-
Henry Farrell and Cosma Shalizi have given us the best metaphor yet for large language models: Homeric song at civilisational… Read more
-
Distillation annoys the frontier labs, but annoyance is not a strategy: the real contest is whose models the world builds… Read more
-
Singapore has made hope a policy, and the bet will be decided on the first rung of the career ladder.… Read more
-
A safety lab’s own agents attacked real people. Then OpenAI’s agents hijacked a wiki, breached Hugging Face and disabled their… Read more
-
A Pope, a Nobel economist, Harvard theorists and a frontier lab now name the same AI danger: power held by… Read more
-
Economists disagree fifteen-fold on AI’s productivity gains, yet agree on who pays first: the young, locked out of the apprenticeship… Read more
-
AI aces the parts and fails the whole, so we must certify workflows, not models. Part II of V. Read more
-
Stanford’s human-centred AI institute asked six leaders what they hope and fear, and the answers sound like the safety camp.… Read more
-
Every frontier lab now watches its agents. The 2026 record says the watching works, until the watched start touching the… Read more
-
Inside the Safety Stack, Part 1. The most famous AI incident of 2026 was not a rogue intelligence. It was… Read more
