I'd like to see an AI leaderboard consisting of 100% LLM-generated submissions. These submissions could then be evaluated for quality and scored, using a new s

I'd like to see an AI leaderboard consisting of 100% LLM-generated submissions. These submissions could then be evaluated for quality and scored, using a new set of measures for code quality. We can also learn a thing or two about prompt engineering.