Introducing Humanity's Sixth Sense: Measuring Intuitive Visual Reasoning
Scale AI, Wednesday, October 7th, 2026
Scale's Humanity's Sixth Sense benchmark shows top AI models score 53.6% on intuitive visual reasoning versus 93.1% for humans.
Scale Labs introduced Humanity's Sixth Sense, a benchmark for intuitive visual reasoning built with Elorian.
It contains 522 open-ended tasks across 288 images and 234 video clips totaling 17.6 hours, spanning temporal and causal, physical and spatial, social, and abstract or contextual domains.
Humans score 93.1%, while the best model, GPT-6-astra, reaches 53.6% at maximum reasoning effort and the median model 30.9%, a gap that agentic image-manipulation setups narrow but do not close.