Hassabis Advances AGI Timeline to 'Within 2-3 Years'

Demis Hassabis, co-founder of Google DeepMind, has revised his outlook on the realization timeline of AGI (Artificial General Intelligence) forward again. He has shortened his previous prediction of 'within 3-5 years' to 'within 2-3 years,' highlighting how recognition of achieving AGI has undergone rapid changes in recent months.

AGI refers to a general-purpose artificial intelligence capable of performing all intellectual tasks comparable to humans, unlike current AI specialized in specific tasks. While long treated as a science fiction concept, in recent years leaders of major AI research institutions have increasingly cited specific timelines, and the discussion across the industry has become more tangible.

Hassabis is a Nobel Prize laureate and currently serves as CEO of Google DeepMind. He provided an outlook of 'within 3-5 years' in the early stages of 2025, but has now accelerated this to 'within 2-3 years.' The update came directly from Hassabis himself and has been reported as a statement made in public.

The background to this revision draws attention because of Hassabis's position. DeepMind is globally recognized for its excellence in fundamental AI research, and the statements of its founder tend to be received not merely as personal opinion, but as the judgment of someone at the forefront of research. Additionally, the repeated revision of predictions in a short timeframe suggests the possibility that AI development progress is advancing at speeds exceeding the expectations of those involved.

However, it is important to note that the definition of AGI varies significantly among researchers and institutions, and the criterion of 'what constitutes AGI' lacks clarity. Even if an organization declares 'AGI achievement,' it may be considered insufficient by other standards. This ambiguity in definition also becomes a factor that further complicates discussions about the timeline for reaching AGI.

From the perspective of the gap between the speed of technological development and social preparedness, this statement carries significant importance. Should AGI be realized within 2-3 years, the question becomes whether regulatory frameworks and social consensus formation can keep pace with that timeline. For policymakers and researchers, such statements can serve as reference points and catalysts for accelerating discussions on response strategies.

A key point to watch going forward is what basis Hassabis has for presenting this outlook. The weight of his statement will vary significantly depending on whether specific milestones or technical evidence are clarified. It remains necessary to continue tracking what research achievements DeepMind and other major institutions are accumulating toward AGI.