There was the headline-grabbing coup: the OpenAI board’s abrupt ouster of its co-founder and chief executive, Sam Altman. We are now on the eve of a second, even more critical, coup that will consolidate control of one of the most powerful and promising technologies on the planet under the leadership of one of this world’s technological titans. country.
On Monday, it was announced that Microsoft was hiring Mr. Altman and another OpenAI co-founder, Greg Brockman. Microsoft had already invested more than $13 billion in OpenAI; its absorption of OpenAI management – and the likely hiring of hundreds of OpenAI staffers who signed a letter saying they would leave to join Microsoft unless the board resigns – effectively completes its redemption. OpenAI could find a way out of this self-made disaster, but any solution would likely require satisfying its irate investors by making its board more accountable to their interests.
It is not ironic that OpenAI’s board, which was apparently concerned about the safety of its wildly popular product, has set in motion events that will likely move it to executives more subject to market pressures for growth fast. The likely outcome of this fight is a nail in the coffin of the larger effort to build a non-commercial version of artificial intelligence that would serve the public at least as much as it sought profits.
OpenAI was founded in 2015 with the explicit mission of creating an alternative to for-profit AI models developed elsewhere. It was established as a non-profit organization and its stated mission is to “ensure that artificial general intelligence benefits all of humanity.”
But ultimately, the cost of developing AI was too high to attract investors to a nonprofit. The cost of training just one of OpenAI’s chatbots, GPT-4, is estimated at $100 million. Thus, in 2019, OpenAI implemented a hybrid model: it remained a non-profit organization but created a commercial arm which it described as “a hybrid of a for-profit organization and a nonprofit” which she said would “increase our ability to raise capital.”
The hybrid model allowed Microsoft to invest billions of dollars and acquire a 49% stake in the for-profit arm of OpenAI. (The nonprofit parent company owns only 2 percent.) In other words, Microsoft had already acquired a significant stake in the promising startup. Poaching its employees would be little more than a formality, although it would be made much easier by the ignorant actions of OpenAI’s board.
Generative AI (large language models lauded for their ability to create writing, speech, and images that are plausibly human-like) has been ceded to the for-profit sector. All the major AI companies – Microsoft, Alphabet, Meta, Anthropic, Hugging Face – are for-profit companies seeking to reward their investors.
If AI has the power that its creators claim to have, should its future rest solely in the hands of the commercial sector, particularly companies whose models have largely been about taking our data and using it to manipulate us?
“The risk of entrusting AI development entirely to demonstrably unreliable Silicon Valley companies is too high,” says technologist Bruce Schneier. He and others are pushing for the United States to support an AI public option — technology built through a government-led program “that could support democracy rather than undermine it.”
A public option could look like a European attempt to create Gaia-X, a European cloud service. Or it could be more like China, which has invested heavily in building AI capacity through public-private partnerships.
In the United States, AI is largely a commercial enterprise. In 2023, the US government is expected to invest $1.8 billion in basic AI research, while venture capital has invested nearly $18 billion in AI start-ups in the past year alone. third trimester. Access to AI is largely controlled by three companies – Amazon, Alphabet and Microsoft – which control two-thirds of the global cloud computing market, a market critical to creating powerful AI models.
U.S. public investment in cloud computing would help Fei-Fei Lee, AI pioneer and director of Stanford’s Human Centered Artificial Intelligence. It is working to develop better AI to help keep hospital patients safe by automatically analyzing video footage from inside the hospital. But it said it couldn’t afford the massive computing power it needed to analyze huge amounts of video data.
“I can’t build the kind of model that I would like to be able to help keep patients safe,” she told me. “The public sector is seriously under-resourced. »
Some have argued that the best approach is to break Big Tech’s control of the cloud computing market – increasingly a choke point for AI, given the enormous amount of computing power needed to collect and analyze the massive data sets that power artificial intelligence.
We have already seen what these companies have done when they control multiple parts of a market. Amazon is being sued by the Federal Trade Commission, which accuses it of using its commercial market power to give its own products an unbeatable advantage. And Google is being sued by the Justice Department for paying Apple and others not to use competing search engines. Similarly, the big three cloud providers also offer consumer-facing services and could easily use their market power, for example, to prevent competitors from accessing their services.
British regulators have already opened an investigation into whether Amazon and Microsoft are abusing their market power in cloud computing by making it too difficult for customers to switch cloud providers.
Barry Lynn, executive director of the Open Markets Institute, argued that cloud computing has become too important to be left to for-profit companies. Not only is it the backbone of artificial intelligence, but it is also the backbone of almost every computer system, from office productivity applications to gaming and social networking.
“This is fundamental infrastructure for our entire online economy,” he said. “The fact that there are only three companies doing this gives them all sorts of powers, including the power to exclude competitors or set prices in a discriminatory manner, and it also leads them not to grant enough attention to stability and resilience.”
Mr Lynn said cloud infrastructure should be separated from other big tech activities and regulated as an essential public service.
Imagine for a moment that cloud computing was a public resource that anyone could use for a small fee, like public libraries. Innovation would flourish. Dr. Lee could use as much computing power as needed to train his patient safety models. OpenAI would never have needed to go to Microsoft for its computing capacity, and it could have been one of many nonprofits building large AI models.
This would be the ultimate blow: taking back the power of calculation in the name of the public.