We are so concerned with ChatGPT making things up but have folks not read HN or reddit comments? People believe made up stuff all the time or make stuff up even without negative intentions. A lot of web content is fictionalized nonsense in general.
Absolutely correct. That is why it is extremely important to be extra skeptical now since now anything from text, images, videos, etc on the internet can be AI generated.
It is only going to get worse from here. Take everything on the internet with a grain of salt.
> A lot of web content is fictionalized nonsense in general
Always funny when you read a Wikipedia entry, and then scroll to the footnotes and click them all. Most of them 404. Some of them are citing some blog post written by some non-credentialed hack and the blog post doesn't even have citations, it's just made up on-the-fly. Then there's citations which lead to a scientific paper, and when you read the paper, everything said is 'backed' by more citations, and those citations 404/don't exist. The whole of Wikipedia is a lie.
People understand that ChatGPT / "AI" is going to be treated as an infallible oracle. Its anonymous inscrutability makes it the perfect Voice of Authority.
Theres also a subconscious fear that everyone will retreat into their very own filter bubble consisting only of themselves, and their AI agents trained to keep their bubble intact. So we want the AI to be worthy of the task.
It is only going to get worse from here. Take everything on the internet with a grain of salt.