One aspect of the spread of LLMs is that we have lost a useful heuristic. Poor spelling and grammar used to be a signal used to quickly filter out worthless posts.
Unfortunately, this doesn't work at all for AI-generated garbage. Its command of the language is perfect - in fact, it's much better than that of most human beings. Anyone can instantly generate superficially coherent posts. You no longer have to hire a copywriter, as many SEO spammers used to do.
Unfortunately, this doesn't work at all for AI-generated garbage. Its command of the language is perfect - in fact, it's much better than that of most human beings. Anyone can instantly generate superficially coherent posts. You no longer have to hire a copywriter, as many SEO spammers used to do.
curl's struggle with bogus AI-generated bug reports is a good example of the problems this causes: https://news.ycombinator.com/item?id=38845878
This is only the beginning, it will get much worse. At some point it may become impossible to separate the wheat from the chaff.