@hipsterelectron@circumstances.run
Post #3218056
2025-03-11 00:43 UTC
@micr0@fuzzies.wtf @esm@wetdry.world in the case of copilot, github/microsoft/openai claim that using code to train an LLM is fair use, which is a novel legal theory currently being tested in court with the class action lawsuit. it is unrelated to the github TOS and the outcome of that case (along with the related hachette vs IA) determines whether any type of scraping can be considered fair use, which would essentially mean (imho) that anything posted publicly on the internet goes into the public domain (this is very abbreviated but i don't think exaggerated)
Replies (2)
-
@jedbrown@hachyderm.io 2025-03-11 15:41
@hipsterelectron@circumstances.run @micr0@fuzzies.wtf @esm@wetdry.world Ah, but you see, it's only public domain after being laundered through "AI", which as a political/power consolidation project, means you have to pay a company for an approximate copy of the original. Making everything public domain would have some redeeming qualities not shared by the AI-laundering model.
-
@veronica@mastodon.online 2025-03-12 21:44
@hipsterelectron@circumstances.run @micr0@fuzzies.wtf @esm@wetdry.world So, by extension, training a model on another model is also fair use. Especially if it undermines the value of the other model. That's cosmic justice even.