I clearly notify where ChatGPT falls short. But accounts for both success and failures, still the level of ability is amazing. That's why I said benchmarking st

I clearly notify where ChatGPT falls short. But accounts for both success and failures, still the level of ability is amazing. That's why I said benchmarking standard, because we can't only cherry picking those failures while completly ignore the success.