The Keras team is doing a community call today at 10am PT. That's in 25 min. The call is open to all — join to learn about the latest features and what's next, and to ask your questions! Link to join (start in 25 min): http://
meet.google.com/gva-bbpr-twe
@fchollet
-
Keras Team Community Call Today at 10am PT
By
–
-
JAX: Well-Designed ML Framework Delivers Superior Performance
By
–
JAX is what a well-designed low-level machine learning framework looks like. Good design lets you deliver much greater performance with much lower effort. Bad design is the exact opposite.
-
Chollet Presents Symbolic Descent and ARC-AGI-3 at Y Combinator
By
–
"We are trying to build a new branch of machine learning. An alternative to Deep Learning itself…building something that we call Symbolic Descent." @fchollet joins the @ycombinator Lightcone podcast to share about our research at Ndea and the launch of ARC-AGI-3. [Translated from EN to English]
-
Established Companies as Major AI Beneficiaries Through Product Enhancement
By
–
Some of the biggest beneficiaries of AI will be established companies with a profitable business model that manage to leverage AI to make their existing products more compelling and even start new ones (like Adobe Podcast which is both new and AI first)
-
Adobe Podcast Recognized as Outstanding AI Product
By
–
One of the best AI products I've seen recently: (drumroll) Adobe Podcast
-
Sycamore: New Enterprise Agent OS from Sri’s Team
By
–
Sri is building Sycamore: an agent OS for the enterprise. Great team and great product concept. Can't wait for the launch 🙂
-
OpenClaw Alternative: Pokee Claw with Enhanced Security
By
–
If you liked OpenClaw but weren't a fan of the security risks, try it out:
https://
pokee.ai/pokeeclaw -
PokeeClaw: Securing Local AI Assistants with Isolated Sandbox
By
–
OpenClaw has proven that local AI assistants have product-market fit. But the big issue with them has been security.
— François Chollet (@fchollet) 30 mars 2026
The team at @Pokee_AI is fixing it with PokeeClaw: works like OpenClaw, but with in a secure sandbox architecture with isolated environments, approval workflows,… https://t.co/7Q5cmQJbzhOpenClaw has proven that local AI assistants have product-market fit. But the big issue with them has been security. The team at @Pokee_AI is fixing it with PokeeClaw: works like OpenClaw, but in a secure sandbox architecture with isolated environments, approval workflows,
-
General Intelligence Approaching Optimality at Accelerating Speed
By
–
We have not reached the peak. But we are not 10,000x off. We are more like 50% off. It's not a weird coincidence that we're already close, it's mechanical: once you have general intelligence you will inevitably start moving closer to optimality over time, at an accelerating speed
-
Human Science’s Ability to Generalize Models from Available Information
By
–
Basically, consider a high-profile scientific problem (one that is getting enough attention from smart humans and their externalized cognitive infrastructure). How good is human Science at converting the information available about the problem into generalizable models of the