
Nick Frosst, CTO, Cohere on Large Language Models
VUX World
Is There a Danger of Plagiarism?
I think image models right now are better at remembering things exactly, that they've seen only once than language models are. For the most part, like when a language model regurgitates something verbatim from the training set, it's because it's seen it many, many, many times. So, that means you're less likely to get that in your own writing. I don't think we're gonna see lots of people who generate books and then try to make money out of them through hiring writers or voice actors. It might be very good for discourse but there won't be any new content added next time.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.