Unlike most congressional hearings involving tech industry leaders in recent years, a hearing involving the CEO of OpenAI (the company responsible for developing the AI chatbot ChatGPT and the image generation tool Dall-E 2) will be held this week. It was incontrovertible. During the three-hour hearing, Sam Altman greeted the audience amicably among the subcommittee members.
The CEO of OpenAI hinted to the U.S. lawmakers in attendance that there would be regulation. artificial intelligence It was essential. “If this technology doesn’t work, it can go wrong.” very It’s wrong,” Altman said during his first appearance in Congress on May 16.
Altman was the most recent person to explode silicon valley. But unlike other CEOs, from Facebook’s Mark Zuckerberg to TikTok’s Shou Zi Chew, OpenAI’s CEO was greeted with a much warmer and more serious demeanor.
Altman spoke positively about the possibilities and pitfalls of new technologies. But to the surprise of many techies, the senators in attendance seemed more than willing to accept his warning.
OpenAI’s CEO acknowledged that AI could “cause serious harm to the world,” adding a plea to some with the warning. regulatory guardrails for this new technology.
How did the testimony of the OpenAI CEO come about?
WASHINGTON, DC – MAY 16: OpenAI CEO Samuel Altman arrives to testify before the Senate Judiciary Subcommittee on Privacy, Technology, and Law on May 16, 2023 in Washington, DC. Greetings to the chair, Senator Richard Blumenthal (D-Connecticut). DC. The commission held oversight hearings to consider AI, with a focus on artificial intelligence rules. Win McNamee/Getty Images/AFP (Photo credit: WIN MCNAMEE / GETTY IMAGES NORTH AMERICA / Getty Images via AFP)
Altman attended a Senate Judiciary Subcommittee hearing, and a simple but difficult question was at the top of the agenda. “What is AI?” regulate technologyespecially when it comes to something as complex and fast-moving as AI, Congress needs to understand it first.
So getting the CEO of OpenAI, the Microsoft-backed startup behind ChatGPT, to provide some insight would be best for lawmakers. What’s more, it was the Senate’s first central hearing on AI. “As this technology advances, we understand that people have concerns about how it could change our lives. So do we,” said OpenAI. CEO said at a Senate hearing.
Other senators echoed South Carolina Republican Lindsay Graham’s likening of AI technology to nuclear reactors that need to be licensed and answered by regulators.
“I will form new agency It licenses any initiative that exceeds a certain scale of capability, and that license can be revoked to ensure compliance with safety standards,” Altman said. bloomberg Report; he added that such US authorities could form a global consensus on AI regulation.
In response, lawmakers in attendance said Congress was moving too slowly to keep up with the pace of innovation, and rulemaking in such a dynamic industry was best left to a new agency, especially when it came to AI. agreed with the opinion of
Connecticut Democratic Senator Richard Blumenthal, who chairs the Senate Judiciary Committee’s subcommittee on privacy, technology, and law, said AI companies test systems before releasing them and disclose known risks. said it should be obliged to do so. Blumenthal also expressed concern about future AI systems destabilizing the job market.
Altman largely agreed, but was more optimistic about the future of work. What is certain is that his CEO of OpenAI himself seemed haunted by his own biggest fears about the technology. Altman largely avoided details, admitting that the industry could cause “significant harm to the world” and “could collapse if this technology didn’t work out.” very error. “
But he then proposed that new regulators impose safeguards to stop AI models that can “self-replicate and self-exfiltrate into nature.” Altman even admitted that OpenAI is concerned about the impact of the technology on elections. “This is not social media. This is not. So we need a different response.”
When companies like OpenAI reach the stage of debating whether development should be halted, Generation AI Senators, like hearing witnesses, said it was unwise to pause American innovation. At the same time, competitors such as China are also pursuing AI innovation.
However, Altman clarified that OpenAI has no plans yet to push the next version of its critical language model-based tool. “We are not currently training to become GPT-5,” he said, adding that there are no plans to begin within the next six months.
