HN user
Do you not think it a bit too hyperbolic to throw scare quotes around experts and imply the only people who can have opinions on systemic risk are software engineers? I don't think it is unreasonable for people who haven't run or worked for a hosting service to have opinions on the policy aspect or economic impact of hyperscalers.
Assuming you're referring to Thaler v. Perlmutter, Thaler claimed to the copyright office that the image at issue was "autonomously created by a computer algorithm running on a machine". So the question of "if you claim the LLM did it itself" is settled (shocker, cf. Naruto v. Slater, 888 F.3d 418), but that definitely did not settle "_I_ used the LLM to do it".
I think the suggestion is that the government use of that public data could be such as to create a chilling effect. That is, the upload and interaction of the user with the private company is almost irrelevant: it is just part of the antecedent to the government's conduct.
If you believe the government would only use that data for just purposes then you probably wouldn't then believe that there is a 1A issue. But if you think the government would use it to identify persons at a protest and then take adverse actions against them on the basis of their presence alone (which to be clear, seems distinguished from the immediate instance) you would probably think there is a 1A issue.
AIUI Because if you have a zero data retention agreement you are necessarily not in the class of records at issue (since enterprise customers records are not affected, again AIUI per platinffs' original motion which might be because they don't think they're relevant for market harm or something).
So I think that this is more so an artefact of the parameters than an outcome of some mechanism of law.
I don't think it is accurate to say that the data becomes the government's or they have to act as an informant (I think that implies a bit more of an active requirement than responding to a subpoena), but I agree with the gist.
Do you think the 4th amendment enjoins courts from requiring the preservation of records as part of discovery? The court is just requiring OpenAI to maintain records it already maintains and segregate them. Even if one thinks that _is_ a government seizure, which it isn't---See Burdeau v. McDowell, 256 U.S. 465 (1921); cf. Walter v. United States, 447 U.S. 649, 656 (1980) (discussing the "state agency" requirement)---no search or seizure has even occurred. There's no reasonable expectation of privacy in the records you're sending to OpenAI (you know OpenAI has them!!; See, e.g., Smith v. Maryland, 442 U.S. 735 (1979)) and you don't have any possessory interest in the records. See, e.g., United States v. Jacobsen, 466 U.S. 109 (1984).
Well, presumably the claim would be that a factor in their not having taxable income was the fact that they didn't have to amortize their development cost.
Meanwhile, software developers spot code fragments seemingly lifted from public repositories on Github and lose their shit. What about the licensing? If you’re a lawyer, I defer. But if you’re a software developer playing this card? Cut me a little slack as I ask you to shove this concern up your ass. No profession has demonstrated more contempt for intellectual property.
Seriously? Is this argument in all earnestly "No profession has been more contemptuous therefore we should keep on keeping on"? Should we as an industry not bother to try and improve our ethics? Why don't we all just make munitions for a living and wash our hands of guilt because "the industry was always like this".
Seems a bit ironic against the backdrop of <https://news.ycombinator.com/user?id=tptacek>:
All comments Copyright © 2010, 2011, 2012, 2013, 2015, 2018, 2023, 2031 Thomas H. Ptacek, All Rights Reserved.
(although perhaps this is tongue-in-cheek given the last year)
There already was a verdict (for the Times) but it was thrown out on appeal. This is a re-do.
ETA: I looked up the docket [0] and in fact, this was the second appeal (See ECF 64). There was also an appeal on a prior MTD, hence the extreme delay.
[0] https://www.courtlistener.com/docket/6081165/palin-v-the-new...
Probably https://free.law/
ETA: which is of course mentioned on the thread root. But RECAP users would be paying, in that case.
I would think it makes some amount of sense if you think they're vegetarian for some moral reason and you think you could court them to become customers.
Why do you think that? This obviously does not preclude copyright interests from existing in works which were generated using "AI" as a blanket rule; rather, this is about the fact that the applicant persistent in insisting that the author of the work was an "autonomous[] computer algorithm".
Do you think autonomous computer algorithms (to the extent we could suppose they exist, for the sake of argument) should have a statutory right to copyright?
What comparison do you think the parent comment is making? They just vaguely gesture at corporate personhood and say "what about that?"
My best reading of it is "can 'they' say that corporations can author works?" and excepting works for hire, no, "they" can't.
Yes, I agree. I don't think I am saying anything inconsistent with that.
Code is not purely functional. If it is, it is not copyrightable (at least in the US; probably true elsewhere but I am less sure) [0]. I would not expect most prompts to be purely functional.
What I can guarantee, is that the prompt itself would be copyright-able.
That's non-obvious to me. Even if the prompt is extremely long and precise, if it is somehow purely functional, it seems possible for it to not be (although in practice, I agree that most prompts could be).
In what way do you think corporate personhood is relevant here?
I don't think it takes that much imagination here. Not sure what good the first step is actually doing you. Might as well just AI-generate your racketeering demand letters without doing that part.
As a matter of law? Sure it does. Thaler said the image at issue was "autonomously created by a computer algorithm running on a machine". He's been trying to walk that back for the last couple of years though. See Thaler v. Perlmutter, 1:22-cv01564-BAH (ECF #24), D.D.C. (Aug. 18, 2023).
Copyright only protects the actual text. LLMs have weights, not exact copies.
Following this logic a lossily compressed image is completely unprotected by copyright.
In any case, saying "if I put in some input and get copyrighted output" is tantamount to copyright violations; if I use a generative tool and generate copyrighted info is it the tools fault?
Do you not think this is obviously fact-specific? If I gzip a bunch of (copyrighted) files, then obviously that doesn't somehow make distributing them not infringement. If I now replace the tool = ungzip + input = files combination with tool = (ungzip and files) and input = (selection mechanism over files) do you think that in the second case distributing the tool is not infringement? I don't mean to say that any of these is precisely the same as the LLM case, but I think your argument is clearly overbroad.
OpenAI at most broke an EULA or some technicality on copyright w.r.t. local ephemeral copies. What's the damage to the NYT though?
One obvious damage claim (if you are skeptical of market harm wrt newspaper/oneline sub sales) is that they were entitled to the FMV of licensing costs of the articles, which is not so hard to value: OpenAI has entered such agreements with AP and others. [0]
[0] https://apnews.com/article/openai-chatgpt-associated-press-a...
I think this is discussed at length in part II.D (starts at the bottom of 17). I would write more but I have spent too long already on this thread :)
I would be a bit careful about trying to liken motivation for something like an EO to a law though; many members of congress voted to pass the exact language in the final bill, and they might not all have agreed with _why_. So I would put to you that the text itself is the primary thing one should consider, especially more in the legislative case than the executive one.
Right, I'm saying they based it on on the "text" of the law, instead of the motivation.
Sure, although they do discuss TikTok's challenge to the motivation ("Petitioners further argue that the Act is underinclusive as to the Government’s data protection concern, raising doubts as to whether the Government is actually pursuing that interest"). I just don't think the quote you had stands for what you were saying.
At what point in the ruling did they wonder what motivated the effective ban?
Above is at page 15. Also, I think you're probably looking for the paragraph starting with "For the reasons we have explained, requiring divestiture for the purpose of preventing a foreign adversary from accessing the sensitive data of 170 million U.S. TikTok users is not 'a subtle means of exercising a content preference.' Turner I, 512 U. S., at 645." (at 12).
I saw elsewhere you likened this to the Trump muslim ban. I don't think that comparison is apt. The First Amendment issues there were not decided by the 9th circuit in the first one (“we reserve consideration of [First Amendment religious discrimination] claims until the merits of this appeal have been fully briefed.” State v. Trump, 847 F.3d 1151, 1168 (9th Cir. 2017)) the stay there was issued due to likelihood of success on the merits wrt due process issues; I don't know offhand about the second one; and the third attempt was upheld.
It absolutely does. (It’s in the opinion.)
The opinion actually assumes without deciding that First Amendment scrutiny applies, so I don't think it "absolutely" does. (But yes, it probably does and Sotomayor and Gorsuch would decide as much)
The quote you posted is about if the exclusion of platforms "whose primary purpose is to allow users to post product reviews, business reviews, or travel information and reviews" means the law is content-based, but the Court is saying that provision is irrelevant because TikTok brought an "as-applied" challenge (and not a facial one) [0] and that provision doesn't change how it applies to them. So they are looking at the parts of the law (and the congressional record supporting them) which actually cause TikTok to be subject to the qualified divestiture.
In America each party typically bears their own costs.
I believe each transaction has to be over $100 and there's a total limit on the number of pay over time plans you can have active. But yes, other than the dollar amount, I've not seen anything not be eligible.
If it's an SJ lookalike or soundalike (and you don't claim otherwise), there's no problem.
This isn't true. At least with respect to "soundalike" see, e.g., Waits v. Frito-Lay 978 F.2d 1093 and Midler v. Ford Motor Co. 849 F.2d 460.
It seems like a fair take to me. Certainly plenty of companies, HashiCorp included, would have fewer choices about how to license future code if they made different decisions regarding contributions. But something like:
I wonder if the community has grounds to sue any of these companies who are ditching the AGPL in favor of proprietary, source-available licenses, especially under “third-party beneficiary contract” legal theories, like @conservancy did in their suit against Vizio.
If the source code originally used AGPL, isn’t it still contractually obligated to ensure those rights to its users, including any new source code added to it?
seems to indicate rather strongly that the original poster doesn't understand the AGPL terms, CLAs, or the nuances of contract litigation. At least OP recognises that at least enough to ask a question about it, but I don't think "random questions I thought up about reasonably well-settled law in Twitter thread form" makes a good post.
I don't believe so (although this is probably fact-specific about what "accidentally" means), at least with regards to retracting the licence. The GPL is an irrevocable licence.
They could choose not to continue to distribute the code though (but good luck clawing back the GPL licenced copies).