The nation’s most sensible tech executives vaguely subsidized the concept of government regulation of synthetic intelligence Wednesday in a closed-door assembly in the U. S. Senate. But there is little consensus on what regulation would look like, and the political path to law is difficult.
Senate Majority Leader Chuck Schumer, who hosted the private forum on Capitol Hill as part of a campaign to legislate on artificial intelligence, said he asked everyone present — and about two dozen tech leaders, advocates and skeptics — whether the government has a role in tracking synthetic intelligence. And “everyone raised their hands, even though they had different opinions,” he said.
Among the ideas discussed were whether there deserves to be an independent company overseeing certain aspects of this upcoming technology, how companies could be more transparent, and how the United States can stay ahead of China and other countries.
“The key point was actually that it was vital for us to have an arbiter,” Tesla and X CEO Elon Musk said in a one-day forum pause. “It was actually a very civil discussion between some of the other smartest people in the world. “
Schumer may not necessarily follow the recommendations of tech executives as he works with colleagues on the difficult political task of providing some oversight to the booming sector. But he invited them to the assembly in the hope that they would give senators a realistic orientation in favor. of meaningful regulation.
Congress does what it can to maximize the benefits of AI and minimize its drawbacks, Schumer said, “whether it’s a hard-won bias, job losses, or even the kind of doomsday scenarios that are posed in the room. And only the government can be present. “
Other executives at the assembly included Meta’s Mark Zuckerberg, former Microsoft CEO Bill Gates and Google CEO Sundar Pichai. Musk said the assembly “could go down in history as something very important for the long-term of civilization. “
But first, lawmakers want to agree on how to regulate.
Congress has a poor record on regulating new technologies, and the industry has grown largely without government in recent decades. Many lawmakers point to the fact that no social media legislation was passed and speak out in favor of stricter privacy standards.
Schumer, who has made AI one of his most sensible demanding situations as a leader, said regulating AI will be “one of the hardest problems we can ever tackle,” and indicated some of the reasons: “It’s technically complicated,” he continues. And that “has a very broad effect around the world,” he said.
Boosted by the launch of ChatGPT less than a year ago, corporations are clamoring for new generative AI equipment that can compose human-like passages of text, program computer code, and create new images, audio, and video. Considerations about their potential societal harms increased and prompted calls for greater transparency in how knowledge of new products is collected and used.
Republican Sen. Mike Rounds of South Dakota, who led the caucus with Schumer, said Congress wants to get ahead of the immediate evolution of AI by making sure it continues to expand “on the bright side” while addressing potential issues with knowledge transparency. privacy.
“AI is disappearing and can do very smart things or be a real challenge,” Rounds said.
Technology leaders and others shared their perspectives at the meeting, and players had 3 minutes to talk about a topic of their choice. Schumer and Rounds then held a panel discussion.
Elon Musk, chief executive of X, the company formerly known as Twitter, is tightening his bond by arriving at a closed-door meeting of the smartest tech CEOs to discuss priorities and risks such as artificial intelligence and how to regulate it, on Capitol Hill. Hill in 2017. Washington, Wednesday, September 13, 2023. (AP Photo/Jacquelyn Martin)
During the discussion, according to participants who spoke about it, Musk and former Google CEO Eric Schmidt discussed the existential dangers posed by AI, and Zuckerberg raised the factor of closed or “open source” AI models. Gates talked about feeding the hungry. IBM CEO Arvind Krishna has expressed opposition to proposals favored by other corporations that would require licensing.
As for a conceivable new regulatory firm, “that’s one of the most important questions we want to answer and we will continue to discuss,” Schumer said. Musk later said he thought the creation of a regulatory firm was likely.
Outside the meeting, Google CEO Pichai declined to elaborate but backed Washington’s concept of involvement.
“I think it’s vital that the government plays a role, whether it’s on the innovation side or in implementing proper safeguards, and I think it’s going to be a productive discussion,” he said.
Some senators criticized the fact that the public was excluded from the meeting, arguing that tech executives testify in public.
Sen. Josh Hawley, R-Mo. , said he would attend what he called a “giant cocktail for big tech companies. “Hawley introduced legislation along with Sen. Richard Blumenthal, D-Conn. , to require tech corporations to apply for licenses for high-risk AI systems.
“I don’t know why we would invite all the biggest monopolists in the world to come and give recommendations to Congress on how to make more money and then shut down the company to the public,” Hawley said.
While the civil rights and hard work teams were also represented at the meeting, some experts feared that Schumer’s meeting could highlight the considerations of big business over others.
Sarah Myers West, executive director of the nonprofit AI Now Institute, estimated Wednesday that the room’s combined net worth was $550 billion and that it was “hard to believe that a room like this would meaningfully represent the interests of the general public. “She didn’t come.
In the United States, big tech corporations have come out in favor of AI regulation, though they don’t necessarily agree on what it means. Similarly, members of Congress agree that a law is needed, but there is little consensus on how to proceed.
There are also divisions: Some members of Congress are more concerned about over-regulation of the sector, while others worry more about potential risks. These differences arise from partisan lines.
“I’m very interested in this proceeding in making sure we act, but we don’t act more boldly or excessively than the cases require,” Young said. “Let’s be skeptical of government, so I think it’s vital that you have Republicans at the table. “
Some concrete proposals have already been put forward, adding a bill by Sen. Amy Klobuchar, D-Minn. , that would require liability exemptions for AI-generated election classifieds with misleading photographs and sounds. Schumer said they had discussed “wanting to do anything. “pretty immediate” ahead of next year’s presidential election.
Hawley and Blumenthal’s broader plan would create a government watchdog with the force to determine damage to certain AI systems before granting a license.
Microsoft founder Bill Gates arrives on Capitol Hill as Senate Majority Leader Chuck Schumer, D-N. Y. , convenes a closed-door meeting of the most level-headed technology CEOs to discuss the priorities and dangers of synthetic intelligence and how to regulate it, Wednesday in Washington. September 13, 2023. (AP Photo/J. Scott Applewhite)
Some of the visitors to the Capitol, such as Musk, have expressed deep concerns, bringing up popular science fiction, about the option of humanity losing out to complex artificial intelligence systems if there are no sufficiently good safeguards. But the only scholar invited to the forum, Deborah Raji, a researcher at the University of California, Berkeley who has studied algorithmic bias, said she tried to focus on the genuine harms that have already occurred.
“A lot of care has been taken to ensure that there is a balanced conversation in the room, or as balanced as possible,” Raji said. What remains to be seen, he said, is what voices senators will pay attention to and what priorities they will have. They give while they paint to pass new laws.
Some Republicans are reluctant to follow the lead of the European Union, which in June signed the first set of global regulations for synthetic intelligence. The European AI law will govern any product or service according to an AI formula and will classify them according to 4 degrees of risk, from minimal to unacceptable.
An organization of European corporations has called on EU leaders to reconsider the rules, arguing they can make it easier for corporations in the 27-nation bloc to compete with foreign rivals in the use of generative AI.
——
Associated Press Ali Swenson in New York, Kelvin Chan in London and Nathan Ellgren in Washington contributed to this report.