Hacker News new | ask | show | jobs
by airstrike 629 days ago
> If a machine spits out some plausible looking text (or some cookie-cutter code copy-pasted from Stack Overflow)

This is not a reasonable take on the current capabilities of LLMs.

1 comments

It’s certainly been my experience with the technology.
But if nearly everyone else is saying this has real value to them and it's produced meaningful code way beyond what's in SO, then doesn't that just mean your experience isn't representative of the overall value of LLMs?
It could also mean a lot of people are misattributing their utility.