I guess Musk doesn’t want to wait for the regulators.
Regulations are for losers!
– Elon Musk, probably
Unless it hurts people I don’t like…
also Elon Musk, probably…
I assume this is the original version when the lawyers drafted it.
They’re nowhere near the idea of “Artificial General Intelligence”. If the lawsuit hangs on this, (and I’ve not read it just what you’ve posted), then OpenAI can easily say they haven’t developed that. An AGI would be an actual thinking machine capable of actually making decisions, rather than what GPT-4 is doing now which is just regurgitating what it’s been fed.
Just my two cents
I don’t think the entire lawsuit hangs on that, but that’s one of the claims. A lot of times in these kinds of lawsuits, the plaintiff will make every conceivable claim, no matter how improbable, because you’re allowed to do that and so you might as well. Kind of a throwing spaghetti at the wall and seeing what sticks approach.
I want to surface the great and pointed joke in the graphic for this story, just because others may have caught it already but I didn’t at first. : )
It’s Musk’s face photoshopped onto the lead pic for this classic from the Onion-affiliated site Clickhole:
OOhhhh - thank you. I felt like “Huh, it’s Elon Musk if he didn’t get he hair plugs.” But it also felt familiar in another way.
Advanced Sommelier exam, LOL. I bet GPT-4 also scored high on the naturalpath exam, and ancient Chinese medicine exam.
Cool okay.
For that particular point then, it can easily be refuted. However, it might not be in OpenAI’s best interest for them to admit on the record GPT-4 isn’t an AGI when some people are convinced it is.
If Musk is convinced GPT-4 is an AGI he doesn’t know the first thing about it.
Concerning…
Musk doesn’t know the first thing about a lot of the things he’s involved in.
Musk has joined the uncanny valley bros with his own AI company. He has a vested interest in driving the tulip mania.
I haven’t read the complaint yet, but may have time after work. I am, however, in the legal and regulatory space of the not-for-profit industry.
I’d have to look at their charter, and at their actual tax status (are they a 501(c)3?), but off the top of my head I can think of a couple places where Mush’s complaint could fall apart.
Not-for-profit is a tax status, not a legal mandate to not earn money. A not-for-profit can make a ton of profit, it just restricts what that money can be used for. For example, it cannot be distributed to owners/shareholders, or benefit a small group of individuals. Generally it has to be reinvested into the organization, distributed to the target beneficiaries, or equitably distributed to staff (such as bonuses). This argument won’t go too far for Mush.
The second is the purpose. Assuming OpenAI is a 501(c)3, then its work is to serve a public or humanitarian purpose. This may be a stronger argument for Mush, but maybe not. If OpenAI makes AI tools available for the masses, this could satisfy their nonprofit status. This could be by providing it on their own website, or through Microsoft. If they provide one AI tool publicly, and a different one licensed to Microsoft, this may satisfy the public good requirements. They can charge a fee to the public, too, and earn a profit. As long as the use of that profit satisfies their tax status, they are good.
And as far as violating their charter? Most companies, and nonprofits, have charters that are very expansive, as in “we’ll work towards our mission using any tool legally allowable”. I wouldn’t be surprised if their charter basically says: “we will promote AI availability to the general public using any method legally allowable.”
Does he have standing? I don’t know. He donated a ton of money, but that is what it is: a donation. He can only impose limited restrictions on those funds, such as “my money can only be used to buy arabica coffee.” However, once those funds are expended, his restrictions and claims end. And that doesn’t mean the recipient can’t use other funds contra to the purpose of his fund, say, using MS funds to buy robusta coffee. The only way he could demand the funds back is if his specific funds were used contra his restrictions.
The way rich people get around this is to give endowments to institutions. That way, their funds are available in perpetuity, and they can exert eternal control through the threat of revoking the endowment.
It’ll be interesting to see if he actually has a case, but the “they are partnering exclusively with Microsoft” is, on its own, not enough to sue.
By definition no-one owns shares of a non-profit. This is interesting in not for-profit law; because there are no owners, in most jurisdictions the board of directors is held responsible for the debts of the organization.
That’s pretty open to interpretation, and usually defined by the board of directors. Again, benefit the public can have wide application, the main issue is what they do with the money at the end of the day.
Ultimately, the thing I find most impressive about Elon Musk is his ability to have no fucking clue about, apparently, any of the things he’s been intimately involved with for years. It really is astounding. (It’s normally the kind of thing you see with executives who only care about the money-making portion of their job, not actually knowing anything about their industries, but Musk doesn’t understand the financial stuff either, and he’s a tech enthusiast.)
I assume it’s the result of powerful self-delusion mixed with the Dunning-Kruger effect mixed with not actually being that bright. He confuses his (shallow) reading of science fiction with reality, and doesn’t care much about the real-world technology itself.
Yes, somehow they are.
That…would actually explain a lot of what he says and does.
Which entity actually owns and creates the technology? Is it the non-profit or the for-profit subsidiary?
The non-profit parent company creating the for-profit subsidiary isn’t that different from the non-profit parent company purchasing some other tool, like a copy machine or truck, that the non-profit then uses to advance the goals of the non-profit.
The for-profit company is allowed to distribute profits to its investors. In this case, the non-profit company which then must use that money based of the non-profit rules. As an owner of the for-profit company, the non-profit company has some control over the for-profit entity.
In this case, my understanding is that MS is an investor in the for-profit subsidiary. That MS is involved in how the for-profit company is run. That the ability for MS to impact the non-profit company is restricted to the fact that both of them own part of the for-profit company. The amount of control each is able to exert on the for-profit company related to how the ownership is split.
Which also makes me think all the technology and AI belong to the for-profit entity. The non-profit is really just there to exert control over the for-profit company instead of the for-profit only answering to general investors. Which I also assume then, that MS owns less of the for-profit company than the non-profit owns. Giving the non-profit majority control of the for-profit company.
Yeah, and based on my very quick look through of their 990 their status looks correct, and their mission as summarized on page 1 doesn’t directly contradict working with Microsoft:
OpenAIs goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return. We think that artificial intelligence technology will help shape the 21st century, and we want to help the world build safe AI technology and ensure that AI’s benefits are as widely and evenly distributed as possible.
None of that prohibits them from working exclusively with MS or making a profit. It’s a vague enough “public good” that it probably satisfies their 501(c)3 status.
Unfortunately there are a lot of dodgy nonprofits, big and small. I just can’t tell here if OpenAI is one of them.