GPT-6 Astra scored 95% on a robot control task, user reports

A user reports GPT-6 Astra scored 95% on a robot control task versus Fable 5.1’s 40%, with 6.2x fewer output tokens at 2.3x lower cost. Another user quoting the result calls it an example of why Astra is not artificial general intelligence (AGI). Their critique draws a comparison with visual perception demos that use brightly colored, simple shapes on simple backgrounds to make objects easier for the system to handle.