At Hugging Face, a former journalist helps make AI more accessible
"Start with what people actually need, then design around that."
Florent Daudens led digital innovation in Canadian newsrooms before becoming press lead at Hugging Face, the open-source AI community. He talks about shaping his feeds to feel more like home, his move from journalism to AI, and why the best way to understand new tech is to start making things.
— Kristina Bravo, editor at Mozilla
What is your favorite corner of the internet?
That rare, quiet part of the internet that actually makes you smarter without making you feel behind. For me, it’s a mix.
LinkedIn surprised me. I used to think of it as stiff and self-promotional, but it’s become where I exchange ideas with people wrestling with the same big questions: What’s AI doing to journalism? What’s worth building?
[X] is still very relevant for everything related to AI news. It’s where I get pulled into weird, fascinating rabbit holes. Someone posts a half-broken agent demo or a wild paper, and suddenly I have 12 tabs open. It’s chaotic in the best way.
And Hugging Face of course, to keep pace with AI releases!
I think what changed everything was narrowing my feeds. Once I stopped trying to follow everything and leaned into what really matters to me – AI, openness, news and creative industries – it all started to feel like home.
What is an internet deep dive that you can't wait to jump back into?
My YouTube recommendations read like a personality test I didn’t mean to take:
obsessive AI build logs. I'm a sucker for "How I made this with that" videos to learn new skills related to AI.
Mandarin tutorials (six years in and still chasing tones...)
vintage French science shows that I now rewatch with my kid — equal parts nostalgia and wonder.
What is the one tab you always regret closing?
That post. You know the one — right under the other one. You meant to open it in a new tab, but you didn’t. And then the feed refreshed and it’s gone forever. A digital ghost.
What can you not stop talking about on the internet right now?
AI-generated videos that are totally unhinged and strangely beautiful. Like:
Or Total Pixel Space — the RunwayML AI film festival winner that feels like visual poetry
What was the first online community you engaged with?
CaraMail, back in France in the late ’90s. It was messy, anonymous, and kind of magical. That early feeling of connecting with people across borders, in French, about anything, was completely new. It opened up so many possibilities and shaped how I saw connection and community, and actually played a role in me moving to Montréal at 18.
If you could create your own corner of the internet, what would it look like?
Actually, I’m lucky; I am building it.
That’s why I moved from journalism to AI. I could feel something shifting, not just in media, but everywhere, and I wanted to help make this foundational technology open, accessible, and collaborative. As a former data journalist, I saw how open-source wasn’t just about sharing code. It was a force multiplier for learning, creativity, and community. With AI, that effect is even stronger.
So yeah, without a doubt: Hugging Face.
What articles and/or videos are you waiting to read/watch right now?
The LangGraph course on DeepLearning.ai on long-term agentic memory (it's niche, I know)
And a new series on MCP, which my colleague Ben kicked off, because I genuinely think this protocol could unlock a whole new layer of what’s possible on the open web.
What’s the biggest opportunity you see right now at the intersection of AI, open-source and public-interest media?
Small experiments, bold new tools, but most of all, building.
With AI-assisted coding, I think the barrier to entry is lower than ever. You can go from idea to prototype really quickly, even without knowing how to code, but just by starting with your words and ideas. And that’s a game-changer.
Take AI agents: the only way to really understand their potential and their limits is to try building one yourself. That forces you into the mindset that matters most: empathy. Start with what people actually need, then design around that.
Open-source supercharges all of this. It lets you remix, test, and share. It makes scaling faster. And maybe most importantly, it’s the best way to stay independent from tech companies. You’re not just using tools; you’re shaping them.
 | A guest post by
|