Hoffman rejects AI apocalypse talk

The LinkedIn cofounder dismissed concerns about AI causing human extinction as remote and controllable, comparing doomers to Cassandra from Greek mythology.

“Look at me, I'm Cassandra, I'm saying it's the end of time”

Frontier AI research has real benefits

Hoffman argues that continuing advanced AI development helps address biosecurity and cybersecurity risks while creating economic prosperity.

“I don't think that slowing down is important, because when you get to the defense against biorisk, the defense against cyber risk, and creating economic prosperity is a reason, I think, to keep going”

Hypocrisy in calling for slowdown

Hoffman questions why Anthropic and OpenAI leaders now want to slow frontier AI research when both companies are leading the world.

“You're in the lead, you can slow down anytime you like”

Recursive self-improvement needs management

Hoffman stresses that frontier labs must properly manage recursive self-improvement, where AI learns to build itself and its successors.

Amodei and Altman take right steps

Hoffman endorsed Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman's calls to pace development and improve AI alignment.

Zuckerberg prefers individual responsibility

Meta CEO Mark Zuckerberg rejected industry-wide pacts, saying each lab should move at its own safe pace. Meta delayed its Muse AI agent for safety improvements.

“Every lab has the responsibility and incentive to move at the pace required to train its models safely”