OpenAI Should Remain Nonprofit

Artificial intelligence Investments Non-Profit
Facebooktwitterredditlinkedin

            Marble Falls     We all know the line about “even a broken clock is right two times a day.”  Please keep that in mind when I say that I have to agree with the evil Trump Rasputin, Elon Musk on one thing.  It almost hurts me to say that Musk, the world’s richest man, now trying to starve millions, fire hundreds of thousands, a hater of immigrants unless they are working cheaply for him and his companies, and for all intents a racist and misogynist of rare proportion, may be right about anything.  Yet, even though he is clearly doing the right thing for the wrong reason, I have to say, I think he’s right to oppose the attempt to change the nonprofit structure of OpenAI to a cash-out, garden variety Silicon Valley for profit money machine.

Speaking for all of the techno-peasants, I don’t know if artificial intelligence or AI for short, is going to be the internet of today and change everything in the world.  I also don’t know that it’s going to create nonhuman control structures that we should fear for our lives and values.  Not knowing, I had felt some assurance that the nonprofit structure and mission of OpenAI to make sure that AI was developed for the betterment of all humanity, rather than another techie get-rich-quick scheme, was a very good thing.

Of course, like everything that has to do with Musk, his tactics have typically made something of a mess of all of this, even though they may be somewhat successful.  His strategy complicates the issue, because he was once part of OpenAI, but is now an open competitor with a separate company also in the AI development race, valued perhaps as much as $75 billion.  There is no way to pretend the he is acting in the public interest, when his self-interest is so blatant, but, still, that doesn’t mean he’s wrong that OpenAI should remain nonprofit or be compensated hugely in any transition.

Musk has filed suit in California to keep OpenAI nonprofit.  California may have some of the clearest and strictest rules for nonprofits in the country, as well as for other corporations, which is one of the reasons that Musk moved all of his legal operations to Texas.  They are watching all of this like hawks, according to the California AG,

Robert Bonta, California’s attorney general and a Democrat, said this week in an interview that the state was scrutinizing OpenAI’s plan to shift to a for-profit structure. “There’s a way to do it right. There’s a way to do it wrong, and we’re monitoring to make sure they do it right,” he said, adding that his office was also closely watching Mr. Musk. We are “monitoring everything he does,” Mr. Bonta said.

            In another brilliant tactical move, Musk, with some of his investors, made an offer to buy the nonprofit OpenAI for about $97 billion.  The board of OpenAI rejected the offer even before receiving it.  Musk didn’t exactly put OpenAI in play as much as begin to place a value on the nonprofit, and what it would cost its CEO Sam Altman and the board to compensate the nonprofit in the process of changing the structure.  In California and other states, we have seen this play out in the process for big players like the nonprofit Blue Cross–Blue Shield being converted, which has created foundations worth hundreds of millions in some cases.  Recently, we saw Blue Cross stopped in Louisiana because of complaints that it was being done on the cheap and for personal inurement.  Remember that Altman and the directors stand to gain millions in this transition, if, and when, they are successful.

Musk is likely just trying to throw rocks in the road for his own benefit and that of other competitors to stall their process, but for those of us in the cheap seats having a nonprofit replacement emerge from an OpenAI buyout is a good thing, especially if it might end up with a $100 billion kitty.  DeepSeek from China offers the possibility that a more open source and accessible AI option is very possible, and that might be part of the new mission.  Finally, having a pile of money to protect the old mission of OpenAI that it benefits humanity and does no harm would also be a win all around.  It might end up as one of the few times that Musk looking after himself ends up helping the rest of us.

 

Facebooktwitterredditlinkedin