"Did you realize that we live in a reality where SciHub is illegal, and OpenAI is not?"

AFAIK the individual researchers who get their work pirated and put on Sci-Hub don’t seem to particularly mind.

Why would they?

They don't get paid when people pay for articles.

Back before everyone left twitter, the easiest way to get a paywalled study was hit up to be of the authors, they can legally give a copy to anyone, and make no money from paywalls

Also, no researcher would even exist if grad students had to pay for the papers they read and cite. A lot of people is not fortunate enough to have access to these publications through their uni. Heck, even when I had it, I'd still go to sci-hub just for the sake of convenience.

Like a lot of services nowadays, they offer a mediocre service and still charge for it.
That's still the easiest way. Email them, don't tweet them.
It still works. The journal websites always include author contact info, just e-mail them.
legally

Not necessarily. They often do not own the copyright, so then it depends on fair use exceptions. The real owners have gone after authors, which may be the reason they don't make their articles downloadable by default.
- The asking makes it legal if I recall correctly.
  
  They can't host a site with all their articles/papers/research, but if anyone asks for a single copy, they can provide it at their discretion.
  
  And since they don't make any money either way, most provide it and are happy to do so.
  
  You mean asking the publisher?
  
  When you publish an academic paper, the journal/publisher makes you sign a transfer-of-copyright-thing. For example, that meant I could not publish my own papers as a part of my thesis. I had to ask the journals for permission to do that. Depending on how that transfer-agreement is formulated (and I imagine every publisher have a different one), an author giving away a paper they authored to someone on twitter or wherever may not be allowed. Only if you'd ask the publisher and get an ok.
  
  It depends. Some publishers ask the authors to transfer copyright. Others don't. Even for the ones that do, the pre-print still belongs to the authors.
  
  What's more likely?
  
  You don't understand the exact details of this?
  
  Or a metric shit ton of published academics are flagrantly violating copyright law and openly encouraging people to do it?
  
  I can easily say that every academic I know and have as friends, which is all but two people, have surely "flagrantly violated copyright law". I have no doubt. They have even asked me for help doing it. I can also tell you that none of those have ever read one of those copyright transfers. I did, once, but I do not understand law-speak and do not remember what it said. I just know that my university had that as a policy -- because of lawyers -- what we had to do to redistribute our articles. That is also why I had a "may not" in my comment and could only refer to anecdotes, because, surprise, I do not understand the exact details about this. But you know this, because that was in my comment.
  
  Not generally. There may be fair use exceptions allowing the sharing in some situations (depending on jurisdiction) or the publisher/owner may allow it as part of the licensing contract. But I don't know in what jurisdiction/under what contract, it would be legal to copy something just because some random person asked.
- deleted
  
  Well, opinions on morality... I think the whole artificial paywalling should be abolished as being against the public interest. A large faction here seems to take a very right-wing view on property, including copyrights, and will always side with owner.
  
  How would you turn your moral intuition into a general law?
  
  deleted
  
  Tricky intuition. It would mean that authors could not transfer all rights. In that sense, it would limit what they can do with their output. Depending on how far you want to take this, it might not matter or it might not matter a lot. EG how much would you pay for the rights to an ebook if the author can always go and create a legal torrent?
  
  Do you really think it should matter if the new owner is an individual or a corporation? If you only limit corporations, then the rights will simply be transferred to individuals.

Academics don't care because they don't get paid for them anyway. A lot of the time you have to pay to have your paper published. Then companies like Elsevier just sit back and make money.

I follow a few researchers with interesting youtube channels, and they often mention that if you ask them or their colleagues for a publication of theirs, chances are they'll be glad to send it to you.

A lot of them love sharing their work, and don't care at all for science journal paywalls.

Other than be happy for that attention and being curious of what extra things you can find in their field, they get quoted and that pushes their reputation a little higher. Locking up works heavily limits that, and the only reason behind that is a promise of a basic quality control when accepting works - and it's not ideal, there are many shady publications. Other than that it's cash from simple consumers, subscriptions money from institutes for works these company took a hold of and maybe don't have physical editions anymore just because, return to fig. 1, they depend on being published and quoted.
- Sure, that's a motivation too, but they were also talking about random people who'd find a reference and were curious about their work, not just other researchers who may quote them. It's not all about h-index.
  
  When a guy literally makes, among other things, regular paleontology news reports and whole videos of his own university course material during summer breaks, and puts all that to youtube it's safe to assume he just likes popularizing his subject.

Don't mind? Hell, we want people to read that shit. We don't profit at all if it's paywalled, it hurts us and hurts science in general. This is 100% the wishes of scientific for profit journals.

I’m starting to think the term “piracy” is morally neutral. The act can be either positive or negative depending on the context. Unfortunately, the law does not seem to flow from morality, or even the consent of the supposed victims of this piracy.

The morals of piracy also depend on the economic system you're under. If you have UBI, the "support artists" argument is far less strong, because we're all paying taxes to support the UBI system that enables people to become skilled artists without worrying about starving or homelessness - as has already happened to a lesser degree before our welfare systems were kneecapped over the last 4 decades.

But that's just the art angle, a tonne of the early-stage (i.e. risky and expensive) scientific advancements had significant sums of government funding poured into them, yet corporations keep the rights to the inventions they derive from our government funded research. We're paying for a lot of this stuff, so maybe we should stop pretending that someone else 'owns' these abstract idea implementations and come up with a better system.

When you publish something in an academic journal, the journal owns the work. The journal also sells that work and it's how it makes its money.

Yes it is, and that's the problem. I work my butt off to identify mechanisms to reduce musculoskeletal injury risk, and then to maintain my employment, I have to hand the rights to that work to a private organization that profits over it. To make matters worse, I then do the work to ensure the quality of other publications for the journal through the peer review process and am not compensated for it.
the journal owns the work.

Fortunately, open access has made some inroads. It is not universally true anymore. The situation is still pretty bad, though.
- I know for the law journal that I used to edit the journal owned the final form of the edited and stylised work and granted the author a license to freely use it in perpetuity with attribution as to the original publication.
  
  So the author was free to share free copies as long as it was in the original form with the journal's name and logo on the first page, or manuscript forms as long as the original publication info was cited. My journal sold electronic and print fornats and had some licensing deals with legal research companies. But we also hosted free electronic copies for anyone that wanted to download an article
  
  For my journal, the significant costs were paid by a foundation and the university that it was a part of. The sales were just to buy like coffee for the office and stuff help offset costs. I know especially in medicine and physical sciences there's a lot more money involved in this stuff.