@protowlf@mastodon.gamedev.place
Post #2702976
2026-04-03 08:35 UTC
@agersant@mastodon.gamedev.place @Forbearance@mastodon.xyz
This is a simplification, but IMO it hits the fundamental point: when a code LLM "invents" something new, it only appears that way because the final words didn't happen to line up with the relations between words it did necessarily plagiarize to produce its output.
But that begs another question: if an LLM is "only" scraping my code to change weights in a model, is that still stealing? I mean, it didn't take my code blocks, it just took the relations between words.
Replies (1)
-
@protowlf@mastodon.gamedev.place 2026-04-03 08:46
@agersant@mastodon.gamedev.place @Forbearance@mastodon.xyz And my response to that is best illustrated with an analogy to image file types: If you download my jpg and save it as a png, you now have a 100% completely different file from mine. It's all new data! But obviously you still took my image. The meaning of the data is what matters. LLMs do not have a consciousness, and their output has to be a function of input(s). When my code was scraped, it's meaning was stored in a roundabout format, but ultimately it was stored.