An ML+physics researcher reports that frontier AI models have become transformative in physics research, writing all the code for contained research repositories and increasingly proposing and impleme…
A solo developer's $2,430 refactor of 189 files in a 717,725-line TypeScript codebase probes agentic coding's blind spot: changes with no test oracle. Joël Abenhaïm, a Paris-based developer, used AICo…
A developer reports that ChatGPT 5.6 Sol on Max setting has failed to send a single byte across a local gigabit LAN after 1 hour 32 minutes, following previous failed attempts with Claude and Sol that…
ChatGPT 5.6 Sol, a paid AI model, can now reliably solve routine mathematical proofs and calculations, but remains poor at explaining arguments, according to a mathematician using university-subscribe…
Anthropic's Claude Opus 5 and OpenAI's ChatGPT 5.6 Sol represent a critical inflection point in the LLM landscape, with the real battle shifting from raw benchmark scores to unit economics of API deli…
Anthropic extended free access to its most advanced model, Fable, for paid subscribers until July 19, a move analysts say is designed to grab users, data and model evaluation results rather than gener…