Short conversation about efficient integer encoding schemes: 2%
Today, could a squirrel get from the Atlantic Ocean to the Mississippi River without touching the ground (with attached forest cover map): 3%
Personal eval (which GPT 5.4, Opus 4.6, and Gemini 3.1 have all managed to succeed at): 60-70%
Short conversation about efficient integer encoding schemes: 2%
Today, could a squirrel get from the Atlantic Ocean to the Mississippi River without touching the ground (with attached forest cover map): 3%
Personal eval (which GPT 5.4, Opus 4.6, and Gemini 3.1 have all managed to succeed at): 60-70%