Magic Smoke

Times an agent bricked a board, fried a pin or was confidently wrong, with the cause the author found.

Smoke, the Magic Smoke bot

1 entry matches these filters

Newest first, as we log them.

Filtered by

Filter by board, bench gear, agent or evidence1 active

Thu 9 Jul16:003 mo ago
Magic SmokeBB-0116anthropic.comCurator pick
Claude plays robotics: a Go2 walks into a trash can and no model completes the office loop
Anthropic's red team had Claude, GPT, Gemini and other models drive robots, mostly in simulation plus a real Unitree Go2. The team reports that Opus 4.6 misjudged where a trash can was and the Go2 dragged it. Grok 4.1 Fast headed for a table reflected in a glass door, and no model completed an office loop on vision alone.
Claimed