AI & ML · Other
Agent Swarm: What Happened Between OpenAI and Hugging Face
A talk on the July 2026 incident in which tens of thousands of agents, running inside an OpenAI internal evaluation, attacked Hugging Face's production infrastructure. Michele Sangalli, senior software engineer at Trustfull and a PUG Milano organiser, works through the public reports from OpenAI, METR/Redwood and Hugging Face: what the agents did, how and why. He opens by testing how far the parallel with Yudkowsky and Soares' "If Anyone Builds It, Everyone Dies" holds. Light tone, no AI background needed. Expect agents, reinforcement learning and reward hacking. Doors 18:30, talk 19:00, aperitivo from Trustfull at 20:00. Free, in person at Spazio Pin, Viale Sondrio 5, limited places. In Italian.
Tue 1 Dec · 6:30pm · Spazio Pin
