Re: Capy: Request for endorsements

Peter Dimov via Boost <[email protected]>
Newsgroups gmane.comp.lib.boost.devel
Message-ID <[email protected]>
Rainer Deyke wrote:
> On 2/20/26 13:06, Peter Dimov via Boost wrote:
> > Rainer Deyke wrote:
> >> Clearly the person who gives the prompt has no copyright claim.  On
> >> the other hand, I can trivially write an "AI" program that
> >> regurgitates its exact training data, even if its internal model looks nothing
> like the training data.  If I "train"
> >> this "AI" on the Star Wars movie, does this means I can use it to
> >> create a copyright-free copy of Star Wars?
> >
> > No. But if you train it on one million Star Wars-like films, and then
> > generate a film that is like them, but isn't a copy of any of them,
> > then that generated film is - this is the argument - not a derived work.
> 
> In other words: if you kill a man, you're a murderer.  If you kill a million, you're
> a conqueror.

Well, no.

Taking one millionth of one million things each is not killing one million.

> > That's because compression works by identifying common patterns, and
> > the common patterns aren't copyrightable. (Because if they were, all
> > these one million films would all infringe on one another's
> > copyright.)
> 
> No, the process is fundamentally different, and it's the process that matters in
> copyright, not the result.
> 
> If I, by pure coincidence, write a book that's word-for-word identical to a book
> that already exists but I haven't read, then I haven't violated copyright because
> I haven't copied or transformed the original work.

Maybe in theory, but you will almost certainly lose that case. You'll lose it
more often than not even if you take extreme precautions and are able
to prove you didn't copy the original.

> If I take an existing picture and apply filters to it until it is unrecognizable, then
> I have legally created a derivative work that is still under the original copyright.

Right.

It's precisely this reductive view - if something is an input to a process, then
the end result is a derived work by definition - that is being challenged by LLMs.

It can't be correct, because if it were, you wouldn't be able to produce a non-
infringing work either, if you have ever been exposed to a copyrighted one;
the exposure has altered your mental state, so your output is tainted by
definition.


_______________________________________________
Boost mailing list -- [email protected]
To unsubscribe send an email to [email protected]
https://lists.boost.org/mailman3/lists/boost.lists.boost.org/
Archived at: https://lists.boost.org/archives/list/[email protected]/message/3VRG7MW5V3EATEC7KC6OLTT2UWSQ6QQJ/
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.