A New Benchmark for the Risks of AI by Will Knight from on 2024-12-04 14:00 (#6SPHM) MLCommons provides benchmarks that test the abilities of AI systems. It wants to measure the bad side of AI next.