A brilliant tech innovator, a billionaire of billionaires, has a deep fear that artificial intelligence (AI) will eventually destroy the human species—an anxiety that much of the world shares. His fears intensify after a debate with the founder of an internet giant who tells him that ultimately, there would be many kinds of intelligence, human and artificial, which would compete for resources; and the best would win. So, he helps to set up and fund an AI non-profit laboratory proposed by a Silicon Valley entrepreneur in 2015, which would work for individual empowerment and humanity as a whole to counterbalance the commercial AI being developed by other tech giants. They would build the technology safely for the benefit of humanity and to protect the world from reckless entrepreneurs who refused to see AI was a threat. On the face of it, this seems to be a simple story on the comic-book lines of good battling evil and coming out successful. Or, in this case, safe and open AI battling for humanity versus the risky AI that benefits just a handful of reckless and greedy tech titans. The truth is more complex than that.
The fearful tech billionaire is none other than Elon Musk of the Tesla and SpaceX fame who founded, set up, named and funded OpenAI along with Sam Altman, chief executive, and Gregory Brockman, president and chairman, to counter Google’s Larry Page, who did not view AI as a threat. But as the California district court begins hearings on Musk’s lawsuit against Altman and OpenAI along with its other partners—this includes Microsoft which came on board in 2019—for changing the original charter and betraying its goals, details of dirty boardroom politics, the funding imperatives of large-scale AI and Musk’s own goals are coming under scrutiny. In his case for damages of US $150 billion from OpenAI, Musk’s lawyers allege that “the perfidy and deceit are of Shakespearean proportions.” They also claim that case against Sam Altman and OpenAl is “a textbook tale of altruism versus greed”.
It is true that Altman set out to form OpenAI in the vision of its original funder by promising to chart a safer and more open course than the rushed initiatives of profit-driven tech giants, in particular Google’s DeepMind that was leading in the race to develop Artificial General Intelligence or AGI. This is a machine having human-like intelligence to undertake a wide variety of tasks. Of crucial interest to developers elsewhere, working with notably fewer resources as in India, was OpenAI’s promise to decentralise its technology by making it open source instead of closed and proprietary. Disappointingly, the non-profit did not implement open source is a meaningful way.
The nub of the case is a decision as early as 2017 by OpenAI’s founders that they should create a for-profit entity to attract big money to fund the scale of research needed to build advanced AGI. Altman says Musk left in 2018 when discussions about who would control the for-profit business broke down. Emails have been provided to validate this point. In 2019, after Microsoft was brought in, OpenAI launched its for-profit division, whose value has soared to an incredible $852 billion now. The question is whether Altman and his partners profited illegally through that 2019 conversion. In his defence, Altman says OpenAI’s non-profit is now a foundation which holds stakes in the for-profit division and has a say in its governance. To a world that is getting increasingly anxious about the dangers of AI, does the outcome of this case have a bearing on the future?
Musk wants OpenAI to return to its no-profit status and Altman removed from the company. As a result of their unlawful actions, Altman, Brockman and other defendants “have been unjustly enriched to the tune of billions of dollars, while Musk, who co-founded their de-facto for-profit startup, has been conned along with the public, whom this vital technology was supposed to benefit,” says the lawsuit. Musk has thus brought this remedial action to divest the defendants of their ill-gotten gains.
For OpenAI and Altman in particular, the hearings have come at a difficult time. There are several pending lawsuits against OpenAI, which is also under investigation by multiple US federal agencies, including the Securities and Exchange Commission and the Federal Trade Commission, and faces a clutch of consumer advocacy complaints. The fate of OpenAI may not be of special interest to the huge constituency of the universe that views the unregulated development of AI with deep anxiety, but it is important in other ways than the immediate lawsuit. It underscores the need for government oversight and regulation of the AI industry, where power is concentrated in just a handful of companies that have managed to keep regulation at bay. According to Karen Hao, an AI expert and award-winning author who has written a book that details Altman’s and OpenAI’s meteoric rise, his extraordinary success in lobbying the US Congress to allow its unfettered growth warns of the sinister impact that the concentrated power of the industry is having on society, even as a small group of the most profitable companies in history try to match the breakneck pace at which OpenAI has grown.
As for open source, the company has done little to foster its founding principle. In an interview last year, Altman admitted that OpenAI “has been on the wrong side of history” but has done little to right that wrong. Other enterprises like Meta AI, a research division of Meta (formerly Facebook) have done more to promote the open-source concept and “help humanity”. This also benefits the company making its technology open, finds Meta’s chief AI scientist Yann LeCunn. “They came up with new ideas and built them on top of other people’s work,” LeCunn wrote in a commendatory post. “Because their work is published and open source, everyone can profit from it. That is the power of open research and open source.”
It is not a philosophy that appears to enthuse OpenAI and fellow tech titans as they pursue unbridled power and profit. The consequences of this will be felt across the universe, but it is doubtful if Musk is concerned about this anymore. Reports reveal that he is pursuing similar goals at SpaceX AI, which has just announced a massive tie-up with Anthropic—a company Musk once jokingly said could become misanthropic.