Google DeepMind's Gemini for Science aims to 'solve all diseases' — but not overnight
Demis Hassabis claims AI will one day cure everything, but the path is complex.
At Google I/O 2026, DeepMind CEO Demis Hassabis announced Gemini for Science, a suite of AI tools including AlphaFold and AlphaGenome. AlphaFold helps researchers understand protein structures—a step that could help unlock cancer treatments; separately, scientists recently found 1,700 new proteins that might do just that. AlphaGenome helps researchers predict mutations in human DNA sequences. These models dramatically reduce discovery timelines. However, Hassabis's bold 'solve all diseases' claim needs context: AI accelerates research but faces ethical, regulatory, and bias challenges before real-world cures materialize.
- DeepMind's AlphaFold has already helped discover 1,700 new proteins linked to cancer, slashing years off traditional protein analysis timelines.
- AlphaGenome predicts DNA sequence mutations, potentially explaining the root causes of genetic diseases, but a Nature study notes important limitations.
- Hassabis's 'solve all diseases' claim risks public misunderstanding; AI shifts discovery speed but cannot bypass ethical, regulatory, and equity hurdles.
Why It Matters
AI-powered drug discovery could reshape healthcare, but hype must be balanced with realistic expectations and ethical safeguards.