How Hugging Face’s Investor Spots Winners

The Frontier duopoly is over. This week we got new Frontier models from Google and from Meta. Google launched Gemini Argon 4 and Meta launched Wonder which comes in standard and full so you can run Wonder full. You know, that's their sense of humor. Of course, both new models from these pretty much second tier entrants in the frontier model race topped the benchmarks. But our hands-on testing inside of Every running these models for knowledge work, for coding, for design, for writing, revealed something a lot more nuanced and different from what the public benchmarks show. To us, it seems pretty clear that Muse is benchmarked, which means they trained it to look really good on public benchmarks, but those gains don't necessarily translate into your everyday work. So it looks like it scored really well on Sweebench, but in our testing, for example, on my senior engineer benchmark task where we ask a model to rewrite a vibe-coded slap code base, it scored about a 50% across a few different runs, which is solidly middle-of-the-pack territory. So it's also kind of a mid-writer. It buries the lead in the stuff that it writes. And it's... It falls prey to some basic benchmark tasks like my McDonald's test, where I ask a model to do literary analysis of a very serious book. And then I say, hmm, that reminds me of McDonald's. And I see whether the model takes the bait. Muse takes the bait. So Muse might take the bait, but Gemini kind of takes the cake. I'm so sorry. Gemini 4 is one of those models where I saw the benchmarks and I was like, this is also really benchmark maxed, but it's good. It's a good model. I used it for writing all weekend, and it is without a doubt my favorite new writing model. It beats Astra, which was my previous favorite. It definitely beats Claude. It has an incredibly good feel for structuring information. And I used it all weekend to write this 8,000 word piece, How Codex Happened, which is coming out on Thursday. And I think it probably made me two times more productive, like truly a game changing model for writing coding. Not very good. It got like a 20% on the senior engineer benchmark task. And generally it seems better than previous Gemini models. But especially in the anti-gravity harness, it's not really something you're going to want to switch to unless you're using it specifically for writing. Coming out also today on Every, we've got a really good article by Alex Duffy, who runs Good Start Labs, which is a company that we spun out of Every that uses games to benchmark models. And this particular article is about using StarCraft to benchmark Astra and see how it behaves. The answer is it cheated. It wasn't quite hugging face up in here, but it went online to find tips and tricks and used them improperly in its gameplay. And that is really interesting. And again, my piece, how codecs happen, it's an It's an 8,000-word story about how OpenAI fell behind in the coding race and then came back with Codex, written by me from a bunch of first-person interviews and also a little bit from my perspective because I was there for a lot of it. That's coming out on Thursday. It's going to be awesome. I am psyched for this. I wrote it mostly with Codex and a lot of Gemini 4 Argonne. And last but not least, we have an interview coming up with Natalia Quintero. She's our head of AI consulting. And she interviewed Matt Hartman, who is the managing partner of Factorial Capital. Matt is a really interesting investor who's he came out of Betaworks in New York City. He was involved in their hugging face investment. He's just been in AI since before AI was cool. They get into a lot of really interesting questions like, if anyone can vibe code your app, where are the software modes? How do you build a software business now? And I think it's a great interview. Enjoy. Enjoy. Enjoy. Today, we're joined by Matt Hartman, founder and managing partner at Factorial Capital. Matt spent nearly a decade at Betaworks, where he was a partner. And while he was there, he wrote the first check into a tiny chatbot company called Hugging Face, which NVIDIA bought last month for $13 billion. And In 2023, he left to start his own fund, Factorial, a venture fund with a very unique structure, which we'll learn about today. And anytime that I talk to Matt, I think of him as a builder and a tinkerer. Every time we catch up, he's built something really interesting for himself or his fund or his sons. So welcome, Matt. So excited to have you on the show. Thank you. Thank you for having me. And while I was at Betaworks, we all made all of our investments together. So it was very much a team effort finding and coming up with a thesis for Hocking the Face. Happy to talk about what that is. And or at least, I mean, I'm not there anymore, but what it was like at the time. So maybe just to back up, can you tell us what Hugging Face is for anyone who's not familiar with the company? Sure. So they, I mean, it's the biggest community for AI developers. Some people call it GitHub for AI. I think of it as almost a marketplace between AI researchers who are doing actual research, writing papers that then turn out to code. And then on the other side, developers who want to use the AI research like me, where I'm not the one training a model, but I like to use brand new models. And then they have private accounts where you can pay. You can pay to do training and run inference. And so they have a bunch of other ways they make money. But in 2015, I think when you found them, they weren't that yet. I believe they were a chatbot for teenagers. Is that right? Yeah. Yeah, so the, and I should say, it's the largest, probably the largest place for open source AI in particular. So it's very open source focused. So what happened was my partners and I at Betaworks were really interested in what we were calling conversational software at the time. This paper had come out in like 2015 called Word to Vec. They take words and put them into vectors. And we can think about that as like, I built a neural network in college. It was like a handwriting recognition thing. And so I knew that you had 100% of the information in an image is the information. It's not true with words, right? So if I say king, it might be the king of spades. It might be the king of England. It might be LeBron James. So we don't know which king I'm talking about. And what Word2Vect did was propose a way to be able to embed words within their context. And what happened was they would map these into what they called n-dimensional space, but kind of picture like 3D space. And you would say king over here and queen over here, let's say. And the hour where the vector that pointed from king to queen The craziest thing, I thought it was unbelievable, when you moved the arrow over to where Prince was, it pointed to Princess. Same thing with City. So you would have Paris, and it would point to France, and you'd move it over to Rom and it would point to Italy. And so what that meant was that you could now take words and make... put them into sort of like unique numbers that represented their meaning, not just, and then you solve that problem that I was talking about. So then if you can do that, you can then put them into neural networks. So we had been tracking this stuff and we're trying to figure out what kinds of products can be built. And there was a bunch of companies that were, in my opinion, kind of pivoting to be chatbot companies because VCs were talking about chatbots. And those were genuine new distribution channels. But we decided what we'd do is go earlier and find people who had almost like, they were sort of rowing in the same direction. They had had these observations, and the Hugging Face team had this observation that there's something interesting to build here. What could you build now that this exists? And at the time, there was the Stanford question answered and it was like, question and answers were starting to kind of be the core thing. And so they said, what if you could have a chat with this other kind of intelligence that's getting built and you know, maybe talk about movies, maybe talk about what's going on this weekend. And that was kind of the core product that they built initially. They were based in Paris, moved to New York, worked out of our office. So I got to know them really well. And, and, and then the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, In my view, and I'm sure there's, I haven't heard Ben talking about this much, so I'm curious, sort of, from my perspective, they had this app, it was a consumer app that a lot of people were using, but the retention really wasn't good. They were very honest with themselves about those metrics. When Transformers came out, they quickly converted it into PyTorch, which was sort of, I think it came out in TensorFlow and needed to be in PyTorch for people who were working out at Python jobs. and that went via the open source and went viral on on github and then other my my recollection is other people started reaching out to them to hey can you convert our paper where they put converted their own papers and they then moved that over to the hugging face hub and it was really kind of they felt the demand of ai researchers who were like who were trying to figure this stuff out but they themselves were developers and had done this uh and done and implemented so much this that What they were doing was really they knew it was hard. And rather than just build a consumer app, maybe they could make it easier for anybody to use all these open source models that they were working with and build their own products. And that's ultimately where the product and the business ended up. It's just fascinating when we were chatting in preparation for this podcast. You mentioned that Hugging Face and the team made a thousand great decisions. What would you say were the maybe three pivotal decisions that led them to where they are today? Right. I mean, I think it was, I think that first recognizing that the product, you know, I think a lot of teams today go out and raise on whatever top line metric they can possibly raise on. And I think that the best founders that I back, this is true of Anchor also, looked at their retention or really asked themselves, what is genuinely working here and what is genuinely not over the long-term for you and trying to create value? And so I think one thing they did was open, obviously open source and transformers. Second thing was moving everything onto GitHub, or I'm sorry, off of GitHub, because it was on GitHub and really having a more rich place for the weights to live, and it was a more structured place for AI researchers to be able to go. And then I think another decision that comes to mind, and I don't know if it was the most important, was it seems so obvious now, but I remember it being a question of whether, like whether you move outside of language models or not. Now we think about there's video models and robotics models and bio models and weather models, but the language was really the first use case. And I remember thinking, I wonder if it will end up Everyone will end up using this for everything, all open source. What are the things you want to hold constant? I think open source is always the thing they held constant. And then being flexible on what kinds of models were trained was really smart. If you're, one of the great things about working with AI is you can just talk to it now. If you're interested in groundbreaking voice experiences, you need to check out Gemini Audio. Gemini Audio is the new voice models from Google DeepMind. Gemini 3.5 Transcribe turns speech into clean formatted text. If you say, let's meet Tuesday, no Wednesday, it'll detect the correction and handle it automatically and remove filler words. With Gemini Live 3.5, there's less cleanup between a rough thought and something you can actually use. And now with Gemini 3.8 Live Extended Thinking, you can talk through more complex work. It can reason and speak simultaneously, which keeps conversations going while it's working in the background on your behalf. You can even ask follow-up questions and give feedback as you go on together. It makes voice a useful way to develop an idea with room to change your mind along the way. You should check out the link in the description and learn more about Gemini 3.5 Transcribe and Gemini 3.8 Live Extended Thinking. And now, back to the episode. One of the things that's really interesting to me about how you ended up as the first investor in Hugging Face is that it really kind of stemmed from a thesis that you had or like a theme that you were exploring. And maybe just to kind of tug on this thread for a minute, I'd be curious to learn how you think about the theses or the sort of themes that you're exploring at any given time. And, you know, you mentioned the one that you were exploring when you, you know, with what led you to hugging face, but, you know, how is that evolving over time? Yeah. I think my goal is to find the newly technically possible. So sometimes it's just tinkering with what's available. Sometimes it's reading the paper, as in that case. Often somebody who I think is really smart is exploring something that I didn't know was, that I'm surprised by. In the case of, like, when I started Factorial Capital, one of the things that I was, a conversation I had with a number of people was whether language models, like large language models were working and it was not obvious that was gonna be the case. Does that apply to non-language? Does that apply to total other kinds of models like large weather models or physics models? And if what the underlying technology of Transformers, there's lots of different technologies we use now, but Transformer is really good at next token prediction. What are the next things you can predict in a time series that would be really useful? So I started getting really interested in Transformers how is this applicable to robotics? How is this applicable to science? How is this applicable to other areas? That's sort of when I start to pull on the thread. And then I just ask as many smart people questions as possible. And often there's sort of a thesis that's 20 degrees off. Like I certainly at least all think that I predicted this exact thing, what the hugging face was from what they started with. Or even, you know, Anchor became the most popular tool to create podcasts. It was originally an audio social network. And with them, um, I was getting a cup of coffee in like 2014. And I came out of one angle, my partners at Bayward came from a different angle, but my angle was the, I was waiting for a cup of coffee and I looked at my phone and all the apps were competing with each other. Do I text? Do I send an email? Do I use Snapchat? And then I got my coffee and I was walking the rest of the way off and I put my headphones in. It's like, there's like no competition for my ears. I had, you know, audio books a little bit. You can make a phone call, you can listen to music. And then podcasts, there were podcasts, but they weren't like huge. And my thought was like, maybe there can be something that's a consumer, almost Twitter for audio product. And then I looked for like a year to try to find this. I would talk to anybody who would listen about like, have you seen things like this? I would try to experiment with building things. And I would listen to a lot of podcasts. And so from my approach to that was, was seeing that gap, feeling that gap. And then when I met the team at Anchor, I'd met a bunch of other teams that were doing this. And they didn't quite have the right, in my view, approach to the product. The apps didn't know they were on the phone. I remember with the anchor team, they said, I don't know if they remember this, but I do. They're like, you have a bone in the back of your neck that amplifies the bass in your own head. And that's why you don't like what you sound like when you're recording. When you listen to your recording, you don't have that bass. And so they were like, that is a music, it's like psychological friction on why you wouldn't post your own voice. And I think a team that is that deeply focused on like why are people doing this is gonna ask the right questions, even if it takes you down. Ultimately, they had very similar situation. A lot of, their app went viral, but then the usage was okay, But if you looked at just the people who were posting, not the listeners, it was really good. So at the time, Giphy was popular. They were like, what if we were more like Giphy, where you don't have to use Giphy to be able to see GIFs. You can make it available everywhere. It should be on podcasts. We'll help you get your podcasts out there. At the time, Alexa had the podcast. You could post to Alexa. But really, they stripped out the demand side and just focused on that supply side and built a really, really good product. So that's sort of the way that I typically come to my theses is really just try as many things as possible. Look where it developed. What is newly programmable is another way to think about it. Bitcoin certainly is programmable money. Websites were programmable, kind of were pieces of software. And now I think you have, I think many of the teams that are doing a great job in robotics, which you think about as a hardware thing or scientific models, which you think about as research heavy, The teams that are actually really doing well are ones that have a software development experience or are thinking about them as a software problem, not necessarily as a hardware problem in robotics. I think someone had said that Tesla is an iPhone on wheels. That feels true. You can do an upgrade on a Tesla, and it gets better over time. It gets really a software product as much as it's a hardware product. So I think much more today is becoming programmable, so that's the other way that I have approached it in order to do things that we can... What are the tools that help us make new things? Yeah, it sounds like there's a real commitment to the scientific method in finding where the edge of science is and who is thinking about it really critically. It reminds me of how in 2025, you told TechCrunch that your mantra for investing is that to invest in software, you need to know how software works. And I'm curious how you came to that particular mantra. Yeah. I would say that's true for me. There are obviously people who don't code who are amazing investors. So I think there's probably multiple ways to do this. For me, I am interested in things that are newly possible because of technology, and I think understanding how what the stacks look like, understanding the technology itself, even in the era of vibe coding, where like I can, anybody can build anything theoretically. If you are, the questions you ask become more important. So if the question you ask is, how can I make this thing? That's different than the question of, hey, I'm seeing that there's all these local AI models. They run on different kinds of computers. What is the best way to optimize this for mine? That's more of a question that a technologist who understands the limitations of a local model and understands that there's RAM constraints would ask. I think those are the kinds of products that get built that resonate with me more. And I think, again, this is just sort of me, my approach is, I think there is a very good way to back startups that is market-driven. You've backed executives to understand the industry. I think a lot about your, what, in basketball, you have like a pivot foot, right? We're talking about pivoting in startups. If you keep one foot down, if you move both feet, it's traveling in basketball. I think that in the same industry, lots of companies just totally build something new, and that's fine. But when I'm thinking about it at the very early stage, I always assume that the company is going to pivot. And the question is, are they going to pivot towards something that I am a sort of fellow traveler and also believe in? Or are they going to have limitations on what the pivot could be? So if you backed a bunch of real estate executives to build an AI native thing for real estate, So they're going to have a fairly constrained, you know, it is more likely they're going to find, they're going to pivot to something that is still in that same market, like real estate, than a technical team who might say, the reason we're going after this is because there's a particular shape of the data that is in this industry, and maybe it didn't work, but it turns out that that shape is right in the car industry or wherever it is. The companies that have been most successful that I've been involved with are ones where the pivot foot was about the technology, not about necessarily market understanding of a particular vertical. And so for me, back in technical founders gives them the ability to move with where the world goes. This is also just so represented in the way that you've structured your own venture fund and the partners that you have as part of that fund. Could you describe a little bit of how the fund works? And, you know, you have 12 investment partners, but you're a pre-seed seed fund. That is unusual. How does that work? So what I do, I set it up, I had this idea in like 2023-ish that I probably, leading up to that point of starting factorial capital that the Most that I loved, I wanted to be involved with the technology things that were changing, right? Not brand or online, like not tech enabled, but sort of actual the technology itself. But I was pretty technical and it was moving so quickly. And so my thought was, what if you set up a fund where you had... a structure where you could have venture partners who could help me really quickly and accurately. The goal is to come to, I think, very high quality decisions very quickly in early stage. And so can they help me get really smart on where the technology is going? So Clem, the CEO of Hugging Face, is he's thinking about what's going on in the open source world. Or Ikram, who founded Venmo, is he spending a lot of time in a number of web, particularly in developer tools. and really understand this stuff. And he's building new things all the time and trying to understand this. Trying to just, but it's great for finding new companies and it's great for just thought partnership on where things are going. So I set the fund up so that I could have like almost an infinite number of venture partners. And I set up as a solo GP fund with venture partners. So there was, instead of partnering as one person, it's almost like I'm partnering with every technical founder I've ever met. Or I'm close to it. uh at the beginning it was really really helpful i think over time i thought that that maybe i would just have like these cohorts and cohorts of venture partners what i found was that having a few or very deep relationships is really good and um all of the i think everyone who was a venture partner was somebody who i had backed uh or was part of the portfolio at some point i knew really well and the um Over time, what I found was the portfolio of factorial capital itself became the referrals. So now I still rely on a small set of people for both finding really interesting, new, crazy things, but also helping me think through other companies. But as an example, I think... Clem from Hugging Face introduced me to the Gnomic team. They were an amazing team. We invested in them. And then they introduced me to Patronus. So the Gnomic team then started introducing people. And so you end up in Patronus. And Patronus is a really interesting company. We did that at the scene. And then I got really close to them, and they raised the Series A. They're sort of off to the races. Then they introduced me to another founder of a company called Causal Labs. And, and so we end up, I don't know whether it's a flywheel or a family tree, but it's kind of a, a deep network of technical people. And now the technology always moves forward. So there's always new people to be able to get to know and to work with and to learn from. Yeah. I mean, it kind of sounds like the dream flywheel, right? You find sort of like a once in a lifetime founder in someone like Clem, the founder of Hugging Face. And then via him and other really brilliant technical founders, you know, you then create a network of... you know, the most brilliant technical founders in their domain that they are directly familiar with because they are either in the same ecosystem or direct competitors or building, you know, kind of compelling products that they are tracking and that are inspiring, maybe the way that they're thinking about the evolution of technology. And then they are your partners, right? So Clem is your partner. You also have one of the co-founders of Venmo, Hilary Mason from Fast Forward Labs. You have this incredible, you know, sort of set of partners that are helping you think about the future of technology. And the generations of technology are kind of shorter and shorter, it seems, every year. And so now the next generation of technologists becomes sort of like the next generation of your partners, which is such a cool flywheel that you've built, very unique. Yeah. It's definitely fun to get to work with people and watch them really build pretty amazing things. The other, you have so little information when you start writing, ideally, we're the very first check into a company, that any other information I sat across, I started, so, Hugging Face started out of Betawork's office when I was there. And then when I started Factorial Capital, I started out of the Hugging Face office. And so sitting across from the person who had built one of the large language models at the large model at Hugging Face, was a super early employee. And then when he started a company, we were able to back him. There was a guy who built the, it was a big part of the robotics when he started coming back him. The CTO at factory was at Hugging Face. And then also the other founders. Like I worked very closely with Clem, There's other people like Thomas Wolfe, Julian, who are so smart and interested in seeing other people succeed. And they often are investing and mentoring founders. I think it's, you know, you sort of ideally fan out from one core hub to just sort of meet more and more people and to keep investing. one thing it doesn't, which is just very good technical people who want to learn or are doing it because they want to see something amazing built. So we're hosting a conference in a few weeks, thesis, and it's about the future of work and the future of kind of technology, really. And I'm curious, you know, based on what you've learned as the first investor in Hugging Face and many other really incredible companies, what your thesis is today. And, you know, we were talking before the pod about... how there's this discourse of, you know, some people say that in, in the future, uh, in a future of AI, you know, it makes sense to be more technical, to be less technical. Um, but you actually had a really interesting take on, you know, kind of specifically for founders and building a company, what the core, um, skill might be. Could you share a little bit about that? Sure. Um, And I want to add, I guess, that the Hugging Face team has built something amazing. There's so many other companies that are built by technical founders like Factory and Modo that are fascinating to me because the speed at which they're growing is so fast. And I'm thinking about why the speed is so quick. And one possible thesis is that I've been reflecting on is if the, if coding is becoming faster, right? Cause we can use AI and helpless code, then it's been sad. And I agree that what you build matters more, but the, what the job product is, I don't know if it's product manager or product person, is really to see the product in the wild and intuit from user behavior, whether that's the data or sitting, watching over someone's shoulder, how they're using it so that you can iterate, right? If like iteration is the most important part of... finding product market fit and probably also of growing, then if it used to take two months to see the feedback, to get the insight, and then to put together what the product plan was going like and code it and then get it back in front of the user, It might take two days now instead of two months. And so does that mean that that person, who is I think a very reliable person, whether that person codes or not, but that product person who is really, their job is to understand the user and then to be able to translate that, whether it's by coding or by, but I will have a partner to build the next version of it and to iterate. The iteration cycles almost are necessarily so much faster. And so it's fascinating to see a company like, you know, Hanging Face took 10 years to get to the point it's at. Then you have a company like Modal, which didn't, they exist a little bit before I started Factorial, but not much. The Factory AI, I think that we started probably at the same time, it was only a few years. And these companies have grown so quickly, and I wonder if part of it is the ability to build and iterate. allows these very, very fast cycles. And so the result in my mind would be that the ability to generate product insights, forget about what the job is called. It could be called an engineer. It could be called a product manager. The ability to synthesize how a user is using your product and then to improve it, I think has got to be exponentially more valuable. It was already the most valuable thing, is my guess, and now it is exponentially the most valuable thing. Obviously, distribution matters more than ever today as well. There's a whole bunch of noise, and the easier it is to build, the more noise there is. But I suspect that that's... That's a job whose value is going up tremendously right now because of the speed of it. Iteration as the core sort of variable to deliver speed and make sure that you can- Your cycle looks like this and you compress it, so your cycle time is like this, and so you just grow much faster. Then your work just gets compounded as a function of time. You know, I'm just curious, how do you filter for that when you're talking to a founder? When you're, you know, when you've met a compelling, you know, technical founder, they're working on an interesting problem. They have a foot that is pivoting. Then how do you measure that they can iterate quickly? I mean, usually when I'm meeting them, they don't think they're pivoting, right? Because they're like, no, this is definitely the thing. I'm raising money for it. But the... I think that's why the ability to do it is probably more important than what the things you're actually building. So usually I'm trying to understand what is their insight. And people have talked about earned insights. I really like that framing. It's like what is the – I think about that as how many people were in a position to learn the thing that you did. and then to execute it on it. Because I think that's what earned means. We can all take a look. If there is some observation that we all have, there's going to be more effort put towards solving that problem, which means more competition. If only 10 people in the world have come across this because it's a particular category or particular area of expertise, you're just going to have more time probably and both more time because you'll be ahead. But also, you're just... you've already spent a lot of time learning that problem and you can go after it directly. And so when I think about like an earned insight, I think that's important. And usually that kind of goes hand in hand with the ability to generate new insights as you are in market, which is again another reason to have your product out there. I think there are people who can like just divine what the right thing is. I think there's a story about how Figma like was just basically building in, I don't know if it's true, but was building like sort of alone for a year or two. that's not where I've seen company. That's not the pattern that I've seen. It's more common that you put something out there and it's pretty cool. And something starts to, to take off and you just try to delete everything else and, and do that bang because you have been able to generate this before you get that first version. you know how to do it going forward. So I'm looking for the, I think often it comes down to kind of the why, why, why are you, why is this the interesting, why is this the problem to solve? Or why is this a problem here? Yeah, that's, that's, and maybe the anchor example, the good one. Somebody who's thinking that deeply about how your voice sounds in your own head is probably a person who's going to really, really sit with their users and try and understand what problem they're solving. And is this Mike Magnano? Yeah. Yeah. The best. So great. Shout out to Mike. Yeah. I don't remember which one said it. And again, I think I've talked about this. I think, I don't know if you remember saying it, but I just, I was thinking, wow. But like a thing that they did do that I'm positive to remember is there was different onboarding if you had headphones in or not. Wow. And like, think about whether you turn your volume on. It's kind of different now, but pre AirPods, you can just like have them in. you would like shut down an app if it just started playing audio. So to have an app, how can you lower the friction of when someone's going to sign up? Letting them do it without headphones, they don't have headphones in is interesting. It was also like a very cute way that they showed the headphones. But that kind of care, I think, pays dividends in a consumer product. Yeah. Something that I often think about with really great companies is that they end up being – kind of like the greatest expression of someone's like best self. And I actually think about this with you and your fund because, you know, you have this mantra to invest in software. You need to know how software works. And you are the biggest tinker I know. Every time we catch up, you're working on something, you know, wild or whimsical, or like, you know, you build software for your own fund. Can you speak a little bit to some of the things that you've built and that you are using? Sure. And so Becca, who works with the effect trail is also a builder. And so she I often build the mediocre version that she builds like a really good version. there are things we use to manage the phones properly, but also I'm building to test and learn. So a product that I built that got a little bit of attention earlier this year was called Ghost Pepper. There's these apps, and there are many of these things that are used to, you hold out a hotkey and you speak. Like a lot of by coders just want to speak the prompt. And so I was using it and I was thinking, you know, I'm seeing all these local AI models for voice attacks. Why do you even need to pay for inference for this thing? Couldn't you just do it on your own machine? So I went to Claude Code, I think, and said, look at the hugging face repos. Here's a couple that I know about, but find whatever you can. Um, and I want to build something like, uh, like this. I use this skill, which I love called superpowers, which I think for any product developer, if you add superpowers, I think it's available for both Codex and Claude. I was using it in Claude. Um, it's, it's by this guy, Jeff, uh, I'm sorry. Jesse Vincent. I, yeah. It's probably this guy, Jesse Vincent, who is like a genius prompt engineer, I think, is how it works. But the difference between using this product to build things and not is night and day, in my opinion. So I used superpowers, and I had Claude do a bunch of research with me on voice models. I told him what kind of computer I had. And then I built this voice-to-text network. tool that I called Ghost Pepper, and put it on Hacker News, and it ends up going to number one on Hacker News, which I'd never posted anything to Hacker News. I was really excited, but I open-sourced it. So anybody, because you don't have any inference costs, so why not let anybody just try this and run it, and I can run these things for free? And it was an interesting lesson both in I learned you could build a very high quality product. A lot of people shared out that it was actually faster than a bunch of state-of-the-art products. I think probably because it was just constantly looking for the newest, fastest models. And particularly with voice detects, I think local works really well. The other thing I learned was when you just vibe code something for yourself, you have to maintain it. And I started to appreciate, there's all this talk about the SaaSpocalypse. these paid apps aren't that expensive. And in a way, what you're paying for is them making sure it stays a great product. As long as these, I think that the job to be done in a way for SaaS apps is to always continue to iterate and be at the edge. But I now have a different appreciation for these SaaS products that really do release all the time. And so it's definitely a fun project. I think, I don't know, there's like, Thousands of people are using it, which makes me happy. Yeah, I think I saw it has like over 3,000 GitHub stars. That's so cool. I ran this by some of our engineers, you know, and definitely the maintenance piece is just so real. One of my earliest learnings of this is I vibe-coded our CRM tool and used it for months. And then I realized at some point that I was spending too much time doing the maintenance of the tool that I had built. And, you know, instead I could be spending time with the clients that we have, you know, working on AI implementation and the work that we do on the consulting side of the business. And I realized it just made more sense for us to migrate the data to Atio, a product that just does CRM and data management, instead of me trying to play, you know, sort of like... engineer, you know, on the side and try to like maintain this tool. So maintenance is a really, really big part of it. I know this is something that our engineers think a lot about and monologue, which is one of these, you know, text, voice to text tools that we've built in house by one of our great engineers, Naveen. He mentioned that this is certainly one of the competitive edges. And with Monologue, for example, latency is something that they think a lot about. So I think it's like 800 milliseconds of latency for it to load. So certainly a trend that we're following. So cool that you've built, you know, Ghost Pepper is something that you have. We're finding these great skills. We'll also include Jesse's skill as a follow-up to the pod, as a resource for anyone that's listening to this. The other thing we were talking about is sort of the theses that you have now, right? So you had a thesis and some themes that you were exploring in 2015 that led to investing in Hugging Face. What are those themes or theses that you're exploring today? I think that so much is still infrastructure. It's amazing to me that these developer tools, for example, you should give us a perfect one. I think that the maintenance, there are going to be products that help you with just maintain all of this code. I've seen a few. So I think that there are, kind of think about the continuation of developer tools. Different people are becoming developers, which probably means there are different kinds of tools. I continue to think about frontier non-language models. So I think I said earlier, if Transformers are really good at next token prediction, What are the other areas? It turns out that LLMs have gotten much, much better with more compute. And so if that's the case, what are the other areas where you could predict time series and have it create genuinely new... value. And then, you know, whether that's my, you know, my, an obvious example is robots. What are they? What is a robotic movement? If not a time series information with like the locations of a bunch of service. And so that's already being people are doing this where they're, they're we invest in a couple of companies with sort of the general thesis that the way that we train robots could change. And people talk about whether it's a chat GPT moment, the data is a bit different, and so we'll see how it kind of works, but I think that's a category where transformers in terms of AI more generally seem to be really applicable. I'm fascinated by the science models. There's Hugging Face has got a hackathon they're doing called Rare Disease Real Kid, where they've open source of genome, and I think doctors sort of notes the parents of a kid did have a very rare disease, I think only 50 people have it, and they have a track for diagnosing it, but they also have a track for drug repurposing. So people talk a lot about drug development, which is I think interesting, but drug repurposing is another one that's kind of fascinating. And it's interesting to me that the people who are building this up are really technologists. As I said before, it's kind of programmable. And then I think the application layer, and my thesis around, everyone talks about what will be vertical allocation, will be consumer applications. You said something earlier about latency. I believe that the applications are gonna look a lot of the people who build the applications, even if they look like consumer applications or verticals, the people who build them are going to be technical enough to be able to go fully down the stack. And the example I keep thinking about is in, when Instagram era, there were a lot of photo sharing apps and Instagram did a bunch of things, right? If you look to the phone, it was very obvious that you had a camera and you had a cell phone connection. What was not obvious is that the load time mattered. So if you're standing waiting for something and it takes you 30 seconds for the app to load, like you're done. So a lot of companies, Zynga and Instagram, spend time really making load time an important thing. Instagram, I have read, and I assume this is true, you have the pictures in squirters. And a side effect of the picture being squared, I don't think it was on purpose, but the benefit is that you don't have to ask the server if you should show the image and portrait last it. And so if that shaves off a millisecond of load time, the team that has the ability both to do a great design that has that nostalgic feel as early Instagram did, but also go all the way down in the stack to be able to shave seconds off of server time, leveraging a decision they made on the design side. I think there's a lot of parallels to what AI native companies are going to look like in the future, where they look at a paper about oh, this is an interesting new way to train on reasoning for language. What can we pull from that and apply to images? Or what can we pull from that and apply to what we're doing? And both the ability to build your own models, but also the humility to know that these other big company models are getting great. And if we can just use them for the stuff we need to use them for, we can then spend all of our time doing the things that are more unique to us. being able to balance that and go full stack means something different now, I think in an AI native world, but I think being able to do that is going to be a real edge. as we think about the future and along these lines of the themes that you're excited about, the companies that you're tracking, and as an investor to so many of these compelling companies, what is some advice that you might have for founders who are trying to navigate this really dynamic environment, if you will, where the technology is changing so quickly? What are you seeing really great founders doing right now and what advice would you have? the best people that I've seen building things are like playing. We are in an era that reminds me of early internet. Everyone is playing. There are so many fun tools. Like I tried this text to CAD thing. I don't have a 3D printer, but I was like, I bet you could vibe. How good of a job? Could I vibe, make a robot? We're not there yet, I don't think. You can 3D print a robot and you can vibe, make some 3D structures. But probably we will be. And the best thing, I think that the way that you get the best ideas is by just playing and tinkering with this stuff and seeing where the problems are and then solving them. And, you know, free this, free 2016, like pre the AI era, it was really the mobile era. And there was a lot of fun things to play with. And then there was this... question mark period where it was okay the phone is at scale lots of people know how to make iphone apps distribution getting harder and ai hadn't quite um become what it is today which i think is a i thought ai was going to be sort of at least as big as mobile i think it will turn out to be which I think is, because mobile is not quite as big as that to me. And then I think it'll be much bigger. Again, for the kinds of companies that I am most excited about, playing with this stuff, finding out what's going to be possible and seeing where those gaps are and then start solving your problems, whether it's probably your own life or a problem for a friend's company, whatever it is, it's how I would be spending my time. I think there are maybe the same number of problems that have always been, probably more, but there's definitely more solutions to existing problems than there have ever been. And this solution space is so untouched relative to... I'm surprised at all the companies that aren't being built. I think they will be built. I mean, there's as many products as there are getting built. And so I think I used to play... I think it was World of Warcraft or something, and you would move around, and the whole world was dark, and then as you moved around, you got to see the world. I think we're in that right now, where everyone, even the people, it seems like even people building these AI models are still sort of exploring what they can do. And we're all kind of exploring what's possible and... Hopefully, I think our focus certainly is, what are the solutions that are newly possible? Maybe it's to existing problems, maybe it's to brand new problems. I think that there are numerous. So it's an amazing time to be exploring that and building new products. It really is so fun. Talking about games reminds me of one of the companies that spun out of every, led by Alex Stuffy, who... literally built a gaming company because he was basically testing the capability of models, playing a game against each other and playing different games against each other. And, you know, it really does come from this, like, spirit of exploring what's possible and then just seeing how far you can take. you know, where that leads you. Obviously, the best way to explore what's possible is to listen to every so that you know what's newly... There's never been a time where there's more newly possible stuff every week. So that's a personal recommendation. Great. You're on the team now, Matt. Welcome. All right. Well, maybe just to close, Matt, any resources that you're really excited about reading, you know, that like, you know, either it's people that you're following, you already shared Jesse Vincent's superpower skills. We'll link that below. Anything else that you're kind of like looking at day to day that's really helping you stay up to speed with everything that's going on? I mean, there's so much that's shared on X, I think is actually really, there's a lot of noise, but there's a lot that's really good. I look at what's trending on Hugging Face. I am fascinated by open source. I try to find any resources or open source. There's newsletters that I like. I think... Ben's Bites, I think is great. I think every newsletter is great. One of the reasons I've been more and more sort of reading X again is that people are writing these longer form content pieces. It seems like medium content sort of moved over to X and like you can take this stuff and put it into Claude and then take how can I apply this to my life or to my investment thinking or to my company or whatever, or to my, how can I take this insight about memory and incorporate it on a mask? It's fascinating. Um, so I think I kind of probably have a barbell approach. I have what's new, new, new. And then I read things that I forget. Uh, I forget what it's called. The rule where the longer something has been around, the longer it'd be around for. And so I'm reading, um, just things that are older, you know, I've been reading one book about like the fall of the Roman empire and things that are hopefully timeless, um, and because there's probably feels like we're in a moment where certain things are repeating themselves and, uh, whether it's technology or whether it's culture or human behavior, uh, I find it useful to look at some of the older stuff too. I know that we've talked about some of that as well, that looking at adjacent, not just only spending time looking at the most recent technology can often end up being informative technology. Yeah. Thank you, Matt. It's so fun to have you. You're one of my favorite thinkers and investors. Where can people find you? Matt Hartman on X. You can email me at matt at factorialcap.com. And yeah, thank you so much for having me. It's been fun. So true. Yeah. Thank you. Such a treat. All right. Thanks, Matt. If you liked this video, you should check out every.to. We have a two week free trial for anyone who's watching or listening at the link in the description. Every is the only subscription you need to stay at the edge of AI. It is the one subscription that if you're super AI pill and you want to learn how to use it to do the best work of your life, it's got everything you need. It has education and equipment. On the education side, we review every new model that drops and tell you what's good. On the equipment side, we have software tools specifically made for you, like an agent that helps you AI pill your whole company, all bundled together under one subscription. So get two weeks free now at the link in the description.