Lorenzo Buonanno🔸 comments on Measuring artificial intelligence on human benchmarks is naive