[D] Is it fair to compare AlphaCode to human programmers?

DeepMind's recent AlphaCode paper has triggered some bloated headlines that compare AI to human programmers. To be fair, DeepMind itself doesn't make such claims. Their blog post says it is "the first time an AI code generation system has reached a competitive level of performance in programming competitions."

Surely, DeepMind has done a terrific job at putting together the right pieces that transform natural language problem descriptions into working programs. And it has been able to compete in Codeforces …

alphacode human machinelearning programmers