Story

On air. The free wire runs 15 minutes behind live. See pricing

World, crypto, finance and tech news, scored for impact.

CryptoTokenPostKorean, translatedFiled 35 h ago

DeepMind presents four paths from AGI to ASI

Original headline딥마인드, AGI에서 ASI로 가는 4가지 경로 제시

Open the sourceReported by TokenPost, tokenpost.kr, 20 November 2026, by 김하린

Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.

Google DeepMind researchers outlined four technical pathways from artificial general intelligence to artificial superintelligence—model and data scaling, paradigm shifts, agent collaboration, and recursive self-improvement—in a technical report titled "From AGI to ASI," detailed in a post by research scientist Tim Genewein on the DeepMind Institute platform. The report defines ASI as a system that consistently surpasses large groups of human experts across a broad range of tasks tied to human interests, and DeepMind noted the institute's views should not be treated as Google's official position. The framework matters because it gives the industry a shared vocabulary for how capabilities could compound beyond human level, with recursive self-improvement flagged as the most consequential and least predictable lever. What to watch next is whether scaling alone continues to deliver gains or whether paradigm shifts and agent collaboration become the binding constraints on progress toward ASI.

Muhabbit's AI wrote this summary in its own words from TokenPost's report and translated it from Korean. It can get details wrong, so check the source.

Named in this story

Google DeepMind, ASI, AGI, Tim Genewein, DeepMind Institute