Demis Hassabis, co-founder and CEO of Google DeepMind and a 2024 Nobel laureate in chemistry, predicts that artificial general intelligence (AGI) will ARRive by 2030. Speaking on Sequoia Capital's video series, Hassabis urged business leaders to treat this Timeline as a concrete business reality rather than a distant possibility.
A Firm Deadline for Business Strategy
Hassabis views the 2030 target as a strict schedule that demands immediate ACTion. He advises companies to pursue aggressive Investment and hiring now to remAIn competitive. Despite public skepticism, he remains committed to DeepMind's original 20-year roadmap. "Our original mission statement at DeepMind was step one: solve Intelligence, build AGI. Step two: use it to solve everything else," he said.
Combining Overlooked Technologies
DeepMind's breakthroughs came from blending undervalued techniques—notably deep learning and reinforcement learning—with the growing power of graphics processors. Hassabis cautions against designing software for hardware that does not yet exist, advising innovators to stay "five years ahead of your time, not 50 years ahead."
From Games to Real-World Science
DeepMind used complex strategy games to prove its systems worked before tackling scientific challenges. The landmark victory over the board game Go marked a turning point, dEMOnstrating that the AI was ready for messy, real-world data. The company launched its dedicated science division the day after returning from the historic match in Seoul.
Transforming Drug Discovery
Hassabis envisions a future where nearly all drug discovery exploration—"99% of the work and the time"—hAPPens inside a computer, with physical lab work reserved only for final validation. He believes this approach could compress drug development timelines from an aveRAGe of ten years down to weeks or even days, bringing personalized medicine within reach for all diseases.
Simulating Economies and Societies
Beyond biology, Hassabis sees AI-powered simulations as a way to safely test economic and social policies before implementing them in the real world. Instead of experimenting with interest rates on live populations, governments could run rigorous simulated trials to make better-Informed decisions in uncertain environments.
A Practical Approach to Consciousness
While Hassabis treats information as a fundamental building block, he prefers building practical tools over philosophical debates about machine consciousness. He recommends developing highly capable AI systems strictly as utility tools first, leaving deeper questions about agency for later exploration.
Balanced perspective
While Hassabis presents AGI by 2030 as a practical planning horizon, independent forecasters remain cautious. A 2025 review by 80,000 Hours noted that while pre-2030 AGI falls within the range of serious expert opinion, no current forecasts are highly reliable. On drug discovery, Reuters reported in April 2026 that Johnson & Johnson expects AI to roughly halve lead-generation time—significant acceleration, but still operating within existing CLInical trial and regulatory Frameworks rather than replacing them. Similarly, the IMF's 2026 AI scenario-planning work treats the Technology as a tool for exploring uncertainty, not yet as a precise policy simulator. A 2025 Reuters report on an AI-simulated Federal Reserve meeting showed how political pressure can distort simulated decision-making, highlighting the gap between simulation and real-world governance.
Comments & Questions (0)
No comments yet
Be the first to comment!