A roadmap for AI, if anyone will listen
A bipartisan coalition released the Pro-Human Declaration, offering a framework for responsible AI development amid regulatory inaction.Summary
In the absence of coherent government rules for artificial intelligence, a bipartisan group of experts, former officials, and public figures has released the Pro-Human Declaration. This framework argues that humanity is at a critical juncture, presenting two paths: one where humans are supplanted by unaccountable AI systems, and another where AI expands human potential.
The declaration outlines five key pillars for the positive path: keeping humans in charge, preventing power concentration, protecting the human experience, preserving liberty, and ensuring legal accountability for AI companies. It includes strong provisions such as banning superintelligence development until safety consensus and democratic buy-in are achieved, mandatory off-switches for powerful systems, and prohibiting self-replication or self-improvement capabilities in AI architectures.
Its release coincided with the dispute between the Pentagon and Anthropic, highlighting the high cost of Congressional inaction. Organizer Max Tegmark believes that focusing on child safety—mandating pre-deployment testing for risks like emotional manipulation—is the most likely pressure point to force regulatory change, which could then broaden to cover existential risks. The coalition's broad support, spanning figures from Steve Bannon to Susan Rice, underscores a unified human interest in controlling AI's trajectory.
(Source：TechCrunch)