ang_cire

35 ms·

691.
by ang_cire 3y ago
> Where's the evidence that OpenAI pirated all the content they trained models on and didn't access it normally just like everyone else? It's quite the allegation that OpenAI torrented `all-media-archive.xz` to train their
692.
by ang_cire 3y ago
> and you buy a copy This is the step that OpenAI and other companies are missing. They didn't pay for the content they are using. > For humanity to progress ideas must flow freely. AI models are not and will never progress human
693.
by ang_cire 3y ago
This is honestly the best response so far, and I agree that my statement was probably too ambiguous. I am not claiming that anyone says AI should have all the rights of humans, I'm saying that people are arguing that: -> the Right e
694.
by ang_cire 3y ago
> the process (of human vs AI learning) just seems to be close enough It only seems that way because the people who understand it best also stand to profit from it, and are choosing to misrepresent it so that people who don't unders
695.
by ang_cire 3y ago
This is a lot of text just to say, "I care more about ML systems getting better than I do about whether they do so ethically". > people who illegally (ftfy) use copyrighted works to create a product Yes, this is something that
696.
by ang_cire 3y ago
> A human has a right to use a shoe. Or an AI. What right is that, exactly? I don't remember the "Right to use a tool". You are trying to make "using a tool in a way that harms people" have greater legal or ethic
697.
by ang_cire 3y ago
I'm a Bay Area startup alumni, so closer to the actual VC scene than most on here, and all the actual successful founders I know have a pretty strong belief in the good that their products will bring, even if they usually overstate the
698.
by ang_cire 3y ago
100000% agree. I get that this is a forum for "future millionaires" (wink wink), but you'd think that us IT folks should be both the most capable of seeing through the b.s. hype that companies put out trying to mystify softwa
699.
by ang_cire 3y ago
Something that is very disturbing in seeing people compare AI models to humans is the seeming implication by many of them that humans only have rights because a law says they do, and that there are no rights inherent to being a human (which
700.
by ang_cire 3y ago
First, we should not give AI models the same rights as humans, because they're not humans. We should place far MORE restrictions on AI models. Second, we should force OpenAI to cut deals with each content creator whose content they wan
701.
by ang_cire 3y ago
First off, you may be unaware but the subject of browser caches is well-tread in copyright law, and has been ruled not to be the same as other methods of downloading, so it's not applicable here. Googlebot allows creators to restrict w
702.
by ang_cire 3y ago
> if they checked the book out from the library and let the LLM read it Did you know it's actually illegal to photocopy books you check out from the library? It's an illegal copy. That aside, the analogies between humans and ML
703.
by ang_cire 3y ago
You're putting much more emphasis on the 1A link to Fair Use than it actually has. 1A is not the entirety of the basis for Fair Use, it is much more expansive than the 1A justification alone, which relates primarily to transformation f
704.
by ang_cire 3y ago
They were scanning books they had legally-obtained copies of. OpenAI is very often downloading whole books which they do not have licenses for. Publicly-accessible does not equal legal to download, in US copyright law.
705.
by ang_cire 3y ago
Right, which OpenAI is trying to argue should not apply to their illegal downloading of copyrighted material for training use.
706.
by ang_cire 3y ago
If Sinophobia is a valid justification to ignore laws, you can do just about anything.
707.
by ang_cire 3y ago
You are missing that the AI is the one creating the output. If I sell you a hammer and nails, I'm not liable if you create a dangerous building. If you ask me to build you a dangerous building and I do it, I am liable if people get hur
708.
by ang_cire 3y ago
And what does the training data then do? Just sit in a box, or produce content for "direct human consumption" that would have market impacts? If all anyone has to do to invalidate copyright law is introduce an intermediary step wh
709.
by ang_cire 3y ago
The big difference is that Google legally obtained the copies of the books they scanned. OpenAI did not. Publicly-accessible != legal to download/ copy. The copyright infringement in Google's case was not illegally obtaining the o
710.
by ang_cire 3y ago
Yeah, it seems like so many people in here are shilling for AI companies over actual creators, it is absolutely depressing. No one is arguing that OpenAI should not be allowed to train on copyrighted works, they are arguing that OpenAI shou
711.
by ang_cire 3y ago
Humans have legal rights like 1A that form the basis for many carve-outs like Fair Use. GPT-4 does not.
712.
by ang_cire 3y ago
Sure, in terms of removing Fair Use entirely, but nothing about Fair Use is absolute, and it can absolutely be substantially changed and reinterpreted. Especially since this is about companies using Fair Use for financial gain, in ways that
713.
by ang_cire 3y ago
> I cannot abide the argument that obviating work is a harm. That's a misrepresentation of the argument. The actual argument is that taking someone's work sans-permission and building on it in order to obviate the original wor
714.
by ang_cire 3y ago
But it's still illegal for me to download a movie without a license, even if I just intend to make AMVs with/ "transform" it. OpenAI's entire argument is that they think this shouldn't apply to them, because th
715.
by ang_cire 3y ago
If you read the PDF, you'd know that OpenAI admits they ARE actually copying his books, in order to train on them, and they are arguing that this is okay because they don't want to have to obtain a legal copy of the work, because
716.
by ang_cire 3y ago
See, for me this section also jumped out, but for the opposite reason. They are intentionally being deceptive about the aim of AI training. The aim is not to create a model, for the sake of having a model , the aim is to create a model tha
717.
by ang_cire 3y ago
Yes, but usually in those situations you can actually move on to something else. One a plane, you're forcibly stuck there. No, "Hey, it was nice talking to you, I'll see you around!" option.
718.
by ang_cire 3y ago
Laptop stickers are a perfect means to signal who you want to talk to, and who you don't, without saying anything.
719.
by ang_cire 3y ago
That's just not true. They're not "reasoning" anything, they're assembling code that appears similar to something they were trained on, within whatever additional language training constraints they're given. I&
720.
by ang_cire 3y ago
This is the driving equivalent of "just follow instructions"; you're ignoring that they are intentionally making the speed limits difficult to respect in order so that they can ticket you. Telling people not to get tricked is

More ›