Lessons from Building Open Source Libraries

Key Takeaways

I'm excited to welcome today Thomas Wolf, the co-founder and chief science officer of Hugging Face. We're here today at beautiful San Diego for the Europe's 2025 conference. So, let's get started. Thomas, you had a very unusual career path before you became the founder of one of the best open source AI companies. You studied originally physics, then even did law and then started uh hugging face. Tell us about how that journey shaped eventually hugging face. >> Yeah, I mean even in physics. So I studied I was at Berkeley. I was working on laser fusion interaction in in some of the team I became the team that did the fusion experiments liver and then I worked on super connecting material. So even the physics part had multiple life. I guess one of the thing for me was always to try to work with people I wanted to work with even more than what I was specifically working on. So that was one of the defining moments and then I just think life is too short to do just one thing right. So once you once you once you did once you do one thing for six years like so I did PhD post and I was like I want to do something else. I really love writing. I was like I I'm very interested in law why not switching? So I switched to became a lawyer. Very different type of things like your time you have a you have a an amount you know your time cost $300 like you start counting the hours totally the opposite of PhD where you spend your hours on like nothing or like falling rabbit holes for two days or something like that. So very different. I think each of these kind of touch me taught me something in retrospect like like like science is this this idea that you can go really deep in one topic and you can actually explore these things to build. low is this idea that you for me was this idea that my time is valuable and I should have signed it very well and then and then hugging face was kind of by uh by accident. I I fell in entrepreneurship kind of like I needed a job in the US basically at that time. So this company was being created which was a game company and I joined and I just started this exploration deep learning that became this opens library that went viral and then we decided to pivot the company around this open source library and then we kind of found our mission along the way that building this idea that community opensource open science there's kind of a mutual win-win that's possible in the AI world instead of just racing and there is this idea that distribution of power is actually extremely powerful and and power catalyzing ecosystem instead of just cannibalizing them which is what you try to do. We try to be this platform on top of it people can build billion dollars company I think is very very uh exciting. >> That's very cool. I mean following this you've been one of the first proponents of really open research as opposed to the efforts that the big labs do. They keep it all private. What do you think are some of the things that the open-source community does a lot better to push AI forward? What are the strengths and also what are some of the limitations? >> Yeah, I think open source is probably the best thing that computer science brought to humanity. It was created really by you know this like couple of like deep hardcore computer science people against a lot of other possibility. Um I think this is even something that I want to see applied in in much more fields of research. I think there's many advantage. Uh there is obviously the collaboration aspect which is like if something is open source in particular code you can try to tweak it and like the way AI research is is progressing is still like something take this model and try to change positional embedding of things. So you need something to start from. If every day if every time you need to reinvent the whole model because everything is closed it's just much lower. So it it catalyze accelerate progress but then more than that I think it it's allowed to explore many areas. So just before the interview we're talking about interactive world model and gaming and how gaming can re be reinventing. The only way to really you know be able to do that or one of the way at least is you take a very good pre-trained model for instance an image and you just add some interactivity component and that's a very quick way to actually explore you know alternative use that maybe the person who trained the model didn't even thought about right so there is this kind of opening of possibility like the model is created it's a very strong base you basically people have distilled in it like a 100 million hours of GPU view and data and you can use that as a very powerful starting point to explore new use cases and if you don't have access to the model you've much more limited right just like TGPT you can basically use it for what OpenAI intended you to use it but if I want to use it outside of the training domain I don't know like very DSL specific DSL they didn't train of doesn't work basically so the only option is to have access to this and tweak it so I think It's really vital for exploration, creativity and and I think creativity like entrepreneur entrepreneurial creativity as well. Yeah. >> So one of the things now following on on this to the audience that are watching here a bunch of founders and builders. You've seen many people build on top of all these open-source models on hugging face. It can be easy to get a cool demo quickly, >> but it's actually hard to get something that actually passes the bar to build a good product. >> Yeah. >> So, what do you think are the things that the audience should think of? Things that look good but don't necessarily translate into real products for for users. >> I mean, to be honest, I think it's the same with closed model. So most of the company right when you want to build your product it most of the time you won't fall exactly squarely in this like good example where JGPT works and what happens so they start building the scaffolding all the time and say oh we need to pre-process we need to be careful about this edge case because now doesn't work and so most of the company I've seen building be on closour on top of open source they they have to anyway like in a way know their domain knowledge their domain really well and just like do kind of the productionalization of this model themselves and it involve lot scaffolding and hopefully as the model improve they're more reliable on a wider range and you can even like some of course you can sell your data and like hope that the next generation of model is going to be better because you get then more data to fine tune but I think we're still at this stage where it's quite rare that you can take a model out of the box and expect it just to work in production environment um the just the way you do it between close source and open source is quite different close source you mostly work with this kind of scaffolding I think and open source you have the option to fine-tune maybe to train it and of course the the associated drawback is that it's a non-trivial still like a non-trivial I would say operation to do so as a as a young startup where you're just like three you have to know you have to know if you want to allocate some of your time to fine tune the model or not and so I would say if if it's a core thing like like this interactive world model whatever you're fine tuning is a core thing because the feature doesn't even exist right but in any case you probably also want to to scaffold or or use I mean more and more by the way this uh fine-tuning I think is something that you can you can find framework that help actually do like tinker like a couple of new company emerging that try to help you do that so probably it's something you won't have to to fully do inhouse but yeah I think in any case there is no super shortcut from demo to production it's still a painful process but as part of the mode and the knowledge and like what you're building is also the pain also I mean that once you made it work you actually solve something non-trivial and that's where your company value lie >> so talking about pain >> you've been very good at delivering at hugging face these cutting edge research ideas and turn them into actually widely used tools and API especially with your role less the chief science officer. What is the hardest part about turning turning research into real products? I mean, we talked about pain. What what are some of the nuts and bolts of how to how do you get the best out of your teams and ship these products? >> I mean, in practice, I'm chief science, but what I really see myself more in terms of open source. So when I created transformer data set was kind of a chief product officer because this is kind of a product and your clients is basically the open source community and they're very difficult to please as clients because everything is free so they want actually the best. It's very easy for them to move from one to the other framework if they have to. So you really have to uh think a lot and I I think for me the key thing I've always tried to work a lot on is um I think there's probably two and they're both equally important. The first one is the kind of on boarding experience which is your first experience with a library. The number of abstraction you will have to learn to be able to master it. the like clarity like the the distance from I download this library to I run the first kind of ex non-trivial example of what I want to do which is this example where you're kind of wow you're like this is something I was not able to do before I had this library this should be an extremely pleasant experience in a way like just like you talk about this unboxing it's a bit the equivalent of unboxing a product so you want people when they come when they start the Apple experience >> it should be extremely obvious like the the abstraction you have you should they should be almost tangible and they look obvious to you. You should have very few abstraction because the more you have you each new abstraction you impose your user to learn is like a friction point. So this is very important and the only way to get this to work is I think to keep before you release or after before each time you add a feature to put yourself in the in the shoes of someone who is trying your library for the first time has no idea what are your abraction don't want to read the doc because no user of any software want to read the documentation so uh it shouldn't even have to write the documentation so everything should look really obvious. And it's it's getting harder and harder as you know your library more and more because you're you know you start to to to not have this fresh mind. So I think getting this fresh mind and so every time I prepare for release in the past I was always keep reiterating on this first step. I install I try this first thing. Does it look obvious? I install I try this first thing. So that's the first one. And then the second one is is where you put your level of abstraction. And there's always a cursor between um how is easy to use, how much control you give and how complex it is. And this cursor is something extremely hard to pinpoint really well. And I feel I feel like the only way is to try it a lot yourself on many use case and see where if you get this balance between flexibility, intuitiveness, right? Uh there's a lot of taste here. I think there's a lot of design and I think a good open source library overall is something that's uh extremely design opinionated. >> Now to close it off, imagine we get to the point which probably should happen at some point. The capabilities of open-source AI models matches those of the closed ones. So what should the world update about the priors of how AI innovation is done because so much funding goes into the private labs right how would the world change >> yeah maybe my take would be that I don't think this is so far off and I I feel like we had regularly this year this kind of challenging on the closest model yeah it started I mean the 2025 started with deepseek at the very beginning of the year saying boom and was the first moment the world realized oh there's this open source AI what is that and it's actually competitive more recently Kimmy Kimmy was also like really close to the state-of-the-art or not so far um I would say yeah for me the prior is that there is still probably uh I mean there's still value in training model I think and for sure this is something that will keep we keep continue pushing frontier but more and more I think the value and uh and that's also the direction of many frontier companies value will be the interaction the app layer with this model so the model is something but basically chipd is very interesting because the interaction friction is very low because it's a product that really works and even entropic has this type of aspect so this user interface and how you use the model in application or um application that exists on your application is I think where the value will really lie. So right now the interesting thing is that the the foundation model labs are also also attacking this but that's something you can attack without training models. So that's there's a lot of room for startups there to do amazing things >> which is awesome news for the audience. So what I'm hearing is that there's just so much more value to unlock at the application layer for models. If we're really now neck and neck in terms of foundation model performance, >> a lot of the cool things and value that exists is building the application stack. >> Yeah. >> Which is great for all the founders. >> Yeah. >> Great. >> 100%. >> Thank you so much for coming and chatting with us, Thomas. >> Thanks. It was a pleasure.