Human coders are still better than LLMs

Re:Human coders are still better than LLMs by allturtles on Thursday May 29, 2025 @02:49PM (#44129003)

I think it's the wrong analogy. The prompt engineer who uses the AI to make code maps to the poorly-paid, low-skill power loom machine tender. The "engineer" is the person who created the model. But it's also not totally clear to me that we'll need humans for that either, in the near future.

[ Parent | on HN ]

Re:Human coders are still better than LLMs by 91bananas on Thursday May 29, 2025 @03:18PM (#44129271)

Not all engineering is creating models though, sometimes there are simpler problems to solve.

[ Parent | on HN ]

Re:Human coders are still better than LLMs by ori_b on Friday May 30, 2025 @12:52AM (#44132978)

At some point, we'll probably have these things write their own prompts, too.

[ Parent | on HN ]

Re:Human coders are still better than LLMs by Jolter on Wednesday June 04, 2025 @04:13PM (#44184982)

Still, the loom tending engineer got paid better than the weaver. Well, my prejudice says he did, don’t actually have the historical facts.

[ Parent | on HN ]