AGI - AI Dynamics

Google développe un nouvel agent d’images pour Stitch

By

–

05 November 2025 16h29

BREAKING 🚨: Google is working on a new Image Agent for Stitch, likely to be powered by Nano Banana 2.

Integration with AI Studio and potentially with Lovable is coming as well, + a possibility to generate a project brief!

Double bananas 🍌 pic.twitter.com/ouhXyaD64p
— 🚨 AI News | TestingCatalog (@testingcatalog) 5 novembre 2025

BREAKING : Google is working on a new Image Agent for Stitch, likely to be powered by Nano Banana 2. Integration with AI Studio and potentially with Lovable is coming as well, + a possibility to generate a project brief! Double bananas

→ View original post on X — @testingcatalog

5 November 2025

Net Positive Beyond Voting: What Else Does It Signal?

By

@andreasklinger

–

05 November 2025 8h44

What does “net positive” say that 79% voted yes already says ?

→ View original post on X — @andreasklinger

5 November 2025

Autonomy Requires Embodiment to Seek Problems and Solutions

By

@fchollet

–

04 November 2025 23h57

If you want it to be autonomous it should be able to seek new problems to solve (and the means to solve them) on its own, which requires embodiment in some form (not necessarily physical)

→ View original post on X — @fchollet

4 November 2025

Autonomous AI Through Dynamic Model Synthesis and Self-Improving Abstractions

By

@fchollet

–

04 November 2025 23h36

The path to autonomous AI is a system that learns to solve new problems by synthesizing models of them on the fly (as code), and that gets smarter over time by adding new abstractions to its own library (also as code), compounding its capabilities. Not a static map — rather, an

→ View original post on X — @fchollet

4 November 2025

@testingcatalog — 2025-11-04

By

@testingcatalog

–

04 November 2025 22h57

OpenAI is gearing up to release a family of new models soon, as teased by Sam Altman. Soon or not soon?

→ View original post on X — @testingcatalog

4 November 2025

IMO-Bench: 400 olympiad problems for AI evaluation

By

@lmthang

–

04 November 2025 21h16

IMO-Bench also includes #AnswerBench, a set of 400 problems with verifiable answers carefully chosen from past Olympiad competitions. The problems span across four IMO categories (Algebra, Combinatorics, Geometry, and Number Theory), were altered by experts to avoid memorization,

→ View original post on X — @lmthang

4 November 2025

IMO-Bench: New AI Model Evaluation Framework for Mathematics

By

@lmthang

–

04 November 2025 18h31

IMO-Bench consists of three benchmarks that judge models on diverse capabilities: IMO-AnswerBench, a large-scale test on getting the right answer; IMO-ProofBench, a next-level evaluation for proof writing; and IMO-GradingBench, a new benchmarkto enable further progress in

→ View original post on X — @lmthang

4 November 2025