The scariest part is that humans are starting to use AI to generate spam comments, which in turn get used to train the models. Will the language capabilities of these models just keep getting worse?

Sent via haiker.app — My handmade Hacker News app