@Forbearance@mastodon.xyz

Forbearance @Forbearance@mastodon.xyz

· 5mo ago

Replying to

@timmy delivers bitcoin to *who*? for *what*? 😕😕😕

7

0

2

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@ana_valdi All I know about libel in the UK is that it's routinely used to suppress information inconvenient to the powerful. But if you're writing and publishing a book, you have a publisher and an editor, and knowing about libel in the UK is their job, right? Unless the publisher is also you.

2

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@agersant@mastodon.gamedev.place @protowlf@mastodon.gamedev.place And of course the voracious AI trainers scoop up a lot of code and other things that they *wouldn't* be allowed to just redistribute as-is, things like freeware or non-free software, and those all get thrown in their pot too. If you can't publish a .zip of it that someone can extract with WinZip, you probably can't publish a giant matrix that someone can extract it from by asking nicely.

2

6

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@agersant@mastodon.gamedev.place @protowlf@mastodon.gamedev.place Traditionally, doing things like making tables of word trigrams and other stuff broadly under the header of "statistical analysis" has been unregulated by copyright law, and has not required any sort of license, on the theory that it's "just" math and not really anything like copying. But this way of thinking about it much predates the current popularity of math that steals your ideas, and might not be adequate to the present moment.

2

5

0

Forbearance @Forbearance@mastodon.xyz

· 5mo ago

Replying to

@theresnotime 🔧

1

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@protowlf@mastodon.gamedev.place I can see this making sense by a sort of "conservation of thought" argument: if those big matrices are definitionally devoid of the spark of original thought, then anything that comes out is definitionally derivative, copied from somewhere even if the sources are so scattered and numerous that they are permanently unidentifiable.

1

0

Forbearance @Forbearance@mastodon.xyz

· 9mo ago

Replying to

@chewie @kkarhan @vikxin @NanoRaptor Windows Central isn't claiming to have *reviewed* it, I think they're just garden variety fooled.

1

0

Forbearance @Forbearance@mastodon.xyz

· 5mo ago

Replying to

On the other hand, Past Software is broken because it is shitty and Current Software is broken because it is evil so

0

3

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@protowlf@mastodon.gamedev.place I looked and found https://lawreview.uchicago.edu/online-archive/plagiarism-copyright-and-ai but that is set in the academic and legal sphere, where a whole idea or concept being spit out without the citation to its originator constitutes "plagiarism". I assume I'm not guilty of plagiarism in the computing field every time I apply a facade or visitor pattern without citing the Gang of Four.

lawreview.uchicago.edu

Plagiarism, Copyright, and AI | The University of Chicago Law Review

Critics of generative AI often describe it as a “plagiarism machine.” They may be right, though not in the sense they mean. With rare exceptions, generative AI doesn’t just copy someone else’s creative expression, producing outputs that infringe copyright. But it does get its ideas from somewhere. And it’s quite bad at identifying the source of those ideas. That means that students (and professors, and lawyers, and journalists) who use AI to produce their work generally aren’t engaged in copyrig

0

12

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@agersant@mastodon.gamedev.place @protowlf@mastodon.gamedev.place When you bring in the GPL, you get the question of whether all those model weights are "software", or whether they in some sense form a single program with all the actual code around them.

0

7

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@agersant@mastodon.gamedev.place @protowlf@mastodon.gamedev.place The MIT license grants the right to "deal in the Software without restriction", though, provided that "[t]he above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software." Is that happening when you throw a lot of MIT-licensed code and their license documents in a blender and come out with a bunch of matrices? If the code is encoded in the matrices, the license probably is too.

0

8

0

Forbearance @Forbearance@mastodon.xyz

· 6mo ago

Replying to

@protowlf@mastodon.gamedev.place Do you have a good source (ironically) that sets out the reasoning behind why we should understand what's going on inside one of these models while code is coming out as "plagiarism"? They absolutely will spit out whole chunks of code that one can point to and say "that was copied from such-and-such without following the license". But they also don't *always* do that. Are they *always* plagiarizing, or only sometimes?

0

14

0