Elon Says Copyright/AI Lawsuits Don’t Matter Because ‘Digital God’ Will Arrive Before They’re Decided

from the that's-not-how-any-of-this-works dept

So, we already wrote about the biggest headline grabbing moment from Elon Musk’s Dealbook interview with Andrew Ross Sorkin yesterday, but there was another crazy, Techdirt-relevant one involving copyright and AI. As we’ve explained over and over again, copyright is the wrong tool to use to regulate AI, and using it will lead to bad outcomes.

But, absolutely nothing in this bit of the interview made any sense at all (from either side):

It starts out with a drop dead ignorantly wrong question from Aaron Ross Sorkin, who seems wholly unprepared for this:

ARS: So, one of the things about training on data, has been this idea that you’re not going to train on… or these things are not being trained on people’s copyrighted information. Historically. That’s been the concept.

Elon: Yeah that’s a huge lie.

ARS: Say that again.

Elon: These AIs are all trained on copyrighted data. Obviously.

ARS: So you think it’s a lie when OpenAI says that… none of these guy say that they’re training on copyrighted data.

Elon: Yeah, that’s a lie.

ARS: It’s a lie. Straight up?

Elon: Straight up lie.

So… there’s a lie in there, but it’s Andrew Ross Sorkin saying that any AI company claims that it doesn’t train on copyright-covered data. Everyone admits that. They say that doing so is fair use (because it is). So the entire premise of this discussion is wrong. Here’s OpenAI admitting in court that, of course, it trains on copyright covered material. It’s just that it believes fair use allows that (because it does).

So, in one sense here, Elon is right to push back on Sorkin’s claim. But Musk is misleading, because he appears to buy into the false premise of Sorkin’s question that AI companies say they’re not training on copyright-protected data. If Musk had any idea what he was doing he would have told Sorkin his premise was wrong, and that no companies deny training on such material.

From there, Sorkin goes on an even more confused discussion, claiming that while snippets of articles on ExTwitter are fair use, because, combined, people might post a full article, it might not be any more… but that’s… not how any of this works anyway. Someone give him a Copyright 101 lesson, because this is embarrassing.

Either way, Musk then made the whole thing… um… fucking weird. Because as Sorkin kept trying to press Musk on the copyright lawsuits, Musk did this:

Musk: I don’t know, except to say that by the time these lawsuits are decided we’ll have Digital God. So, you can ask Digital God at that point. Um. These lawsuits won’t be decided on a timeframe that’s relevant.

If someone you knew started saying stuff like that, you’d have them checked out.

Whether or not you believe that AGI (Artificial General Intelligence) is on the way or not, or that it might create “Digital God,” the idea that this is coming before these lawsuits are decided is… um… not realistic. But, even if we do somehow reach AGI within the next few years as these lawsuits play out, the idea that such an AGI would obsolete the courts and/or copyright law is similarly wishful thinking.

Hell, we’ve argued for years that the internet itself has already obsoleted copyright laws, but they’re still sticking around and getting dumber all the time. I’d love for it to be true that technology further obsoletes copyright law and moves things to a better overall system… but it’s not going to happen because “Digital God.”

Of course, perhaps if Elon truly thinks Digital God is coming in the next few years, it explains why he doesn’t care about advertisers on ExTwitter any more.

Filed Under: , , , ,
Companies: openai, twitter, x, xai

Rate this comment as insightful
Rate this comment as funny
You have rated this comment as insightful
You have rated this comment as funny
Flag this comment as abusive/trolling/spam
You have flagged this comment
The first word has already been claimed
The last word has already been claimed
Insightful Lightbulb icon Funny Laughing icon Abusive/trolling/spam Flag icon Insightful badge Lightbulb icon Funny badge Laughing icon Comments icon

Comments on “Elon Says Copyright/AI Lawsuits Don’t Matter Because ‘Digital God’ Will Arrive Before They’re Decided”

Subscribe: RSS Leave a comment
94 Comments
PaulT (profile) says:

Re:

“The more he speaks, the more he destroys his own “genius” mystique”

I think that’s long gone at this point, outside of a small number of cultists.

I forget who originated it, but there’s a quote that goes something like “I know nothing about cars, people said he was a genius with cars so I believed them. I know nothing about rockets, so when people said he was a genius I believed them. I know a lot about software, and Musk has said the dumbest things I’ve ever heard about it, so I now know he’s not a genius and will avoid his cars and rockets”.

He’s probably even destroyed the mystique around CEOs generally as well. At one time I might have conceded that there’s some work involved even if they tend to earn too much money even while failing, but this guy is CEO of several companies and seems to spend his whole time shitposting on Twitter. At least others tended to pretend to work for a living.

Andu says:

Re: Re: its malignant narcissism.

plain and simple. Same with Trump. I have witnessed narcissism in different forms up close and personal my whole life. This is textbook. what makes it a lot worse than my mom or any other random person is the wealth and power he weilds. Which both makes him more dangerous but also his narcissism more unchecked.

He surrounds himself with yesmen. That’s a narcissist’s wet dream. Constant adulation, ever word out of their mouth is profound to those around them, etc.

There is no cure for narcissism. Self awareness is the only thing that can check it, and I’d venture human beings would go extinct before Musk got an ounce of self awareness in him…

This comment has been deemed funny by the community.
Anonymous Coward says:

Re: Multiverses are not allowed under existing copyright law

Courts have never found that creating new universes based on existing universes is fair use. You could make a substantially different universe but infinite universes obviously violates the DMCA. And good luck creating a stable universe without violating the fine structure constant patent.

This comment has been flagged by the community. Click here to show it.

Anonymous Coward says:

Do you think “fair use” would allow a film student to go to Pirate Bay and download every Disney movie he can find?

Fair use allows him to learn from the movies and produce works that imitate their style, but the first step is still to buy a legitimate copy of the movies.

And AI developers generally haven’t bother with that step, because it’s too expensive. Their copy of training data is no different than any other pirated file.

This comment has been deemed insightful by the community.
Ethin Probst (profile) says:

Re:

Uh, no, but this isn’t how fair use works. 17 U.S.C. 107 clearly states:

Notwithstanding the provisions of sections 106 and 106A, the fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include—

(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
(2) the nature of the copyrighted work;
(3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
(4) the effect of the use upon the potential market for or value of the copyrighted work.

The fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made upon consideration of all the above factors.

This section doesn’t bar a finding of fair use in other scenarios that are not delineated within. If your interpretation of fair use was realized, it would make even making AI completely impossible to begin with, because in order to make an LLM you need an incredibly large corpus of data. The total licensing cost of that would be potentially in the trillions of dollars, if not quadrillions. So I’m pretty sure a court would find that trying to make fair use in those circumstances inapplicable would make the burden on those who wish to use the work far too burdensome and therefore would annul the original complaint of copyright infringement to begin with. And we’re just talking about LLMs! If you want to make it impossible to train actually good AIs, then we can realize your ridiculous notion of fair use. But even realizing that kind of interpretation risks running afoul of the copyright clause, which is very clear: “[the United States Congress shall have power] To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” The important part is “promote the Progress of Science and useful Arts”. Making it untenable to make AIs without paying impossible-to-pay copyright licensing fees would go against that part of the clause, because AIs, in order to be as good as they can be, need huge datasets.

Anonymous Coward says:

Re: Re:

If the only way to make an LLM requires breaking the law, then it’s illegal to make an LLM.

Not that it actually requires “quadrillions” of dollars. Some content libraries have offered to license their collections for the purpose of training LLMs, with royalties paid to the authors. But AI developers should not expect to get something for nothing.

This comment has been deemed funny by the community.
Anonymous Coward says:

Re: Re: Re:

Making it untenable to make AIs without paying impossible-to-pay copyright licensing fees would go against that part of the clause, because AIs, in order to be as good as they can be, need huge datasets.

If the only way to make an LLM requires breaking the law, then it’s illegal to make an LLM.

“That post can’t stop me because I can’t read!”

Anonymous Coward says:

Re: Re: Re:4

“So education of people is not fair use, then.”

Not in any educational establishment that isn’t wholly funded by the state, according to your disingenuous interpretation of fair use. The fact is that fair use (fair dealing in other countries) doesn’t apply only when you can’t afford licensing fees. It applies when you can show that your copying and subsequent use falls within one of the narrow exceptions to copyright law regardless of ability to pay fees, and you can be found guilty of copyright infringement if your copying and subsequent use don’t fall within one of the narrow exceptions, and any complaint of “But I couldn’t afford the licensing fees!” would very quickly get short shrift.

Anonymous Coward says:

Re: Re: Re:5

So everyone should remain dumb if they’re poor, then.

And stay in their corp-assigned positions unless they “prove to be capable” without any sort of proof proving they deserve a promotion (that they should pay for).

No upward class/economic mobility then. 90%, if not more, should remain serfs or worse, slaves.

Because, as you said, education, unless it’s “state-owned”, is not fair use or free and the price determined by copyright maximalists, then.

Anonymous Coward says:

Re: Re: Re:7

But when you’re sued for pirating a book, that argument won’t make any difference.

That’s a fair point to make, and why you don’t see publishers suing students for pirating their books – the publishers know that there’s no money to take, and shitting on poor students over what’s generally considered a universal good like education is pretty shit for PR. It’s the distributors they go after.

Now things are a little murkier when it comes to university students and researchers downloading from Sci-Hub instead of paying Elsevier, but Elsevier has long since indicated that they’re entirely willing to look like moneygrubbers, to the point where the governments of Taiwan and Germany and other countries have basically told their universities to tell Elsevier to go fuck themselves instead of paying their exorbitant asking prices. Sympathy for publishers is as close to rock bottom as it’s been, and frankly, the publishers mostly did it to themselves.

Anonymous Coward says:

Re: Re: Re:7

Oh, I am familiar with your shitty copyright maximalist argument.

And to that, I say, if the product I have to pay for is garbage, I don’t get to find a better product?

That you also seem to hate education as fair use is also telling, and I am all for reasonably spending for your education.

Since if education is a premium, then only the rich can afford it, and the poor are left to rot.

Is that what you want? A class divide so apparent that Karl Marx is the only way to change things? Or worse, a general violent revolt?

After all, if even trying to communicate involves breaking the law

This comment has been flagged by the community. Click here to show it.

Stephen T. Stone (profile) says:

Re: Re: Re:2

There is actually a significant legal difference between simple downloading and torrenting, in that torrenting involves redistribution.

So does downloading an ebook from a website. The difference is that redistribution via torrent is a multi-user multi-source process instead of a single-user single-source process.

Anonymous Coward says:

Re: Re: Re:2

You can legally download a pirated copy of a work only if you paid for a legitimate copy of the work. So it is illegal to download a copyright protected work without paying, even if you do not upload it to anyone else.

From a practical perspective, you are more likely to be sued if you upload a work than if you download it (because there are no exceptions for the former).

However, the AI developers in this lawsuit never paid for legitimate copies of the works in question.

This comment has been deemed insightful by the community.
This comment has been deemed funny by the community.
Anonymous Coward says:

Re: Re: Re:3

“You can legally download a pirated copy of a work only if you paid for a legitimate copy of the work.”

You’d best let Project Gutenberg know that, then. I’ve been downloading Charles Dicken’s books from them without paying anything for years, and they were pirated in the 19th century.

bhull242 (profile) says:

Re:

And AI developers generally haven’t bother with that step, because it’s too expensive.

First, how do you know how the content was accessed to create the training data? You’re just guessing here.

Second, that is irrelevant to whether or not the use is infringing, which is what’s at issue here. They aren’t being sued for copying the training data; they’re being sued for using that data to train AI.

Rocky says:

Re: Re: Re:

And AI developers are being sued because they allegedly obtained training data from pirate websites.

Perhaps you should actually read what the lawsuit claims before making sweeping statements.

I’ll leave it up to you to read and actually understand the lawsuit since there is no point explaining it to you until you do: https://storage.courtlistener.com/recap/gov.uscourts.cand.415175/gov.uscourts.cand.415175.1.0_1.pdf

This comment has been flagged by the community. Click here to show it.

Anonymous Coward says:

Re:

You put on clown shows, people gonna watch. People gonna talk about it.

Sorry that the people who laugh at clowns because they’re stupid say different stuff from the people who like clowns because they think they’re awesome. But pretty sure that the clown fans talk waaaay more about the clowns than the, uh, lamestream sheeple.

This comment has been flagged by the community. Click here to show it.

Anonymous Coward says:

Re: Re: Re:2

On the other hand, I kinda really want to see Musk being dumb enough to do a Prenda Law and file stacks of evidence in the form of Techdirt comments in a bid to subpoena all our personal information.

Not because I think he’d succeed, but because I still get a kick out of knowing that in John Steele’s rabid fury he put my comments calling him an assclown in the official record.

Anonymous Coward says:

Re:

“writing ceaselessly about Musk and Twitter. He’s like Donald Trump was to the failing lamestream media!”

Perhaps there is this so called ceaseless writing about Elmo and his fav toy because Elmo ceaselessly fucks up everything he touches. If Elmo were to sit down and act like a human being for once then maybe …. oh idk – he is not capable is he.

Anonymous Coward says:

Re:

Is fair use a consideration when i read a book or look at a painting?

Only if you’re a copyright maximalist. Consuming media does not fall under copyright law. Hell, consuming media isn’t illegal in any country, unless there are very specific laws regarding the acquisition and consumption of media regarding certain topics.

Is it a consideration if i deeply study works, so they inform my style (consciously or unconsciously) later?

The case law is shaky at best, but yes. It is fair use if you use the Mona Lisa and react to it. It is fair use to make Star Trek fanfiction. It is fair use to make fanart of Game of Thrones characters.

And yes, it is also fair use to learn from a fucking painting or novel. Just like how it is MOTHERFUCKING FAIR USE TO LEARN MATH, ENGLISH, AND SCIENCE FROM A FUCKING TEXTBOOK.

If you say otherwise, then you are worse than even the Romans, Greeks and even the kings and queens of the Middle Ages. Who still thought that some form of education was still useful for their subjects.

Anonymous Coward says:

Re: Re:

“It is fair use if you use the Mona Lisa and react to it.”

No it’s not. The Mona Lisa was painted from 1503 to 1506, so is therefore in the Public Domain. Since anything in the Public Domain is out of copyright by definition, there is no consideration of fair use necessary because that only applies to works still under copyright.

Anonymous Coward says:

Re: Re: Re:

Since anything in the Public Domain is out of copyright by definition, there is no consideration of fair use necessary because that only applies to works still under copyright.

Eh… I wouldn’t count on that. Copyright fanatics have proven that they are entirely willing to push for works to be removed from the public domain.

Anonymous Coward says:

AGI will be interesting

AGI (Artificial General Intelligence) will be an interesting time. From a scientific standpoint, it will an astounding time to be alive, as our knowledge of science should expand exponentially.

Applying AGI to other matters that are subjective, such as court cases will be interesting too. Not just on how to train an AGI (plenty of cases have had objectively terrible rulings), but then how AGI’s rulings on new cases will be evaluated based on each evaluator’s political and religious leanings.

Anonymous Coward says:

Re:

“as our knowledge of science should expand exponentially.”

This will happen in a time when a significant number of people think the earth is flat and a significant number of our politicians are actively destroying the education system …lol.

AI, by definition, does not think outside the box, aka its training. Humans on the other hand, do.

Anonymous Coward says:

So I'm down with seeing how this pans out...

In my experience, prophets of some kind of deity’s apocalyptic arrival and the true believers in their cult have this tendency when the predictions do not come to pass of attempting to… uhm… “expedite” the process en masse on a more personal level.

This could solve a lot of the problems TD has been struggling for the last year to keep up with all the newest developments in.

Patrick says:

AI = An Idiot

AI is a theocracy pushed by fascists dictating what the future should be. confirmed.

They could also call it Digital Satan, though, cause it appears attractive but will deceive you in every way.
But same thing, Digital God, isn’t it?

It’s all a promise of a utopia, as well, but a utopia for the rich where the poor are useless and somehow “magically” pay themselves. Cause for sure the rich isn’t going to pay them cause they make AI to get even richer.

Absolute lolcows.

Add Your Comment

Your email address will not be published. Required fields are marked *

Have a Techdirt Account? Sign in now. Want one? Register here

Comment Options:

Make this the or (get credits or sign in to see balance) what's this?

What's this?

Techdirt community members with Techdirt Credits can spotlight a comment as either the "First Word" or "Last Word" on a particular comment thread. Credits can be purchased at the Techdirt Insider Shop »

Follow Techdirt

Techdirt Daily Newsletter

Ctrl-Alt-Speech

A weekly news podcast from
Mike Masnick & Ben Whitelaw

Subscribe now to Ctrl-Alt-Speech »
Techdirt Deals
Techdirt Insider Discord
The latest chatter on the Techdirt Insider Discord channel...
Loading...