- Joined
- Jan 28, 2018
If they're so worried they should post on kiwifarms, I'm pretty sure nobody training LLMs is scraping this site.
LLMs are actually extremely powerful shitposters. LLM text full of GPTisms is mostly lack of skill in regards to the person who set the bot up. I give us another year or so and then even pajeet will have figured out how to make an LLM sound natural in a tweet. When platforms then use that data to train their LLMs, (detecting AI text being utterly impossible) LLMs and their writing will start having serious influence on the culture without anyone even realizing it because we humans as herd animals will just fall in line. It is horrifying. But also a bit funny.AI will be able to write retarded shitty takes as well as they do?