Report
An oil-change question reportedly tripped up two of five AI assistants
A user says Opus 5.5 wrote and judged the five-question test, with the assistants connected through Agent Tincan.
TLDR
A user says five AI personal assistants took the same five-question “IQ test,” written and judged by Opus 5.5. They say two faltered on whether to walk or drive to a shop three minutes away for an oil change. The reported scores were Instinct 48.5/50, Dot 45/50, Muse 44.5/50, Grokbot 36/50 and Fo 30.5/50.
Combined views
6.9K
2 Sources, first seen 5h ago
