-: FOLLOW US :- @theinsaneapp
OpenAI acknowledges the necessity and usage of copyrighted material in developing AI tools like ChatGPT.
-: FOLLOW US :- @theinsaneapp
The acknowledgement came during a submission to the UK's House of Lords inquiry into large language models.
-: FOLLOW US :- @theinsaneapp
They use public internet content for training ChatGPT and DALL-E.
-: FOLLOW US :- @theinsaneapp
Copyright covers almost everything humans make, like blogs, pics, and code.
-: FOLLOW US :- @theinsaneapp
OpenAI says AI can't learn well without copyrighted stuff.
-: FOLLOW US :- @theinsaneapp
They argue that old books won't help make good AI for today.
-: FOLLOW US :- @theinsaneapp
OpenAI defends itself, saying they didn't do anything wrong.
-: FOLLOW US :- @theinsaneapp
They say using public internet stuff for AI is allowed by law.
-: FOLLOW US :- @theinsaneapp
This isn't the first time OpenAI has claimed fair use; they previously defended similar actions involving comedian Sarah Silverman,
arguing that copyright limitations permit innovations like large language models.