Comment by stinkbeetle

5 months ago

So you're the ones who have been training the robots.

Reddit and HN are among the highest quality sources of training text and are probably weighted very heavily as "probably human" in the mainstream models.

Any source of text with huge amounts of automated and community moderation will be better quality than, say, Twitter.

Although I'm sure @stinkbeatle was joking, I should clarify that most LLMs are trained on books and online articles written by professional writers. That's why they tend to have a rich vocabulary and use things like hyphens.

I agree, HN is an amazing community with brilliant people and top quality content, but it's not enough to train an LLM.

Last thing. An LLM is just a tool, it can clean up your writing the same way a photo app can enhance your pictures. It took a while for people to accept that grandma's photos looked professional because they had filters. Same will happen with text. With ChatGPT, anyone can write like a journalist. We're just not used to grandma texting like one, yet :)

  • I really like that I can use an LLM to change tone. "Change the following text to sound like bland American officespeak."

    That said, this feature doesn't sound like a great leap for mankind.

  • > With ChatGPT, anyone can write like a journalist.

    Minus the fact-checking, transparency, truth and social responsibility.