All Watched Over

Computer scientist Boaz Barak argues we're repeating history with artificial intelligence. In an essay on LessWrong, he draws a parallel to the nineteen-seventies, when activists and engineers fought against IBM's centralized computing monopoly by pushing toward personal computers and decentralization. Today, AI systems are large, expensive, and getting larger—concentrated in a handful of data centers and companies—while many people who fear the technology are refusing to engage with it entirely. Barak worries this could leave the future of AI governance to a few powerful entities or the AIs themselves. But he argues this outcome isn't inevitable: we can design safeguards, prevent monopolies on intelligence, and preserve human autonomy without sacrificing safety. His point: the distribution of AI power should be built into the foundation of how we develop and deploy the technology, not left to chance.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton