Anonymous9 days ago
Anthropic co-founder Jack Clark and policy researcher Marina Favaro have called for a globally coordinated mechanism to slow or pause frontier AI development, warning that recursive self-improvement — where AI systems design their own successors — could arrive within two years. The proposal arrives amid sharp contradictions: Anthropic dropped its own hard safety commitments just months earlier, is heading toward a near-trillion-dollar IPO, and faces accusations of regulatory capture from both venture capitalists and open-source advocates.