Founder @ Ordalie | https://teapawt.org/t.co/1tcHhyBaxW
Paris, France
Joined February 2019
- Tweets2K
- Following1.5K
- Followers593
- Likes19.9K
Baudouin retweeted
After working on it for a year, I can finally see the end state of the harness
On the backend, it's an execution environment that gives the model maximum degrees of freedom and on the frontend it's a free-form canvas that the model can use to communicate with the user
Baudouin retweeted
Introducing celld: a self-hosted, distributed Durable Objects and Workers implementation
- celld = V8 + S3 + SQLite + LTX+ Tokio
- programmed with exactly the Cloudflare Workers/DO JavaScript APIs and config
- order of magnitude cheaper at scale
- writes are durable before they're acknowledged, RPO=0
- only depends on an S3 compatible bucket - no other infrastructure
- ~4 MB memory per cell instance overhead
- hibernated cells cost ~nothing, just their S3 storage
- open source, Apache 2 licensed
Baudouin retweeted
might be late but my 2 cents on the “read your code” debate:
2024 > do your own research
2025 > write your own code
2026 > read your own code
2027 > think your own prompts?
take model development and its economics seriously, and start thinking about tomorrow’s problems today.
Baudouin retweeted
A post of ours on vendor lock in and LLMs. We don't like the growing trend of AI companies quietly hiding your data while stripping away your control. We think that's bad for users and bad for the ecosystem. Here's are our thoughts: earendil.com/posts/session-p…
Baudouin retweeted
The answer is deceptively simple imo.
There’s a notion of a general-purpose harness (like, say, CoT reasoning, proper compaction, or an RLM) that’s essentially just what *the* model should become. A better “model” in the API will eventually wrap it and train accordingly.
And a separate notion of a harness that’s last-mile specification over the underlying bland model. This is irreducible to the extent that your task/context diverges from the mode, regardless of capability. Here, the closer your complexity to the minimum number of bits required, the better.
Baudouin retweeted
Folks want other people to take responsibility. If you do it yourself, you give up that privilege...
Baudouin retweeted
I mean, yes if you don’t immediately see an issue with an eval ranking gpt-oss-20b higher than gpt-5.1.
Baudouin retweeted
Breaking: we release a fully synthetic generalist dataset for pretraining, SYNTH and two new SOTA reasoning models exclusively trained on it. Despite having seen only 200 billion tokens, Baguettotron is currently best-in-class in its size range.
Baudouin retweeted
Ordalie x Le Barreau de Paris 🤩
Merci @PierreHoffman_ et toute l'équipe @Avocats_Paris pour votre soutien !
Baudouin retweeted
anthropic’s thinking campaign is just so damn tasteful… it feels like a warm room. the aesthetic is comforting as hell, i kinda get a sense of claude as ai helping you be you from it.
that’s why we’re weaving claude into everything we build. i’ll post more about sonnet 4.5 later too which is just a ridiculous model.
i walked into air mail wanting to buy the hat & someone just told me it’s free & i was shocked haha. also got a free latte. great work @mikeyk & @alexalbert__. please promote whoever came up with this thinking campaign, it’s just so damn beautiful.
Baudouin retweeted
Bloombert GPT was really done two years too early. Plenty of techniques now to push small models much further than generalist ones on niche knowledge/logic.
