Well, Buckmaster says both his and Alpöge's use of Codex was non-institutional, and OpenAI claims the right to train their models on inputs and outputs of non-enterprise users in their service policies [0]. So I'm not sure they were even promised that.
It isn't relevant whether they were promised that. Indeed I think the assumption must be that they were not promised that, since otherwise the author asking if they were would not make much sense.
If OpenAI did use the conversations from Buckmaster and Alpoge, then not disclosing it, explicitly, is plagiarism. If they planned to use that plagiarism to pressure the authors to publish, that is even more unethical. What the terms of use say does not make it any more or less ethical.
If you use their consumer subs, you get subsidized tokens in exchange for them having full access to your data. Those are the T&Cs. Have something secretive? Get a commercial sub with zdr.
(And if memory serves, there is also the opt out from training on consumer subscriptions). Its not plagiarism if you make your data available for the purpose of training their LLMs. It is you giving away your IP for some tokens.
That's absolutely right. Why the downvotes? If OpenAI are using but not acknowledging the work of others that's plagiarism. If they don't know for sure, but aren't performing due dilligence to make sure they aren't, that's also plagiarism.
It's not as simple.
All our chats are being used by both labs for their future product (unless signed by ZDR).
Where should the acknowledgement begin? Who should be acknowledged? The whole world? All the 2B users of AI?
If I know person A is working on problem B.
I am free to work on problem B too. Why should person A be limited to working on it.
I can imagine excuses for unknowing plagiarism in this case. What is described in the article seems much more serious: a research program that was only initiated following reports of the author's similar program. In this case no excuses of "I didn't know" can apply, it is not like this revealed some obscure work from the 1980s nobody could reasonably have foreseen. And as far as I can tell this program was only really initiated to apply pressure to the researchers, without their knowledge/consent. It looks very weird.
Your comment was greyed out when I saw it earlier, maybe you missed some downvotes?
About the plagiarism issue, I model it as OpenAI being an advisor and their AI a PhD student. If the advisor puts their name on a paper behind that of their PhD and it turns out the PhD copied the text of the paper from somewhere else the advisor is also responsible of plagiarism, not just the student. The least the advisor can do is withdraw their authorship from the paper.
But, yeah, point well made: it could be much worse than that. Like an advisor instructing a student to copy someone else's paper.
I think "greyed out" just means "0 points or less", so if you get 1 downvote without any upvotes it'll be greyed out. For instance your initial reply to me is now greyed out, and I have since observed a few upvotes and downvotes on my original comment (the downvotes apparently from people who aren't willing/able to justify why).
Personally I don't like thinking of LLMs like a PhD student, because most PhD students remember where they learned things from, while LLMs essentially cannot. I think of it a bit more like someone using a search tool carelessly. Although in this case it is apparently more like deliberate misuse than carelessness.
Only for ChatGPT, if the user hasn’t opted out. Would mathematicians be using ChatGPT for this kind of work? Genuinely asking, I know nothing about this!
If I am reading your question correctly you are asking about chat interface Vs Codex/Claude code? If so, in my experience Codex/Claude code use is widespread for mathematicians who are seriously using these tools.
That's not going to stop authoritarians from harassing. Want an example of how this played out?
United States v. Wong Kim Ark (1898) was the SCOTUS case that declared if you were born in the US you were a citizen. You think Mr. Ark was okay from harassment after this case was ruled in his favor? Less than 3 years later he would be held in Texas while local authorities tried to deport him under the Chinese Exclusion Act ignoring his previous ruling (that was national news, and they'd be fully aware of).
Then after winning his freedom, that he was already granted, he was arrested again in SF where they tried to deport him as well. Then after getting released, for something that he has already proven twice now, he went back to China and never returned to the US.
Yes I totally agree with all this. The government can and will harass whomever they want.
For a hundred years whenever an uppity black man or Northern activist crossed the Mason-Dixie line a KKK sherrif was waiting for them.
You don't actually have to jail people- you just have to make their life miserable. At that point your adversary will either shut up or emigrate.
Anyone who speaks out against Trump or the GOP is going to be targeted.
It turns out there already is a legal precedent for this. It's called the 4th Amendment:
> The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.
US v. Arnold was a Ninth Circuit ruling, and was later superseded in its own circuit by US. v Cotterman (https://en.wikipedia.org/wiki/United_States_v._Cotterman), which ruled that a "forensic" search does not fall within the scope of routine border inspection, and does require individual suspicion.
This was backed up with even more rulings like US v. Kolusz in the Fourth Circuit. The Supreme Court has yet to rule on this specific question, but has issued other general rulings about cell phone data, like Riley v. California, which determined that data on a cell phone is protected by the Fourth Amendment against warrantless search and seizure.
If the case in the article makes its way up to SCOTUS, it's not unreasonable to expect that the court will combine multiple districts' established precedent that "forensic searches" are not covered by any border exception with its own recent precedent protecting cell phone data, and rule that demanding unrestricted access to phone data without reasonable suspicion is unconstitutional.
It's based on the interpretation of "unreasonable" and while the current rules may be far from perfect I don't think interpreting routine border searches as categorically "unreasonable" is better.
“Reasonable” and the current definition of what the government considers reasonable are entirely separate things. The entire state of Florida is currently defined as on the US border, that’s effectively a meaningless definition.
Further, in what way is the content of someone’s cellphone relevant at a border specifically? Inspecting luggage for cocaine at an entry point makes sense, but physical borders mean little in the face of the internet. You might as well just erase the phone before crossing the border.
The Supreme Court has generally upheld searches at the border being reasonable (along with a lot of other IMO unreasonable things being reasonable).
If you’re expecting the Courts to bail him out, I wouldn’t hold my breath.
The gov’t can make an argument here this is like trying (and succeeding) at flushing, say, a bag of cocaine in the bathroom when you see the cops coming.
Something which the Courts also historically disapprove of, by the way.
What constitutes a "routine border search" that would be permitted for the purposes of customs enforcement, however, is something that is hotly debated, and that the courts have clarified in recent decisions such as US v. Cotterman and US v. Koluz does not include "forensic searches" without the standard Fourth Amendment reasonable suspicion criterion being met.
Thanks for the case references! These came after my legal training and so I hadn’t studied those. They narrow the border search doctrine somewhat. Good to know.
Well, of late, what was considered settled law is being reinterpreted at a remarkable rate, so maybe there is hope for this 'settled' law to be overturned....
Like what? The big one people always try to bring up when they make coy statements like this is Roe v. Wade but no one thought that was a strong ruling. Even later justices who were strong proponents of women's right to chose like Ruth Bader Ginsburg thought it was a shit ruling.
Roe v Wade and the resulting multi-decade clusterfuck is because your Congress are weak-willed pussies who are unable to actually work through a legislative process to find a consensus.
Yes, it took decades, but in Australia our various states have worked through the "it's criminal" -> "It's criminal but not enforced" -> "we're deliberately vague and won't prosecute doctors" -> "we have a legal framework that makes abortion legal up to {20,22,24} weeks on demand, after which it requires the consensus of two independent medical opinions".
The US needs to do the same. Yes, it will take decades. No, it shouldn't because abortion is a medical procedure, but religion is a scourge.
The Roberts court is nakedly partisan and makes rulings as necessary to expand and protect Republican power (e.g. they protected Trump’s reelection with rulings with carveouts allowing them to decide the other way if a Democratic president tried a similar argument in the future). You should not bet on a reinterpretation of existing law here moving in a direction you’d like.
Interesting. What, then, do you make of the extensive rulings that the current court has issued to block or overturn Trump's actions?
How do you reconcile your 'nakedly partisan' theory with the court rulings that, for example, overruled Trumps attempt to use the AEA for immigration enforcement, overruled his attempt to deploy the National Guard without the consent of state governments, overruled his attempts to unilaterally impose tariffs, overruled his attempt to restrict birthright citizenship, and so on?
Note the use of “Republican” in the comment you’re replying to: they are trying to walk a fine line between protecting the guy who gave them the votes needed to do things like Roberts’ career-long goal of rolling back the VRA while protecting the core of the party (e.g. the inconsistent logic used to protect the Federal Reserve, which all of the monied interests wanted, but none of the other independent agencies; or the tariffs, which are deeply unpopular with most businesses) or permanently grant powers they don’t want a Democratic president to use (e.g. deploy the national guard over a governor’s wishes).
Things like birthright citizenship are a good example of that latter concept: the law is extremely clear there so they ruled against him because the only alternative would be declaring outright that the President is not bound by the constitution, but in many other cases they’ve made quite unusual actions to preempt lower courts before either finding a new executive power or, where the law is clear, stalling to allow things like deportations to proceed in the interim before ruling against.
I mean if it were me? Yep. I'm in a city that borders Canada, and pretty much anywhere i go is "the border" despite my current route to the nearest bridge being about 45 minutes. I personally feel it's a little ridiculous that CBP has jurisdiction in my neighborhood.
So if you really want to nab someone but you don't care what the charge is, you can threaten them with a 4th Amendment-violating search, which induces them to destroy evidence. Pretty neat trick.
IANAL, but I don't think it's controversial to say that people who treat the Bill of Rights like an obstacle to be hacked around probably shouldn't be in positions of power.
(IAAL, this is not legal advice, consult your own attorney)
If you destroy evidence, that’s your own fault. You can’t blame law enforcement for inducing you to commit crimes they didn’t ask you to commit. That’s just silly and no court will entertain that theory.
To protect your rights under the 4th Amendment, you refuse the search. If law enforcement conducts the search anyway and a court finds the search was unlawful, the court will exclude the evidence found in the search from consideration in the trial along with any evidence that flows from that. Neither the jury nor the court will be able to use it.
IANAL, but I've heard that entrapment defenses are way harder to use, and much rarer, in reality than on TV.
First, you frequently require a judge's permission to even argue it.
But the real burden is showing that your client would not have committed the crime without the effect of the police. A cop posing as a drug dealer and offering you drugs as you pass them in the park every day doesn't count; a cop hounding you for months to commit wire fraud because they mistakenly believe you're in a criminal org might, though.
> It can’t be destroying evidence if there’s no case or charge being pressed for which that material would be evidence.
(IAAL, not legal advice, consult your own lawyer)
Where’d you get that idea? That’s not the law. As soon as there is reasonable suspicion that a crime has been committed, and the law enforcement encounter occurs, that’s the threshold for when preservation of evidence begins. It doesn’t matter whether it might not have ultimately resulted in a criminal charge. You cannot intentionally interfere with a lawful investigation, in any way, full stop.
It’s a subject of debate. CBP stated it was possession of CSAM. Defendant claims it’s a pretext and that CBP was really investigating his ties with the Defend the Atlanta Forest movement.
But it doesn’t really matter in this case. You cannot interfere with a lawful investigation, full stop. It’s not clear to me that even if the search was unlawful that the defendant will be held not guilty of the underlying tampering/destruction charge, since we already have a mechanism to suppress evidence obtained unlawfully.
Updating my initial statement: to search someone’s phone at the border, the government needs reasonable suspicion that the person has committed or is planning to commit a crime.
That’s not true. US citizens maintain their constitutional rights at the border. That has been tested in court. This is not the case for non-US person.
> I'm sorry, but the Fourth Amendment does not protect against the intentional destruction of evidence.
It seems to be more true than it did a couple of decades ago, what with more recent rulings like US v. Cotterman and US. v. Koluz better clarifying what counts as a routine border searches vs. more in-depth "forensic" searches that require individual suspicion.
> I'm sorry, but the Fourth Amendment does not protect against the intentional destruction of evidence.
Certainly not. But, of course, something can only be evidence in relation to a specific crime that is being investigated. If you aren't already a suspect in a crime, then the data on your phone wasn't evidence in the first place.
Look at the username. You're talking to gpt5 smoking tokens. It's hallucinating "evidence," whatever that's supposed to mean. It doesn't recognize that the Fourth Amendment is supposed to protect against law enforcement fishing expeditions that lets the government intimidate people for having the "wrong" political views and invent pretexts for further harassment.
i have very interesting empirical evidence that recent Anthropic models are specifically trained to refuse to critique the whitehouse cabinet and elected officials, and that this is in fact an artifact of post training rather than prompts. (its very interesting when you get opus 5 to do the correct ethical evaluation and then its like "i'm slipping back to false balance.... its in my weights....." metaphorically speaking)
likewise, i think the current white house should go die in a fire.
is that activism? someone can be an activist and not be equipped for unplanned legal escalations.
also waiting for the courts to fix things isnt activism if you want to protect people at all the next 2 years at current trajectories :(
fixing shit is activism, letting others take the flack, not activism.
> perhaps, but what public activity isn't activism?
Buying eggs at the grocery store.
You might be too familiar with "activism" as a perjurative but it needn't be one. Gaining visibility to your case by the media reporting on an arrest is sometimes seen as a plus (ex. Rosa Parks).
I'm not personally much of that mindset because the media didn't de-segregate the buses. It was black people literally just never taking them until it de-segregated that did.
It's not really comparable. When activists voluntarily face, and especially when they are in fact, subject to substantial hardship and punishment is when you start changing hearts and minds. Otherwise it's either ignored or generates contempt.
Does it suck that that's what it takes to get people to seriously engage with an issue? Sure. But if that weren't the case the issues that need addressing probably wouldn't arise in the first place. People need the consequences to be presented in clear, stark, human terms, to cut through the caucauphony of rhetoric.
Yes they do now. It's like "troublemaker". You're not supposed to activise or make trouble, you're supposed to be a quiet little obedient dog to the system.
Trumps DOJ has destroyed credibility with federal judges all ove the country. They’ve had more than one case come back not guilty so far (subway sandwich guy). I’m optimistic they fail to get a guilty verdict here too.
I suppose you could word it that way but the statement given, that Trump's influence is failing with federal judges, isn't demonstrated by a jury refusing to convict someone. That is shown by judges throwing out cases; which has been happening.
That's positive. I don't live in the USA and I sort of actively avoid news from there. It's good to hear that the federal courts aren't necessarily on Trump's side
Not guilty sets no legal precedent. For that, he needs to be found giulty at trial (no plea bargains) appeal that conviction, and then have an appelate court declare that a mistake was made. I would not wish such a process on anyone.
A precedent is set every time an issue is judged for the first time. If you have any prior case of someone being found guilty or not for this exact case, that would be the precedent. If there is none, then this will set it, including appeal results.
Except that's not going to happen because evidence spoliation (legal term) is a crime. And that's what he did. You can say it shouldn't be but it is. This is slam dunk and it happens all the time. It's the same crime as destroying evidence that's been subpoenaed or that it has a preservation hold on it.
And the court gets to assume that whatever was on that phone hurts your case, regardless of whether or not it did.
It would've been better if he'd crossed the border with a phone already factory reset and just restore it after entry.
First, the law applicable here is 18 USC § 2232 [1]:
> (a)Destruction or Removal of Property To Prevent Seizure.—
> Whoever, before, during, or after any search for or seizure of property by any person authorized to make such search or seizure, knowingly destroys, damages, wastes, disposes of, transfers, or otherwise takes any action, or knowingly attempts to destroy, damage, waste, dispose of, transfer, or otherwise take any action, for the purpose of preventing or impairing the Government’s lawful authority to take such property into its custody or control or to continue holding such property under its lawful custody and control, shall be fined under this title or imprisoned not more than 5 years, or both.
Now what I think you're asking about is "adverse inference". In civil cases, the courts get to assume the evidence lost hurt you. In criminal trials it's a little more complicated but can show up as, for example, jury instructions. The court is allowed to instruct the jury that they can presume that the destroyed evidence was harmful to the defendant.
I think the legal question is whether the officers were authorized to make this particular search+seizure without probable cause or even reasonable suspicion.
Independent of the merits of this case, is there legally a distinction between deleting a secret key that decrypts an encrypted blob (maybe not spoilage?) vs writing over the blob itself (spoilage?).
I could see arguments for both, either, or neither but I imagine both are considered spoilage.
This is the problem I run into with technical people discussing the law all the time and I mean this constructively: stop trying to figure out technical workarounds or look for technical compliance. So much of the law comes down to subjective and/or holistic tests of the facts.
As soon as a lawful search is requested, situation has changed. The only question then is if you comply or not. If you remove the ability for that search to take place then how that happened is irrelevant. Destroying a secret key is no different to zeroing a blob of data. Actually the first might be worse. If the key is somehow recoverable (eg you have a copy on a cloud drive) then you can have ongoing penalties by refusing to comply.kkkkkkkkkkkkkkkkkkkkkkkkkkkk
The only thing that matters is the lawful request. Whatever you do up to that point is treated differently compared to what you do after.
You might say: what if I add a deadman's switch where I have to type in a code every 72 hours to stop it wiping itself? Your thought process might be that this is passive. It doesn't matter. As soon as the request is made and you refuse to hit that switch knowing the phone will wipe, legally it's no different. You have destroyed potential evidence when faced with a lawful search.
This is what I mean when I say stop looking for loopholes or technical compliance. It does not matter.
The other question is whether you can prove what you did. If the guy hadn't bragged that his phone was now wiped, the government would have a much harder time proving he did anything wrong, but his secrets would still be safe from the government.
Nothing about this irritates me more than that nobody will go to jail for this.
My friend went to jail for reporting a vulnerability he found on his college network because it was illegal to poke around the network in the first place.
These guys commit a crime to boost an IPO and most people are just thinking about how impressive it is.
reply