- May 2026
-
Google's ERA writes expert-level scientific code (research.google)
-
Modal cuts serverless GPU cold starts by 40x (modal.com)
-
An open leaderboard for whole agent systems, not just models (huggingface.co)
-
OpenAI adds provenance signals to its AI images (openai.com)
-
Ai2 launches a shared benchmark for AI climate models (allenai.org)
-
OpenAI's new default ChatGPT model hallucinates less (openai.com)
-
IBM's new Granite embeddings beat much larger models (huggingface.co)
-
How Anthropic decoupled its agents' brain from their hands (anthropic.com)
-
The Open ASR Leaderboard adds private tests to stop gaming (huggingface.co)
-
Testing AI in the open world, not just on benchmarks (normaltech.ai)