Demis Hassabis's AGI prediction: ~50% by 2030
Where Hassabis sits
| Forecaster | AGI timeline |
|---|---|
| Elon Musk (xAI) | By end of 2026 |
| Leopold Aschenbrenner | 2027 |
| Demis Hassabis (DeepMind) | ~50% by 2030 |
| Metaculus community | 50% by 2033 |
| Academic survey (n=2,778) | 50% by 2040 |
Why his view carries weight
Hassabis leads one of the two or three labs actually building frontier systems, so his estimate reflects daily contact with what current models can and can't do. That vantage tends to price in the friction between benchmark capability and genuine autonomy — which is why he lands more cautious than single-author extrapolations like Aschenbrenner's 2027.
How it's holding up
As of mid-2026 the evidence is consistent with his window: capability is strong (~83% on GDPval, ~80% on SWE-Bench Pro) but autonomous AI research is undemonstrated. His 2030 horizon has years left to run, whereas the nearer calls — Aschenbrenner's 2027 and Musk's 2026 — face reality first.
Frequently asked questions
The DeepMind CEO puts AGI at roughly 50% by 2030 — more cautious than Aschenbrenner (2027) and Musk (2026), more aggressive than the academic-survey median (2040).
Yes, but carefully. He expects transformative AI within the decade (~50% by 2030) while repeatedly warning against over-hyping near-term timelines.
He builds frontier systems daily, so his caution on the timing of full autonomy looks defensible as of mid-2026. Aschenbrenner has been closer on compute and capex trends. Both expect transformative AI this decade.
The live scorecard updates as models ship and verdicts change.
View the live scorecard →Get the weekly AGI progress briefing
Verdict changes, lab milestones, and what they mean for the 2027 clock. Free — no hype, just signal.
Subscribe free →