White House AI adviser David Sacks criticizes Anthropic's mixed messaging
David Sacks, the White House's top AI adviser, just criticized Anthropic for warning about big AI dangers while still pushing ahead with its own powerful tech.
He pointed out the irony, saying, "You compare it to nukes... threaten half of white-collar jobs... warn recursive self-improvement could end humanity... then race ahead anyway."
Anthropic urges pause on frontier projects
Anthropic's latest report highlights how its AI now writes most of its code (over 80%) and fixes bugs way faster than humans ever could.
But even with this progress, it admits stopping risky AI development is tough ("training runs are easier to hide than missile silos"), and it's calling for a pause on frontier projects.
Despite chasing a $900 billion valuation, Anthropic says it's committed to responsible innovation and working with global policymakers.