Nobel Laureate Hassabis Predicts AGI Within Five Years, Hinges on Key Breakthroughs

Описание: Demis Hassabis, CEO of Google DeepMind and Nobel laureate, has recently forecasted the arrival of Artificial General Intelligence (AGI). He suggested that AGI might be only one or two key technical breakthroughs away and offered an optimistic timeline, stating we could witness AGI within the next five years.

Hassabis pointed out that current large language models like ChatGPT, despite their capabilities, suffer from "jagged intelligence." They are essentially sophisticated "statistical prediction machines" that lack a genuine understanding of the physical world and capabilities for long-term planning. Merely scaling up data and computing power is insufficient for achieving AGI.

He identified two crucial missing pieces for AGI: first, "World Models" that would enable AI to truly comprehend and simulate the rules of the physical world; second, "Agentic Systems" that would grant AI the ability to proactively plan, execute complex tasks, and correct course in real-time. DeepMind is actively conducting research in these areas.

Hassabis emphasized that the core value of AGI lies in revolutionizing science. AI will become the ultimate tool for scientific discovery, dramatically accelerating breakthroughs in areas like drug development, material science, and clean energy. The transformation it brings could be ten times the scale and speed of the Industrial Revolution.

Regarding the global AI landscape, Hassabis noted that the capability gap between Chinese AI models and their Western counterparts has narrowed to "months," acknowledging China's strong engineering prowess. However, he raised a deeper question: whether China can transition from being excellent "engineers" adept at replication and optimization to becoming genuine originators of frontier innovation, which will be a key determinant in future competition.
https://x.com/Ric_RTP/status/20125232...