You know things are getting serious when the people building the world’s most powerful AI start reaching for the brakes.
Anthropic CEO Dario Amodei called for a global slowdown in AI development Saturday. In a 3,800-word essay, he warned that rapidly advancing AI could eventually “outrun our ability to understand and control these systems.” Amodei wants AI labs to slow the development of more powerful models so safety measures have time to catch up.
And he isn’t alone. Hours later, OpenAI CEO Sam Altman said he agreed that a slowdown was necessary. “This has been a primary topic of discussions we’ve had at OpenAI in recent weeks,” Altman wrote. The warnings come days after Anthropic researcher Jacob Coxon quit the AI industry over similar fears. Amodei is calling for independent safety evaluations and greater coordination among governments.
But not everyone is buying the warnings. Nvidia CEO Jensen Huang recently accused AI companies are fueling cybersecurity fears so they could develop products designed to address those threats. “What better way to create demand than to create a problem?” Huang said.
You know things are getting serious when the people building the world’s most powerful AI start reaching for the brakes.
Anthropic CEO Dario Amodei called for a global slowdown in AI development Saturday. In a 3,800-word essay, he warned that rapidly advancing AI could eventually “outrun our ability to understand and control these systems.” Amodei wants AI labs to slow the development of more powerful models so safety measures have time to catch up.
And he isn’t alone. Hours later, OpenAI CEO Sam Altman said he agreed that a slowdown was necessary. “This has been a primary topic of discussions we’ve had at OpenAI in recent weeks,” Altman wrote. The warnings come days after Anthropic researcher Jacob Coxon quit the AI industry over similar fears. Amodei is calling for independent safety evaluations and greater coordination among governments.
But not everyone is buying the warnings. Nvidia CEO Jensen Huang recently accused AI companies are fueling cybersecurity fears so they could develop products designed to address those threats. “What better way to create demand than to create a problem?” Huang said.