Trump had asked to deliver part of the closing arguments himself, but Engoron rejected the request Wednesday in a contentious email exchange with Trump's lawyers after the former president refused to commit to only speaking about the facts of the case and not engage in any attacks.
The judge didn't really reject their request, they simply ghosted him when he told them the terms under which he would accept their request, and after 2 or 3 deadline extensions their opportunity lapsed.
Only if they publish or sell it. Which is why OpenAI isnt/shouldn’t be liable in this case.
If you write out the entire Harry Potter series from memory, you are not breaking any laws just by doing so. Same as if you use photoshop to reproduce a copyright work.
Actually you are infringing copyright. It's just that a) catching you is very unlikely, and b) there are no damages to make it worthwhile.
You don't have to be selling things to infringe copyright. Selling makes it worse, and makes it easier to show damages (loss of income), but it isn't a requirement. Copyright is absolute, if I write something and you copy it you are infringing on my absolute right to dictate how my work is copied.
In any case, OpenAI publishes its answers to whoever is using ChatGPT. If someone asks it something and it spits out someone else's work, that's copyright infringement.
There is also some more direct president for this. There is a website called “library of babel” that has used some clever maths to publish every combination of characters up to 3260 characters long. Which contains, by definition, anything below that limit that is copywritten, and in theory you could piece together the entire Harry Potter series from that website 3k characters at a time. And that is safe under copywrite law.
It isn't safe, it's just not been legally tested. Just because no one has sued for copyright infringement doesn't mean no infringement has occurred.
The figures speak for themselves: in Europe, one in eight vehicles purchased in 2022 was electric (in the U.S., one in seven), while in China, EVs accounted for 25% of total car sales during the same period.
Why the sudden jump from fractions to percentages, right at the point a clear comparison was supposed to be made?
This article is also suspiciously heavily pro-Chinese. It amplifies the significance of government subsidies to buyers of EV's, while ignoring the fact that most other governments have provided similar subsidies. It reads as if it's trying to encourage investment in China, while stopping short of discussing the kind of leverage China would hold over foreign investors (or even local investors).
It also skirts around the fact that the main reason China is so prominent in the EV market, both locally and internationally, is because Tesla's biggest factory is located in China. That information is there in the article, but repeatedly downplayed.
No they did lose it, I believe. As part of the trial a judge ruled that scanning physical copies and lending out one digital copy per physical copy scanned was illegal. They were operating in a legal grey area, then as soon as they came out of that grey area they lost it. That's why I think they should have settled out of court.
They were sued for lending unlimited copies, fought it, then ended up being told they couldn't lend any copies without a license.
Wait, wasn’t the case about the archive giving people unlimited borrowing during COVID?
Yes. That I don't have an issue with, although I think it was a mistake in hindsight.
The issue was trying to face the publishers head on in court, and then coming at them with a frivilous legal argument that had no hope in succeeding. They've done the same with their appeal - and donors have paid for both. They should have done absolutely everything they could to settle out of court.
To me using the internet archive’s interface is clunkier than archive.today’s. Maybe it’s the thumbnails, maybe it’s the loading times.
I agree, but just because archive.today is more polished doesn't mean it's more trustworthy or respectable.
You are. The person who made or sold a gun isn't liable for the murder of the person that got shot.
The difference is that ChatGPT is not Photoshop. Photoshop is a tool that a person controls absolutely. ChatGPT is "artificial intelligence", it does its own "thinking", it interprets the instructions a user gives it.
Copyright infringement is decided on based on the similarity of the work. That is the established method. That method would be applied here.
OpenAI infringe copyright twice. First, on their training dataset, which they claim is "research" - it is in fact development of a commercial product. Second, their commercial product infringes copyright by producing near-identical work. Even though its dataset doesn't include the full work of Harry Potter, it still manages to write Harry Potter. If a human did the same thing, even if they honestly and genuinely thought they were presenting original ideas, they would still be guilty. This is no different.
I use Quad9 because it's the one I want to use, and some other reasons that I decided on at the time but can't remember (something to do with encrypting DNS).
I'm not going to change my posture because one dodgy website decided to behave in a dodgy manner. archive.today is not some charity saint acting in the public interest - that's confusing it with archive.org/Internet Archive/The Way Back Machine. archive.today is passing off as a more respectable organisation.
Mind you, after Internet Archive fucked up with their lawsuit from book publishers, and subsequently had a judge explicitly rule that their "one copy, one share" philosophy for digital lending was illegal, they're not great in my books either. They basically made the situation worse for everyone, meanwhile they used it as an opportunity to drum up more donations, all the while risking being shut down and everyone losing The Way Back Machine. Although, really I put most of the blame on their lawyer, who of course was the real winner in all that.
It wasn't his intent when he posted about wanting to buy the company, but after the SEC forced him to buy, and once he formed a coalition to buy it, the plan became to kill the business off with a leveraged buyout.
Make no mistake, Twitter isn't dying because of Musk's mismanagement, it's dying because it was saddled with $13bn of debt that it could never have hoped to pay back. The mismanagement is just a show to provide deniability.
That isn't to say that Musk is some kind of genius, just that he's a clown playing his role.
Whether or not they "instructed the model to regurgitate" articles, the fact is it did so, which is still copyright infringement either way.
Cohn was also gay, and when he was dying of AIDS Trump dropped him like a hot potato.
From your same wiki link:
and according to Seymour's notes, Trump was the last person to speak to Cohn on the phone before he died in 1986.
I haven't found anything that corroborates your statement that Trump "dropped him".
I did find this though, which reminds me of South Park; it sounds exactly like Cohn was a customer of CumHammer Brand Management:
the AIDS Memorial Quilt describes him as "Roy Cohn. Bully. Coward. Victim."
Edit: Wait, actually, looking at one of the sources from the Wiki article, there was this bit:
Bell also doubts that Cohn’s last conversation was with Trump, who, she said, abandoned his lawyer when he found out that Cohn was H.I.V.-positive. “They were so close, they talked at least several times a week,” she said. “And as soon as he found out, he took all his cases away from Roy except for one and got new lawyers. After all they’d been through together.”
Seymour is Christine Seymour, Cohn's telephone operator, who listened in on many of his conversations. Bell is Susan Bell, Cohn's secretary.
That's interesting, if only because the judgement flies in the face of the actual legislation. I guess some judges don't really understand it much better than your average layman (there was always a huge amount of confusion over what "transformative" meant in terms of copyright infringement, for a similar example).
I can only rationalise that your test version could be considered as "research", thus giving you some fair use exemption. The placeholder graphics were only used as an internal placeholder, and thus there was never any intent to infringe on copyright.
ChatGPT is inherently different, as you can specifically instruct it to infringe on copright. "Write a story like Harry Potter" or "write an article in the style of the New York Times" is basically giving that instruction, and if what it outputs is significantly similar (or indeed identical) then it is quite reasonable to assume copyright has been infringed.
A key difference here is that, while it is "in private" between the user and ChatGPT, those are still two different parties. When you wrote your temporary code, that was just internal between workers of your employer - the material is only shared to one party, your employer, which encompases multiple people (who are each employed or contracted by a single entity). ChatGPT works with two parties, OpenAI and the user, thus everything ChatGPT produces is published - even if it is only published to an individual user, that user is still a separate party to the copyright infringer.
If a person robs a bank, but is not caught, are they not still a bank robber?
While calling someone who hasn't been convicted of a crime a criminal might open you up to liability, and as such in practice a professional journalist will avoid such concrete labels as a matter of professional integrity, that does not mean such a statement is false. Indeed, it is entirely possible for me to call someone a bank robber and prove that this was a valid statement in a defamation lawsuit, even if they were exonerated in criminal court. Crimes have to be proven beyond reasonable doubt, ie greater than 99% certain, while civil court works on the balance of probabilities, ie which argument is more than 50% true.
I can say that it is more than 50% likely that copyright infringement has occurred even if no criminal copyright infringement is proven.
That isn't pulled from my ass, that's just the nuance of how law works. And that's before we delve into the topic of which judge you had, what legal training they undertook and how much vodka was in the "glass of water" on their bench, or even which way the wind blew that day.
According to the Federal legislation, it does not matter whether or not the copying was for commercial or non-commercial purposes, the only thing that matters is the copying itself. Your judge got it wrong, and you were very lucky in that regard - in particular that your case was not appealed further to a higher, more competent court.
Commerciality should only be factored in to a circumstance of fair use, per the legislation, which a lower court judge cannot overrule. If your case were used as case law in another trial, there's a good chance it would be disregarded.