this post was submitted on 25 Aug 2023
88 points (100.0% liked)
Technology
37739 readers
500 users here now
A nice place to discuss rumors, happenings, innovations, and challenges in the technology sphere. We also welcome discussions on the intersections of technology and society. If it’s technological news or discussion of technology, it probably belongs here.
Remember the overriding ethos on Beehaw: Be(e) Nice. Each user you encounter here is a person, and should be treated with kindness (even if they’re wrong, or use a Linux distro you don’t like). Personal attacks will not be tolerated.
Subcommunities on Beehaw:
This community's icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
A slightly compressed JPG of an oil painting is still, at least for purposes of intellectual property rights, not distinct from the original work on canvas. Sufficiently complex and advanced statistics on a work are not substantially different from the work itself. It's just a different way of storing a meaningful representation.
These LLMs are all more or less black boxes. We really cannot conclusively say one way or another whether they are storing and using the full original work in some form or another. We do know that they can be coaxed into spitting out the original work, though, which sure implies it is in there.
And if the work of a human that needs to be fed is being used by one of these bots -- which is pretty much by definition a commercial purpose given that all the relevant bots are operated as such -- then that human should be getting paid.
Only very rarely, under extreme cases of overfitting. Overfitting is a failure state that LLM trainers want to avoid anyway, for reasons unrelated to copyright.
There simply isn't enough space in a LLM's neural network to be storing actual copies of the training data. It's impossible, from a data compression perspective, to fit it in there.