The OTW and AO3 are currently embroiled in a controversy around legal chair Betsy Rosenblatt's recent-ish interview on training AI using fanworks. Back in February, Rosenblatt offered support for the idea that AI bots should be allowed to scrape and "train" on fanworks on the assumption that this will make them less outdated, i.e., racist, sexist, etc. Here's her exact quote:
(I won't comment on how itchy it makes my eyes to see "fan works" as two words.)
This interview was intended for a legal audience, not an audience of fans, and indeed, it slipped below the radar until the OTW decided to highlight it as part of the OTW Signal on May 6 (link points to Archive.org because the OTW has since removed the link to Rosenblatt's interview and issued this sorta retraction-apologyish thing).
(This Tumblr post has a nice analysis of Rosenblatt's interview and why it was unremarkable as a piece of legal thinking but managed to send much of Fandom into a froth.)
Interestingly, on May 5, the SWG mods picked up on a post on our Discord where someone mentioned making "fanart" with AI, and this sent us all into a collective wide-eyed "ummmm ..." on our mod channel, followed by a very rapid decision that we needed an AI policy sooner rather than later. The kerfuffle around the Rosenblatt interview was kicked off in the next day's OTW Signal, which has made for interesting timing and perhaps put the reaction to the interview more on my radar than it normally would have been, living under a rock as I do. The tl;dr of the whole thing is that fans are pissed that someone "high up" in the ranks of the OTW would cheerlead the nonconsensual harvesting of their fanworks for AI training. But viewing this through the lens of a small archive owner, which I can't not do at this point, then I notice a few things.
( Read more... )
One of the things that excites me—which is probably a bit off to the side of what most people are talking about with AI and copyright—is that AIs are reading fan fiction now. For a long time, machine learning relied almost exclusively on data sources that were known to be in the copyright public domain, such as works published prior to 1927 and public records. The result of that was that machines were often learning archaic ideas—learning to associate certain professions with certain races and genders, for example. Now, machine learning is turning to broader sources from across the internet, including fan works. That means that machines will learn how to describe and express a much more contemporary, broad, inclusive, and diverse set of ideas.
(I won't comment on how itchy it makes my eyes to see "fan works" as two words.)
This interview was intended for a legal audience, not an audience of fans, and indeed, it slipped below the radar until the OTW decided to highlight it as part of the OTW Signal on May 6 (link points to Archive.org because the OTW has since removed the link to Rosenblatt's interview and issued this sorta retraction-apologyish thing).
(This Tumblr post has a nice analysis of Rosenblatt's interview and why it was unremarkable as a piece of legal thinking but managed to send much of Fandom into a froth.)
Interestingly, on May 5, the SWG mods picked up on a post on our Discord where someone mentioned making "fanart" with AI, and this sent us all into a collective wide-eyed "ummmm ..." on our mod channel, followed by a very rapid decision that we needed an AI policy sooner rather than later. The kerfuffle around the Rosenblatt interview was kicked off in the next day's OTW Signal, which has made for interesting timing and perhaps put the reaction to the interview more on my radar than it normally would have been, living under a rock as I do. The tl;dr of the whole thing is that fans are pissed that someone "high up" in the ranks of the OTW would cheerlead the nonconsensual harvesting of their fanworks for AI training. But viewing this through the lens of a small archive owner, which I can't not do at this point, then I notice a few things.
( Read more... )
Tags: