Google DeepMind's Demis Hassabis claims AI could help 'solve all disease,' but the company's new Gemini for Science tools are designed for researcher use, not immediate cures.

At Google I/O, DeepMind CEO Demis Hassabis stated that the company hopes to 'reimagine the drug discovery process with the goal of one day solving all disease,' referring to Gemini for Science, a collection of experimental AI tools designed to encourage researchers to explore and make new discoveries.

Gemini for Science builds on Google's AlphaFold (which helps researchers understand protein structures) and AlphaGenome (which predicts mutations in human DNA sequences). Researchers have used AlphaFold to help develop malaria vaccines, discover a key protein behind LDL cholesterol, and understand a protein behind early-onset Parkinson's disease, though AlphaGenome has not been validated for personal genome prediction and struggles to capture cell- and tissue-specific patterns.

The article warns that such breakthroughs are likely to take at least 20 years, probably more, and that AI is ultimately a tool requiring expert input and collaboration—it does not eliminate the need for FDA drug trials, animal testing, or other long-established regulatory processes.