Sam Altman urges lawmakers against regulations that could ‘slow down’ U.S. in AI race against China
Sam Altman stressed the importance that the efforts of the American AI are not “slowed down” by refined regulations, as the CEO of Openai testified on Thursday in Capitol Hill for the first time in two years and faced an audience of largely friendly and deferential legislators.
The audience, focused on the victory of the race against China to dominate AI around the world, has marked the first appearance of the Altman Congress since its high -level testimony in May 2023, which propelled it on the world scene and sparked generalized media coverage only six months after the launch of Chatgpt.
Dressed in a conservative manner on a dark gray suit, a blue tie and a white shirt on Thursday, Altman faced a few hostile questions or a hindsight of the American senators.
In response to Senator Ted Cruz, who asked how close China was in the AI, Altman replied: “It is difficult to say how much we are in advance, but I would not say a lot of time.” He said he thought that the models of Openai, Google and others are the “best models in the world”, but added that continuing to win will require “reasonable regulations” which “does not slow us down”.
This spirit of innovation and entrepreneurship is only American, continued Altman, stressing that “none of this is rocket science. We just need to continue doing the things that have worked for so long and not make an idiot error. ”
Oh, what a difference made two years. Altman's testimony was away from its appearance in 2023, when the main objective of the legislators was the security and regulation of the AI. Altman himself urged Congress at the time to implement regulations for AI technologies, emphasizing potential risks if they are not controlled. He proposed the creation of a new federal agency responsible for granting licenses and audit of AI models, in particular those who have capacities that could present significant risks.
To be fair, the hearing of Thursday, in which Altman was joined by the CEO of AMD, Lisa Su, the CEO of Coreweave Michael Intrator and the president of Microsoft Brad Smith, was all about the geopolitical issues at stake in the AI era of today-that is to say the race to dominate the AMA overall by reinforcing the American capacities.
He offered the perfect opportunity to Altman to develop OpenAI for countries, the global expansion of his $ 500 million project Stargate Mega Data Center announced yesterday. The future of artificial general intelligence (AG), said Altman in his pre-written testimony, “can be almost unimaginably brilliant, but only if we take concrete measures to ensure that an AI-led version, built on democratic values such as freedom and transparency, prevalent on an authoritarian version.” This is particularly important, he said, because AI systems become more capable and people want to use them even more.
Responding to this demand “requires more fleas, training data, energy and superordinators,” he said.
“Infrastructure is fate, and we need it much more.”
Several of the questions have launched on Altman, Openai's commercial policies and practices, but legislators have dealt with altman in a contribution unlike the grills that other CEOs from Big Tech received on Capitol Hill. For example, senator Amy Klobuchar, who said that she had never heard of AI's “hallucinations” (although they have been widely reported since the launch of Chatgpt), seemed satisfied when Altman said that even if Openai models are not 100%precise, “users are intelligent”-that is to say that people understand what they can do. »»
And when Senator Lisa Rochester (D-DEL) slightly questioned Altman on the calendar of Openai's efforts to restructure her for-profit business, Altman said that it was grateful to have the chance to explain a “complicated thing that I think it was distorted”. The company's plan has always been to have a robust non-profit organization, he explained: “We hope that a non-profit organization will be one of the best, perhaps one day, the best non-profit organization in the world.” The restructuring of the for-profit arm as a company of public benefits with the same mission, he added, “would allow us to lift the capital necessary to provide these tools and services to the level of quality and availability to which people wish to use them, but always remain on our mission.”
However, there was a notable absence of references to the security of AI in the testimony of Altman, which contrasts strongly with its comments in 2023, which mentioned dozens of times in terms of IA security.
But Openai has transformed beyond the recognition of the AI research laboratory which burst into public awareness when Chatgpt made its debut. From last summer, half of the Openai staff who once focused on the long-term risks of the super-power AI had left the company, in particular the co-founder and scientist in chief Ilya Sutskever. Almost 20% of the company Now works in sales. And yesterday, the CEO of Instacart, Fiji Simo, joined Openai as CEO of applications, supervising the product offers and business operations of the company.
“Openai is not a normal company and will never be,” said Altman in his pre-written testimony. But he seemed perfectly relatable by closing with an optimistic personal anecdote about growing up in Saint-Louis and obtaining his first Mac computer at the age of eight, which he called a “division line” in his life.
“I can trace a straight line at the Openai Foundation,” he told the group of senators. “I am a child of the Internet revolution and I am proud to be one of the many parents of the AI revolution.” During a recent visit to Saint-Louis, he remembers going through his old house and looking up at the window on the top floor.
“The light was lit, and I thought, I hope there is a child there who stands late at night, playing with Chatgpt, determining how he or she will start the company that will come afterwards and whatever the next thing after the AI happens here too. It is, for me, the magic of this country.”
This story was initially presented on Fortune.com
