GPT-4 and professional benchmarks: the wrong answer to the wrong question

Absolutely correct.

We already know this is about self-driving cars. Passing a driver's test was already possible in 2015 or so, but SDCs clearly aren't ready for L5 deployment even today.

There are also a lot of excellent examples of failure modes in object detection benchmarks.

Tests, such as driver's tests or standardized exams, are designed for humans. They make a lot of entirely implicit assumptions about failure modes and gaps in knowledge that are uniquely human. Automated systems work differently. They don't fail in the same way that humans fail, and therefore need different benchmarks.

Designing good benchmarks that probe GPT systems for common failure modes and weaknesses is actually quite difficult. Much more difficult than designing or training these systems, IME.

"SDCs clearly aren't ready for L5 deployment" Apologies for the tangent to the OP topic. The metric to watch is 'insurance damage per million miles driven'. At some point SDCs will overperform the human driver pool, possibly by a large margin. Wouldn't that be the point where SDCs are clearly ready for L5? Not even sure if that point is in the past or the future, does anyone -- not named Elon ;) -- have reasonably up-to-date trend charts and willing to share?