Thursday, March 7, 2024

So Why Is Musk Suing OpenAI Anyway?

Wesley-tingey-mlGJB5D9IS0-unsplashWhile I'm sure that more could be said about Musk's lawsuit against OpenAI, I want to end the week exploring why he might be suing. This post, of course, is going to be purely speculative--I don't know Musk and I don't know his motivations. Still, there's at least some value in nonprofits having a sense of the reason disaffected donors might sue.

So what might be motivating Musk?

  • He's worried about the risks of artificial generative intelligence and thinks the nonprofit form is the best way to control those risks

I think he legitimately is a believer. I don't personally have any idea whether his concerns about AGI are real or are overblown, but he seems legitimately concerned that AGI could be harmful to humanity. So the first half of that sentence may have some explanatory power for why he filed the suit.

But the second half? Maybe he also does believe that using a nonprofit to develop AGI is necessary to contain that risk. But there are a couple things that give me pause. The first is, he's developing his own AI in a for-profit Nevada corporation. (Nevada's corporate lookup tool doesn't seem to let me link directly to a company filing, but if you enter "X.AI Corp.," you'll go to the registration for a "Domestic Corporation." If you enter, for example, "Three Square," you get a listing for a "Domestic Nonprofit Corporation." So maybe Musk believes that it's critical that AGI be developed through a nonprofit unless he's the one developing it, but at best, the existence of X.AI Corp. undercuts his stated concern.

On top of that, OpenAI has released a (select) series of emails from and to Musk, dating back to 2015 in which they discuss the need for OpenAI to raise more money than it realistically can through charitable means. Musk acknowledges the concern and suggests that Tesla (a for-profit corporation) should get involved.

So even if he's legitimately worried about the potential destructive power of AGI, he seems fine with its development intertwining with a profit motive.

  • He wants the money he donated back

I suspect that this, too, is true. But I'm not convinced that it's motivating his lawsuit. According to the complaint, in total, Musk contributed $44 million to OpenAI.

Now $44 million is absolutely real money. But as of this writing, Forbes pegs Musk's net worth at about $196 billion. So he's fighting over 0.02% of his net worth. And frankly, litigation can be expensive. Eight years ago, Above the Law speculated that Irell partner Morgan Chu was billing $2,000 an hour. Mr. Chu's name is on the complaint, and I can only assume that top-flight attorneys are even more expensive now than they were eight years ago.

And also, whatever the possible financial reward of this suit, it has to be adjusted for risk. As I've laid out this week, I suspect the chances that Musk will win are low; on a risk-adjusted basis, I think this lawsuit has an expected negative value (especially in light of the emails OpenAI released, which undercut his assertion that the Founding Agreement, made up, among other things, of emails, unequivocally prohibited any type of for-profit investment in OpenAI). 

  • He wants to hurt a competitor

As discussed, Musk has now formally entered the AI space, meaning he's directly competing with OpenAI. Maybe he filed the lawsuit to get a leg up on the competition.

I'm a little skeptical of this, though. It's true that they're competitors, and it's true that defending the suit will require OpenAI to expend some degree of time and assets to defend, but the suit strikes me as more of a nuisance, on that will likely be dismissed at the motion to dismiss phase. I don't think it poses any real threat to OpenAI, and I don't think that Musk could possible believe it would, at least not after sitting down with competent attorneys.

  • He's trying to emphasize a point

Frankly, I find this the most likely. Over the last number of years, performative litigation (that is, litigation meant for communicative purposes) has become tremendously popular among that class of people who can afford the costs. Sometimes it's meant to punish, but here I suspect it's meant as an exclamation point at the end of a sentence that reads "Artificial general intelligence could be really dangerous if we don't do it right."

Full disclosure: I'm entirely opposed to performative litigation. I think it's bad. But I also think it has a lot of explanatory power, and here, I think it goes a long way toward understanding why Musk filed this suit.

Samuel D. Brunson

Photo by Wesley Tingey on Unsplash

Current Affairs | Permalink


Post a comment