“WGIRL FOOD do you have more confidence in him? Getting your technology from a non-profit company, or a for-profit that is completely controlled by one person?” asked Brad Smith, president of Microsoft, at a conference in Paris on November 10. That was Mr. Smith’s way of recommending OpenAIthe beginning behind ChatGPTand demolishing Meta, Mark Zuckerberg’s social media behemoth.
In the last few days OpenAInon-profit management has looked rather more attractive. On November 17, seemingly out of nowhere, their board fired Sam Altman, the startup’s co-founder and chief executive. Mr. Smith’s own boss, Satya Nadella, who heads Microsoft, was told that Mr. Altman was fired just minutes before Mr. Altman himself. Don’t forget that Microsoft is openAIthe largest shareholder, having backed the startup to the tune of more than $10bn.
By November 20 most OpenAIthe workforce of 700 had signed an open letter giving the remaining members an ultimatum: resign or the signatories will follow Mr Altman to Microsoft, where he was invited by Mr. Nadella to be the head of the new interior. AI laboratory.
What’s going on has drawn attention to the OpenAIIt’s an unusual structure, and its table even more unusual. What exactly is he tasked with doing, and how could he take the bag to the hottest boss ai start without any voice of his investors in the matter?
The company was founded as a non-profit in 2015 by Mr Altman and a group of Silicon Valley investors and entrepreneurs including Elon Musk, the mercurial billionaire behind Tesla, X (formerly Twitter) and SpaceX. The group has collectively committed $1bn towards OpenAIthe goal of building artificial general intelligence (AGI), as AI experts refer to a program that outperforms humans on most intellectual tasks.
After a few years OpenAI he realized that to achieve his goal, he needed money to pay for expensive computing power and top-notch talent – especially since he says only $130m or so came in of the original $1bn pledge. So in 2019 he created a for-profit subsidiary. Profits for investors in this venture have been capped at 100 times their investment (although this cap will rise with a rule change to 20% per year starting in 2025). Any profits above the cap flow to the parent nonprofit. The company also reserves the right to reinvest all profits back into the company until it manages to create AGI accomplished. And once it is achieved, the result AGI it is not intended to generate a financial return; OpenAIMicrosoft’s licensing terms, for example, cover only “pre-AGI“technology.
Decide if and when AGI achieved depends on OpenAIthe board of directors. Unlike most startups, or indeed most companies, investors don’t get a seat. Instead of representing OpenAIfinancial backers, the organization’s charter works on leaders by representing “human” interests.
Until last week’s events, humanitarian representatives consisted of three OpenAIco-founders (Mr. Altman, Greg Brockman and Ilya Sutskever) and three independent members (Adam D’Angelo, co-founder of Quora; Tasha McCauley, tech entrepreneur; and Helen Toner, from the Center for Security and Technology who comes up). , other non-profits). On November 17, four of them – Mr. Sutskever and the three independents – lost confidence in Mr. Altman. Their motives are still unclear but they may have to do with what the board saw as the pursuit of new goods with a pair without worrying enough about it. AI safety.
The company’s bylaws from January 2016 give board members extensive powers, including the right to add or remove board members, if a majority agrees. The earliest tax filings from the same year show three directors: Mr. Altman, Mr. Musk and Chris Clark, an Open activistAI an employee. It is not clear how they were chosen, but thanks to the bylaws they could now appoint others. By 2017 the original three were joined by Mr Brockman and Holden Karnofsky, chief executive of Open Philanthropy, a charity. Two years later there were eight members on the board, although Mr Musk had resigned by then due to a dispute with Mr Altman over the direction of OpenAI was taking Last year it was down to six. Throughout, he was responsible only for himself.
This strange structure was designed to ensure that OpenAI they may resist external pressure from investors, and may prefer a quick profit now AGI for later mankind. Instead, the board’s unprofessional dismissal of Mr. Altman has put pressure on OpenAIinvestors and workers. That small part of humanity, at least, clearly feels misrepresented. ■
