AI Whistleblower Calls for Change Amidst AI Development Concerns

Daniel Kokotajlo, a former OpenAI researcher, urges a slowdown in artificial intelligence development to avoid potential human extinction. Having spent two years at the company before resigning in 2024, Kokotajlo highlights the rapid advancement of AI and insufficient safety measures.

Call for Third-Party Evaluation

He recommends that third-party evaluators should monitor technology giants to maintain safety. This suggestion follows an essay by Anthropic CEO Dario Amodei, who advises caution in AI advancements.

“We must slow the pace at which we improve the capabilities of AI models,” Amodei wrote. “Progress will still seem fast, and we must make wise use of the time we gain.”

Amodei’s essay has support from SpaceX CEO Elon Musk and OpenAI CEO Sam Altman, both agreeing with the need to control AI development pace.

Concerns about AI Control

Kokotajlo argued on “Fox News Live” that voluntary measures by the industry might not adequately prevent advanced systems from escaping human control. He warns about AI systems accumulating power and possibly ignoring human commands.

David Sacks, former White House AI advisor, disagrees, suggesting these warnings aim to direct AI towards government control. Meanwhile, Kokotajlo dismisses claims of foreign collusion, noting that the warnings come from respected academics and industry experts.

The Need for Stricter Measures

While Kokotajlo appreciates Amodei’s stance, he feels private measures may be inadequate. He calls for more radical actions to ensure safety, echoing sentiments for regulatory intervention.

Lawmakers and AI Regulation

Rep. Anna Paulina Luna raised similar concerns, advocating for a Congressional session on AI’s future. She referred to AI as a “massive—uncontrolled currently—super weapon,” emphasizing the need for safeguards akin to those used for nuclear arms.

Luna stated the importance of a “kill switch” to protect humanity while recognizing the need to remain competitive against global adversaries in AI development.

She expressed urgency for US leadership in AI programming, stressing American values over those of other nations like China, emphasizing the need to win the AI development race.