The daily AI briefing people actually read.

Tencent says its new model, with open weights promised, trained a small AI that beat the original on six of eight tests. Its own charts show how far it still trails OpenAI and Anthropic.

MIT researchers had OpenAI's model rewrite 0.016% of a small Qwen model's weights, with no training, and its math score jumped from 7.9% to 48.1%. The same day, Anthropic told investors it may soon reach its own threshold for AI-accelerated progress.

Three weeks after hundreds of French stores were hit, suspected hacks reached fridges and freezers in at least seven European countries. Nobody has claimed them, and no evidence links them to AI.