[Wu Blockchain News] Vitalik: Slowing the Hyperscale Expansion of AI Compute May Do More Good Than Harm

Ethereum co-founder Vitalik Buterin pushed back against the backlash to the anti-data-center movement, arguing that algorithmic efficiency gains could deliver in roughly four years the AI capabilities that would otherwise require far larger hyperscale facilities within two. He said curbing compute expansion could buy society time to address cybersecurity and biosecurity risks, and noted that scientific progress depends on factors beyond AI, so a one-year AI slowdown need not cause an equal research delay. The intervention matters because it challenges the assumption that ever-larger data-center buildouts are the only path to capability gains, a premise underpinning much of the sector's capital spending. Watch for how AI labs and infrastructure investors respond on efficiency versus scale, and whether the debate shapes power and compute procurement plans.