Demis Hassabis predicts general artificial intelligence within two years and accelerates AI-driven drug discovery

Demis Hassabis, co-founder of Google DeepMind and winner of the 2024 Nobel Prize in Chemistry, has revised his forecast for the arrival of artificial intelligence with human-like cognitive abilities from three to five years to just two to three years, according to a Wall Street Journal profile. Currently serving as chief scientist at Alphabet and chairman of DeepMind, he also leads Isomorphic Labs, a company focused on using AI to discover new medicines.

Demis Hassabis and the AGI prediction

This prediction refers to general artificial intelligence, or AGI, a system capable of learning and solving problems across diverse fields, transferring knowledge between tasks rather than being limited to a single function. To illustrate this progress, Hassabis posed a challenge: could an AI replicate Einstein’s 1905 discoveries using only information available up to 1901? This question distinguishes between reproducing known answers and generating original knowledge.

From scientific research to medicine: AI’s role

DeepMind’s journey demonstrates how learning techniques can be applied to science. In 2016, AlphaGo defeated Go champion Lee Sedol. In 2020, AlphaFold 2 achieved a breakthrough in predicting the three-dimensional structure of proteins, and now has predicted over 200 million structures available to researchers. This work earned Hassabis, John Jumper, and David Baker the 2024 Nobel Prize in Chemistry.

As AI systems grow more powerful, safety assessments are becoming a central focus in discussions about the risks associated with increasingly capable AI.

Isomorphic Labs and the billion-dollar investment in drug discovery

At Isomorphic Labs, scientific ambition has taken a corporate form. In May 2026, the company announced a $2.1 billion funding round led by Thrive Capital, with participation from Alphabet and GV. The funds will be used to expand the drug development platform and accelerate the progression of candidates to clinical trials.

The company uses the IsoDDE suite of models, which goes beyond predicting protein structures. The technology also estimates how molecules interact with their targets and helps researchers select promising compounds. The strategy integrates AI models, data generation, software, and physical infrastructure to cover all stages required to bring potential treatments to clinical evaluation.

Prospects and challenges

If Hassabis’s prediction proves accurate, AI could not only accelerate scientific discoveries but also transform the pharmaceutical industry by reducing time and costs in the search for new drugs. At the same time, the scientific and business communities are discussing how to ensure that progress is accompanied by robust safety and governance mechanisms.