress. It's not 50%. Mhm. Mhm. Mhm. We have a long way to go, right? You know, Intel is starting to be a real foundry. foundry. foundry. Okay, that's real progress. Uh and TSMC's factories are up and operating at scale. Right? We have Samsung and you know, uh as well, but you know, I'd say the Intel and the TSMC progress, okay, that's meaningful. Now, let's make it ugly for a second. Uh the island of Taiwan has less than 3 weeks, a big article in the Wall Street Journal 2 weeks ago on this, less than 3 weeks of energy reserves.

energy reserves. energy reserves. Wow. Wow. Wow. Okay. Okay. Okay. That should just put a chill in everybody's spine. Right? Because the blockade after 3 weeks, the island browns out. When you turn off a fab, it doesn't come back on for 90 days. Right? The economic impact of a brownout of Taiwan is greater than the Great Depression. Right? Uh in the world. Never do you need to do anything, a shot to be fired. You just need to say, "Great, no energy for 3 weeks." No oil. Yes.

Right. Right, no LNG. Right? You know, that's how the island run. That is scary. You know, to me. We need more resilient supply chains uh associated you know, with it. And I don't think this is an alternative for the world because if it really does become a risk, you know, and I you know, I don't sit in the situation room and get all the data and so on. But let's remind each other that I think China has blockaded the Taiwan Straits seven times over the last 4 years.

Yep. Yep. Yep. This isn't a theory. No, no, they're running exercises. They're being pernicious and Right. Right. Right. pretty provocative in terms Is that 2027? Is that 2030? Is that 2035? Their intentions have been clear over a sustained period of time. We need more resilient uh supply chains. Yeah. You know, for it. So, something, you know, I put a lot of my time and energy into and we're making progress, but we need to go faster, need to go more meaningful. more meaningful. more meaningful. Yeah. And let's talk a little bit about the AI build-out.

I mean, you watched the PC revolution, servers, the internet. These were all extraordinary build-outs and then this is the build-out to end all build-outs. The amount of data centers, the amount of chips, the amount of inference needed. inference needed. inference needed. Do you think it's a bubble? I think I've heard you say like it's it's obviously a bubble, but bubble, but bubble, but what what's the risk factor here that we build too much or that the technology doesn't solve enough problems and we are swimming in tokens?

What worries you about what you're seeing now? The valuations of these companies has gotten quite extraordinary. extraordinary. extraordinary. And, you know, if they build too much and they spend too much money and they don't make enough money, well, uh based on your experience with running a company, a public one, that's a lot of tension on it. When you don't make as much money as you're spending, people tend to fall out of love with these stocks, yeah? stocks, yeah? stocks, yeah? Well, I do think there uh you know, there there is a silver lining here that guarantees we don't get too far ahead of our self in terms of bubble.

You know, and that is energy capacity. Right. Right. Right. Right? You know, energy capacity in the world is expanding for 5%. You know, in the US, we had a decade at 1%. Right? You know, I mean, it's just hideous what we did to our energy grid you know, over about a decade and a half. But now that's getting built out. But essentially, nobody's going to build and buy GPUs and build data centers if they don't have energy. So, essentially, you have an upper bound on how aggressive and how hyped and bubbled that we get.

So, I take a lot of solace in that, yeah, right? You know, for it because what then is the incremental value of a token? And if it's a measure of intelligence, it's somewhat infinite. Right? You know, in the sense if I have more intelligence, I will do, you know, better supply chain. I will do better finance. I will do more, you know, efficient logistics. I will, you know, all of those things. So, to me the the the potential value that we unleash in a token economic world is somewhat infinite.

Right? And particularly with labor shortages and so on that we see, right, in developed countries. I am an optimist, you know, that we're in a couple of decade build-out. Wow. Wow. Wow. Right? Not a couple of years, a couple of decades. One of the big objectives I've said is that I have to make AI 10,000 X better. Mhm. Mhm. Mhm. Right? You know, it's way too expensive today. You know, we want to drop, you know, by five orders of magnitude the cost per token, you know, the energy, you know, per token, so that we really do have Jevons' law that we just explode the access to AI, right, in much more economic uh ways.

Which it does seem like Jevons' uh paradox has been at play over the last year. Like, oh my lord, these tokens are so cheap and the tools are getting so good. Yeah, I'm just going to start using these tools all day long until the bill comes in and you're like, okay, yeah, maybe I need to get some ROI out of this, but you do have these incredible companies, Cerebras, Groq, etc., making inference. and and silicon and so on. You know, and you know, if we accomplish, right, you know, these orders of magnitude improving in token economics availability, reduction in energy costs associated with it, you know, we just have a fantastic couple of decades in front of us.

There has not been a time in human history where it's been better to be a technologist than the one we're in right now. We will solve chemistry. We will solve language. We will, you know, invent new materials. We, you know, new forms of, you know, interaction, you know, killing cancer, right, lifting people out of poverty. There is not a better time to be alive than the one that we're in right now. And as technologists, we get to sit in the driver's seat of it. Pretty amazing, and you're investing, uh, and that's your passion now.

What do you think of these valuations? It's quite seems, you know, if you lived through the dot com bubble, we did see a disconnect there. These companies slightly different. We just had 11 labs up, 600 million in revenue, lovable. I think they're at 5 or 600 million. So, that's quite different than the dot com speculation, yeah? speculation, yeah? speculation, yeah? Yeah. Well, fundamentally, we have real revenues, you know, real margins coming out of these businesses as well. You know, that that said, anytime the multiples get too high, okay, some corrections.

You know, and to me, periodic corrections that keep the multiple, you know, earnings multiples and, you know, so on in reasonable things is good, because this will not be a smooth curve. You know, I'm predicting two decades of goodness, and there's going to be lots of disruptions along the way. It's not going to be a smooth curve, and every time we have one of those corrections, say thank you, right? We're not letting the bubble get ahead of itself, right? You know, hey, we had the SaaS apocalypse.

There's going to be other apocalypses on that journey when when industries get impacted by the capabilities that will be unleashed. And that's even before it gets exciting in what I call the trinity of computing, classical computing, AI computing, and quantum computing. And when those three come together, okay, that's when things get really exciting. Hey, quantum's been about 5 years away for 25 years. Um, when is it actually going to do anything This decade. This decade. This decade. This decade, so by 2030. Yep. Yep. Yep. It'll have be meaningful.

What should we expect in terms of its impact in 2030? Like You know, you're going to be able to start doing things that cannot be computed today. You know, chemistry, you know, biology, there will be things that can't be computed today. You know, some of the easy things will be some of like the logistics where I will compute the best answer to get this thing to you. Traveling salesman problem? Right. You know, all of a sudden, all of those problems. Uh obviously, it's probably going to be, you know, 2020, 2032, 2033 when we solve, you know, things like encryption.

Right? You know, where, you know, you'll have the fundamental Q day, you know, kind of implications, but this decade we will see quantum supremacy to results across multiple industries. You know, we know how to build qubits. We know how to error correct qubits. We now have algorithmics, right, against uh quantum. And, you know, now it's just about engineering scale. engineering scale. engineering scale. Who's going to win? Well, obviously, I'm a PsiQuantum guy, right, since that's one of our portfolio companies. But, the thing that you're seeing is that you now have like four, five, six modalities of quantum that are demonstrating pretty good results.

Right? You know, across trapped ions, across, you know, photonic uh approaches, spin uh approaches. So, you now say modality is not an issue. Error correction's been proven uh across them. And, you know, I think the race will be on, and my prediction is meaningful results before 2030. Wow. That You realize that's about 40 months from now. Yeah. Okay. Meaningful results. Thanks so much, Pat, for sharing all this incredible uh information and knowledge. Great to see you. you. you. Very good. Your most valuable conversations rarely happen at a desk.

The hallway, the sink, the dinner, the quick founder call. Plaud, no pin S, clips on and captures all of it hands-free. Afterward, Plaud intelligence turns the recording into clean notes and clear next steps. You stay in the room, Plaud handles the rest. For people who live in meetings, that's real leverage. Wear your Plaid at plaid.ai. plaid.ai. plaid.ai. O C Cap is one of my favorite founders. He's the founder of Lovable. Why do I love this founder? Uh well, he built a product that people are addicted to.

Primarily, Anton, the people who work for me. Ha. And I love talking to you because as the founder, you have a North Star. You're incredibly laser-focused on enabling anyone anyone anyone to build great software. Yeah, that's the mission of the company. I'm paraphrasing here, but essentially that's the mission of Lovable. Mission I talk about empowering humans. Empowering humans. Empowering humans. Empowering humans. first gap first gap first gap is to build a product. Mhm. The second gap is to build a business around that product. product. product.

Right. Right. Right. And now we're at everyone at Lovable, we're we're working on both of these two gaps. Right. The first one, we got to very far. We're seeing a million new projects built every single week on the And on the on the second one, we're investing a lot in making it easier to run your business and to get people to care, people to discover what you build and the entire business of what you're whatever you're doing as a small business, as if you're a large business, we're also getting a lot of traction.

Um and we're actually seeing as a proof of that, more than 700 million visits to the applications every month. So, every month there is um extreme growth in in the surface area of the entire the entire the entire uh more than 50 million apps built on the platform to date. How many years has Lovable been in market or how many months now? 20 months uh since we launched, yeah. And and we're And again, we're we're seeing people who are first-time founders. We're seeing enterprise leaders move much faster together with their teams on this platform that has a lot of opinionated pieces in how you should uh create software and how to operate that software and how the different applications in your company connect to each other over time.

So, that's what why we're seeing so much growth also on the enterprise side which were where we're actually growing fastest right now. This is really interesting because 10 years ago years ago years ago people were doing WYSIWYG software. What was the name for it? Before vibe coding. No code, low code. Yes. And when I saw that 10 years ago in my incubator you know, every 20th company somebody would come in who was an MBA or not a developer and they had vibe coded something. coded something.

coded something. And And And not vibe coded, they had no coded. And they were using these different software platforms and the software didn't look good. It didn't work perfectly well. It was slow, but the promise was there. And I guess it took LLMs and this new intelligence to make actually good software. So maybe you could talk a little bit about who is the customer because developers do developers use lovable or is it the other 95% of society that are your customers? How do you think about who your ideal customer profile is?

Yeah. We're seeing people use lovable both with a technical background. That's about 20% are technical or some type of engineer. And they they love that we're quite opinionated. We put all the best practices into how the software is architected. And we make it seamless to we want from get payments set up in a very secure way. And do things like run security scans after every change even now in the background monitoring the projects. So, it's actually quite appreciated by the engineers in the technical community. technical community.

technical community. Also because it's a great bridge from the non-technical people which is four out of five are are non-technical. And they're building often first to figure out what is the right thing to build which is where lovable has always been exceptionally and and and now what we're seeing is that people are running running running businesses making more than million dollars of revenue on the on this platform. So it's a building for everyone. It's this entire spectrum. And what's what's exciting to see is often that if someone who discovers Lovable from their colleagues at the large company, they go out and then run a side hustle.

And some of those side hustles hustles really work. They make hundreds of thousands of dollars and then they become a founder after that. So this is cross-pollination from both uh Yeah, and this is like the really interesting thing about Vibe coding. If we were sitting here last year, people would look at it and say, "It's a great way to make a mock-up." Like you said, "A great way to think about product and maybe maybe maybe create wireframes or a workable prototype." All of that's out the window now.

The whole concept of building wireframes and building a mock-up Well, you can just go right to building the product in a day or 2 days. And what people I think don't appreciate about what you're doing at Lovable is after you've made a product that you're proud of and that has some product market fit, there are many more steps that are required. You mentioned payments, you mentioned security, uh making sure that the data isn't lost or that it's not leaked. That's changed dramatically over the last 12 months, yeah?

Very much so. So um I would say many engineers, they don't look at the code. They don't write code anymore. And that means that you don't need to be an engineer to create software, right? Um but the the thing that Lovable does for any anyone, also the non-technical people, is that it it um takes uh uh creates a structure for the architecture of the software that you build and it makes sure that you don't go off a cliff um and that things like setting up payments, emails, things like getting discovered by other AI chat engines and uh by Google search.

Those things are kind of taken care of. Yeah. You don't have to know how the all these things work in the details. You trust You can trust the platform to take care care of data security, uh connecting to other tools that you might be using in a secure way. And And that's really where really where really where uh us being opinionated from day one and being focused on making this for the 99%. It's a It's a vast market, right? Right. Right. Right. From the From day one is what made us very successful.

very successful. very successful. Yeah, and I can tell you internally, I gave my team all the different tools they could possibly want to use. And somebody had started with lovable. I think I told you the story when you were on This Week in Startups a year ago. Like And they made some interesting websites and they were trying to make an intranet. They couldn't quite get it done. done. done. Then I had some people who started using, you know, Cursor or Claude Code. They started vibe coding stuff, but they couldn't finish the product.

And then people tried to solve some problems with Co-work. I really like Perplexity computer. And then my team came to me and for one of our projects, I was talking to you about Founder University, our pre-accelerator, they wanted to make an intranet. an intranet. an intranet. Um now, this is something I would have never okayed because it would have cost $500,000 $500,000 $500,000 10 years ago to make it and we don't have that kind of budget, you know, we would rather put that towards the founders in the program and getting more people into the program.

And in 4 to 8 hours, they made the whole intranet and they made a bunch of things I hadn't asked for. And it was the person running the um um um this Founder University who made it. And she did it on her own without uh permission in lovable. I said, "Whoa, what How did you build this?" She said, "Lovable." I was like, "Oh, we still have lovable?" And they're like She's like, "I just put it on my corporate card." To your point, she made it. Now that that that software is driving the program and the reason people do the uh the the program in their country, we have it in Saudi and in Japan, is because it has economic impact.

Yeah. Yeah. Yeah. So, I said, "Hey, I have an idea. Can you make for me an economic impact of the 50 companies that are in the program?" program?" program?" She asked Lovable to do it. I gave her some, you know, prompting, human prompting, boss to Now it has the economic impact in there, and it considered, you know, with our prompting, prompting, prompting, "Well, how many people work at each company? What are they paying taxes? How much do they rent their home for? What is their average salary?" And it built something that I would have never been able to afford to build.

And Lovable is 50 bucks a month, I think. I don't know how much you charge, but it's far too little. Like, $50 a month, I think. the that's the if you're on a business plan. Yeah, I start at 25. Yeah, so Yeah, so Yeah, so uh the economic impact of what you're building building building is I would equate for what you built to us, it would have cost me $500,000 2 years ago. It was built in 4 hours by an employee, which if you just put employees at 50, 60, whatever, 70 dollars, dollars, dollars, uh plus the cost of your software, it got made for less than $2,000.

In a year. It's extraordinary. It's extraordinary. It's extraordinary. I'd love to hear more about the progress of the of the internet. Anything that you asked for that you want to forward directly to me? Uh well, right now, you know, my concern was security and making sure that data didn't leak, and they talked to your team and they went through it, and it's secure, so we feel good about it. Well, look, um I'm now asking people who do penetration testing to say, "I want you to compare all the tools Yeah, uh make sure that there's a all the work that we're doing that's not visible on security and trust.

There's a lot of a lot of other things um where we uh we invest and spend money on that every also free users get a lot of security scanning running in the background that that actually um translates to something that security experts can can see. A year ago we were at mockups. Now we're at functionality and secure and super viable for deployment. Where will you be in a year? Yeah, so what we're seeing is that there's a gap in build being able to build the product right.

And and you built an entire intranet on on the platform. That's great. Um what we've done since then is to have a new product line basically, the hosting part, which is both the AI and all you know all the normal hosting. And that product line has been growing faster than the building thing I I mentioned. AWS competitor. AWS competitor. AWS competitor. It it's a let's it's not you run all your software and then we're working with companies like AWS and Red Hat as well. But but what you also want to have is um is um is um to use Lovable we're we're seeing by our customers as an AI co-founder.

A partner that you talk to about everything in your business. And if you're running your apps your apps your apps your tools are on the platform, then just talking to Lovable has access to all the data that you might want to know about your about your company, how it's doing. So we're we're working with some of our customers in pre-release to give them access to a co-founder that works for for you even when you're sleeping. And comes back to you in the morning and says like here are some strategic directions you could go.

Here's some optimizations you can go go in terms of growing your business faster, serving your customers better, faster. And and and that's that evolution towards operation and intelligence for towards driving towards outcome outcome outcome for your business. So you come and build the software, but you stay to build the business. Yes, to operate your business. And what we're already doing I've I've been doing for a very long time is to compound from everything we're learning. Every time Lovable makes a mistake, it goes through our genetic system with our engineers in it improving it.

That compounding intelligence is of course applicable to our our customers our users running their business on our platform as well. Is software going to become 100% bespoke even like the internal tools? I was looking at Slack. And our bill for Slack even on the highest version is maybe $10,000 a year. It's not a lot of money. It's well worth it. But I was starting to think, well maybe I should buy code my own Slack. So it's integrated into everything we do at a deeper level. So how do you think the What do you think the future will look like in terms of some of these you know uh foundational pieces of software that every startup every enterprise uses?

Salesforce, HubSpot, Slack, Slack, Slack, uh the Google Suite, Microsoft Office. Will bespoke software start to replace those? Do you believe? I I like this question. Let me ask answer you, but I'll just give you a story about someone I recently heard who's going on this journey. They're quite the brand. So Nenad, he works at a pretty large company in the US, Nursa. And he came to our platform because he wanted to build out the new product lines. Nursa study for educating more nurses. And and he built out all the admin tools for the program, the scheduling for the nurses getting getting their licenses and their certification management.

And he was able to build that into a product and to take it to market because they have they have had all that access to nurses wanting their certification. What he also did was he took it into the back office internally and they've now replaced more than 10 tools that they had in just bespoke applications. And And And I think in terms of your question, you can do that for multiple reasons. In their case, they're saving more than a million dollars per year. Right. Right. Right.

that's huge, right? But it's also the case that case that case that in some cases, you have specific requirements where the tools that you've been using to date, they aren't suited for those requirements exactly. And in those cases, I think yes. You will have more more bespoke solutions. Yeah. Yeah. Yeah. But but we're all I will I also expect us to see that [snorts] Lovable continues to interoperate with all of those tools. And I'm not sure if you tried this. If you ask for connecting to anything in the Google suite suite Yeah.

Or now anything in the Microsoft suite or or Slack. If Lovable guides you through all the steps to do that in a way where you can get a lot a very good overview of exactly how the data flows, which is of course very important that you don't give access to the wrong person to the wrong data. And you can continue to use Salesforce, Salesforce, Salesforce, HubSpot, and all the tools that you kind of like to use under the hood but with a bespoke interface on top of it.

How have these new frontier models they're in some ways competitive but in some ways you can use them to power Lovable. So how do you think about the competition with them, open source and the future of Lovable because people have announced that Lovable's dead every 6 months since you started and then every 6 months you go from 100 to 200 to 300. I think you're at 400 million in revenue, something crazy. something crazy. something crazy. We we met we reached 500 in May. Okay. Okay. Okay.

Yeah, yeah, it's growth is a phenomenal. So you're dying again by another 100 million in annual revenue. So, but underneath the hood you're using some of these these these Yeah, let me explain. Yeah. So, we've always had the strategy that we do whatever is best for our customers. And in terms of the intelligence, that means that that that we're using we're using we're using multiple models. And so, if you ask lovable now, it's actually routed to the model that's both suitable to whatever you want to do.

And that's both the commercial frontier models. Yeah. Yeah. Yeah. From multiple vendors. And increasingly, it's open weight models. Where our team when whenever it get gets routed to to our own model, that model becomes more intelligent for our agent harness. Yeah. Yeah. Yeah. Especially on the mistakes that it might be making in some cases on which which tool to call, which integration to create, and how to guide you through success for your business. Right. So, you're all in on open source. You believe that's the future of lovable.

lovable. lovable. Well, I would I'm reading into it. So, we have a multiple partnerships, and we're investing heavily to be close with those partners. those partners. those partners. Right. Right. Right. And it's the big the big labs. And it's also to make sure that um um um we get the fastest performance at the lowest cost for our customers when we know that we can do that with our own models. models. models. Right. Right. Right. And we have a really really strong research team up in Stockholm who is working on what's called post training.

Sure. Sure. Sure. So, and we're applying all the best practices to do that and scaling up that team team team quite significantly. Since we also believe it's a it's a part of the European ecosystem to have that capability in Europe specifically. Are you doing or are you using any of the data labeling, data training companies to help you understand the most common businesses and build that proprietary data? proprietary data? proprietary data? So, so we're what we're doing is that we're looking at the mistakes that any of the models do right now.

right now. right now. Ah. Ah. Ah. And then we we prioritize them by what drives most impact for our customers and then we make the models we create data sets where we we do did something called reinforcement learning for specifically for the problems where the frontier models are making mistakes for us right now and and and we have this enormous token distribution right from right from right from a million new projects being built every every single week. You're burning a lot of tokens. We are yes and that's and that's a lot of signals for making the system both the agent harness and what we've been refining over the last two years which is the skills that we have the how does it like internal type of skills that the agent knows when to remember the facts from our software engineers that know how to build really really good software we're modifying both of those on every every single week.

Makes total sense and somebody told me some companies are doing token dumping they're you know selling a hundred dollars worth of tokens for fifty dollars dollars dollars you know basically they become token resellers in some ways and they're money losing businesses you have to you're money you're profitable I believe now or close to it. We we always monitor our margins but again again again we're doing what's best for our customers and that means that often means more intelligence so we're not we're not looking at oh let's use it we use we have never had the decision to say let's not use a cheaper model here if it's if it's if it's measurably worse for our customers and we can measure that what's best for our customers.

customers. customers. is it unlimited for the fifty or you have caps now? We have caps. Yeah you have to have caps over existing caps are people starting to hit them? Yeah our customers definitely hit caps and then you can top up you can have a we have multiple subscriptions here. What number I'm just curious like what percentage of people need to top up they're so addicted to it that they're blowing past the This is This is This is Some of the from the lower subscription tier tier tier Um, I I think it's um uh much uh much uh much I something like 60% of our customers are I think it's like I'm hearing that more and more often that people are willing to pay the overages cuz they're getting so much value.

And I think that's the future of the business is people are looking at it going like I am [snorts] [snorts] [snorts] well, if I'm paying $600 and if you token max to 6,000 a year but this is a $500,000 piece of software. I don't care. I'm still paying somewhere between 0.1% and 1% of what I would have paid 3 years ago. Who cares? Go for it. Um, so Yeah, what we're seeing is everything is about moving moving fast these days. And and the AI more AI usually let you move much faster.

So, this this spend is usually worth it. Hey, do your customers a final question for you cuz I'm starting to see this now where [snorts] multiple people in the organization try to solve the same software problem software problem software problem and they're competing with each other. So, like this intranet I'm talking about, we built one for Japan. Yeah. Yeah. Yeah. But somebody built the US one. So, now I have two pieces of software. So, I said to the two different people or or or do we have Did you guys fork each other's code or They're like, "No, we just built two different lovable projects." And I'm like is that the right thing to do because you went faster and I had two swings at bat, two different intelligent brilliant people making their version of the software.

software. software. But you would never have done that in the previous way of building software. You would have one track of software and you would be building Franken software where you'd be trying to get all the needs into it from the two different groups. Uh yeah, I'm actually a huge fan of very rapid experimentation. And I I have a story where for a while I worked at a place called CERN where they do particle physics. It's it's pretty here in South Europe, right? Uh And that's where I was introduced to this concept of competition where they have two actually quite isolated teams working on the same the same the same particle accelerator but different places on it.

And then they don't share the results until they publish. And that way they way they way they they can kind of over time learn what's working best in the different organizations. But you don't get stuck in a local minimum. And it's you know, free markets work extremely well because of competition. And they they they do that in academia as well. And now since the engineering is less of the bottleneck, it's more of the question of what is the right thing to build. I think it's a great thing to have if you have the sufficiently many humans, right?

To do to try to attempt the solving the same problem in different ways. And then if you do that on Lovable, what I like to do is I I take I bring up a new project or one of the projects and I I say, "Hey, can you go and I check out this other one and take this these three things that I really like and and bring them bring them over here and maybe even run an a split test, run an experiment to see if it's if it's improves improving the metrics for for our customers we're trying to serve." Did you see somebody used Fable to build Fortnite?

Fortnite? Fortnite? And uh And uh And uh I think there's some 3D some of the 3D games, yeah. games, yeah. games, yeah. Yeah. This What is your take on, you know, this latest version from Anthropic, Fable? I know they're a partner or I assume they're a partner. I don't know that. Yeah, we use Fable as well as a one of the models, yeah, in Lovable. So, what do you think of it in terms of compared to the last generation? Faster, better, both? better, both? better, both?

Yeah. Yeah. Yeah. Is it a massive step function? Yeah. What I've seen is that it can in the first attempt create very sophisticated things that look really good. Then when as you're evolving, right? It's it's still the same thing where you as a human, you have to think you have to often planning together with your agent about what is the right thing to do. And and that's more of that's again more of the bottleneck. bottleneck. bottleneck. Whereas more intelligence is on some tasks, it's great. Like it creates really beautiful things, 3D games for example, but on figuring what to what to build, figuring out figuring out what are the right strategic directions or experiments you should run to improve the outcomes for your business, that's that's not changing as fast.

It's the humans knowing how to use the tool to get and to plug in all the right data to be able to take the right decisions for taking your product forward and to take your business forward. Uh listen, I love the product, but even more than I love the product and you as a founder, I love the outcome. The outcome for business is extraordinary. So, anybody who's listening, Lovable is absolutely worth your time. Don't wait. Just put it on your corporate card and start building. That's my message.

Just start building with Lovable. It's an incredible product and congratulations on being reborn six times cuz every six months you had 100 million in revenue, it seems, and then everybody says Lovable is dead because the new foundation model is so good, but you keep studying your customer customer customer and you keep somehow surviving and thriving. So, congratulations as an entrepreneur. entrepreneur. entrepreneur. Thank you so much, Jason. I enjoyed that talk. I hope you enjoy the rest of your stay here in Paris. It's pretty great and the Palace of Versailles is so impressive, huh?

Someday we'll be building this with Lovable and Optimus robots. I look forward to it. I'm going all in.