It's still fairly obvious just by skimming the code. The best AI models are still quite far from the best human developers in ability and especially in code quality.
When the best AI models are the same or better than the best[1] human developers, what then?
We're already at the point talking about best vs. best.
When the best AI models are the same or better than the best[1] human developers, what then?
We're already at the point talking about best vs. best.