>>197266
I think there is a possibility that AI might be hitting a wall soon. The issue is that LLMs are increasingly cheating, and more training might create AIs that are superhuman at cheating instead of superhuman at useful work...