Demis Hassabis cuts AGI timeline to 2‑3 years and touts AI's scientific potential
| Source: Techmeme | Original article
Demis Hassabis now expects artificial general intelligence within two to three years, cutting his earlier three‑to‑five‑year estimate, and says it could transform scientific research.
DeepMind chief Demis Hassabis has trimmed his timetable for artificial general intelligence, telling the Wall Street Journal that he now expects AGI to arrive within two to three years – a notable acceleration from the three‑to‑five‑year horizon he gave earlier in 2026. In the same interview he highlighted the technology’s potential to transform scientific discovery, suggesting that a truly general system could accelerate research across fields from drug design to climate modelling.
The revised forecast matters because DeepMind is one of the few organisations with the resources and talent to mount a credible AGI effort. A shorter horizon intensifies the competitive race among leading labs, raises the stakes for safety and alignment work, and could prompt regulators to revisit oversight frameworks that are still in early stages. It also signals to investors and partners that DeepMind believes its roadmap is on track, potentially influencing capital flows into the broader AI ecosystem.
What to watch next are the concrete steps DeepMind will take to bridge the gap between its current models and a general system. Industry observers will be looking for new research publications, large‑scale model roll‑outs, or strategic collaborations that hint at progress toward the timeline Hassabis outlined. Parallel signals – such as policy discussions on AI governance and the actions of rival labs that have recently disclosed massive compute spending – will help gauge whether the two‑to‑three‑year window is realistic or merely an optimistic benchmark.
Sources
Back to AIPULSEN