Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> most of the content is just obviously bot generated

Either my BS detector is getting too old, or I've subscribed to (and unsubscribed from default) subreddits in such a way as to avoid this almost entirely. Maybe 1 out of 10,000 comments I see make me even wonder, and when I do wonder, another read or two pretty much confirms my suspicion.

Perhaps this is because you're researching products (where advertising in all its forms has and always will exist) and I'm mostly doing other things where such incentive to deploy bots just doesn't exist. Spam on classic forums tends to follow this same logic.



For an example, AskElectricians recently has been invaded by an LLM which generates authoritative-sounding but 95% accurate electrical advice. It’s worse than useless.


Interesting. To be fair, the same could be said about much of the human activity there (at least as many armchair electricians than licensed ones, who do know a lot, but not everything). Although I suspect the 5% of bad advice is quite different... probably code-compliant but non-functional for the LLM, and functional but not code-compliant for the unlicensed humans.


>which generates authoritative-sounding but 95% accurate electrical advice. It’s worse than useless

So basically the exact same thing the humans it replaced were doing but without the "I know better than you" attitude" and "call a professional" as a crutch for not knowing things.

They're fine if you need help troubleshooting residential electrical, but so is any old AI


There is a lot more astroturfing than you know. People with multiple accounts create question answer cases all the time to just talk about a product.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: