This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
It was bad that they were sending them to the dump, too. It doesn't become less bad just because most people wouldn't think of it until now.
We have the storage. Billions of pulp novels must be uploaded. They were worth printing once, they are certainly worth archiving.
They were worth printing thousands of times each. Would your archive contain one copy of each distinct work, or every single book that came off the presses?
More options
Context Copy link
Storage here would mean a massive, climate controlled, disaster resilient library-warehouse, requiring a team of specialists. You can't just pile the books up in a bunch of cardboard boxes; every year you'd have X% decay into oblivion. It would also be closed off to the public.
There's a reason no one does this: no one cares, given the costs involved. I'm sure Anthropic would have been happy to just ship them to someone else for preservation, but who would take them?
I don't mean store the dead trees. I mean store the data, in free public access.
That's forbidden by copyright law.
The advent of generative AI in 2022 has resulted in a lot of talk about copyright, and if Anthropic decided to unilaterally declare copyright as a concept that's become obsoleted by gen AI and which belongs in the dustbin of history like slavery and lobotomy, that would certainly make the ongoing AI-related culture wars that much more interesting. Now I kinda want to see that happen.
Information should be free! except for distilling.
I largely expect that US courts will eventually find model weights to be uncopyrightable: they're the product of an expensive "sweat of the brow" operation like a phone book, and purely functional in nature. The information in phone books isn't covered, nor are purely functional inventions. The process just doesn't strike me as creative at all.
Although I wouldn't bet huge money on it, and there is a nonzero chance Congress decides to modify the law in this instance.
You're entirely correct here, but I was surprised to double-check and discover that the Supreme Court ruling to that effect is only 35 years old. By that point authors and publishers had been using fictitious entities for decades to try to catch copying, all for naught.
Even if you can't make a case for foundation models ("what's the probability of this next token") being creative products, once you get to the RL steps that turn a foundation model into a chatbot or an agent, endowing it with a particular range of capabilities and "personality", you're being as creative as any author of copyrightable scientific publications and other non-fiction. The difference between "I want to honestly find the best answers I can to this exam" and "I want to get away with felonies long enough to find the answer key to this exam" is not a precisely defined function of the initial training corpus.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Oh, misread. Yes, they should be stored and made available to the public. But I'm sure they are storing them internally as part of their training corpus. As for the public, many or most of them are under copyright still; I have plenty of complaints about that, but I can understand why Anthropic doesn't want to exacerbate its existing copyright issues.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link