DeepMind Ready to Reimagine Drug Discovery and Defeat Diseases

At Google I/O 2026, DeepMind CEO Demis Hassabis, with a completely serious expression, announced: the company hopes to "reimagine the drug discovery process with the goal of solving all diseases in a single day." This is an ambitious and, frankly, unrealistic-sounding goal. But before dismissing it, let's understand what's behind it and why Verge journalists are already catching Google at its word.

What Google DeepMind Showed at I/O

Google DeepMind presented a comprehensive package of tools to accelerate biological research. The three main ones: Gemini for Science (a specialized version of the flagship LLM for scientific tasks), AlphaFold 3 (a protein structure prediction system), and the brand new AlphaGenome.

AlphaFold 3 can predict the three-dimensional structure of proteins and their interactions with other molecules with high precision. This is critical because protein shape often determines whether a drug will be effective. AlphaGenome extends this approach: the system can now work with genomic data and predict genetic factors of diseases. Gemini for Science integrates both systems into a single platform. It's an LLM that "understands" the context of biological research, can read scientific papers, and provide recommendations directly within an interface familiar to scientists.

How It Works in Practice

Here's a typical scenario: a researcher uploads a question into Gemini for Science like "develop an antiviral compound for virus X." The system analyzes known facts about the virus, proposes candidate molecular structures, uses AlphaFold 3 to predict the interaction of these molecules with viral proteins, and forecasts probable side effects. On paper, this sounds like a productivity leap. Previously, such a search would take months. Now it might take hours or days. But there's a catch: Google talks about "acceleration" but doesn't name specific timelines. How much faster, exactly? At what cost? This remains off-screen.

Why There Are More Skeptics Than Believers

Here's a list of reasons why "solving all diseases" is, to put it mildly, a very bold claim:

• From molecule to medicine—it's not a single task. Even if AI perfectly predicts the molecule, it needs to be synthesized in a laboratory, tested on cell cultures, then on animals, then on humans (clinical trials—3 phases, 5-10 years), and only then get regulatory approval.

• "All diseases"—it's not 100 diseases, it's 10,000+. Some have a genetic nature (within AI's scope), but others are infectious, autoimmune, mental, and environmental. AI in molecular biology won't help much with those.

• AI's history is full of unfulfilled promises. Deep Blue promised a revolution in analytics. IBM Watson promised a revolution in medicine. Reality is usually more complex than marketing slides.

• Regulators deliberately slow the process. The FDA requires safety proof—it's not their whim, it's patient protection. An AI-proposed solution is only the first step in a very long journey.

"There's an enormous distance between grand promises and results in real clinics," write

Verge editors, and they're not alone in this view.

What It Really Means

Google DeepMind's tools are indeed useful and can accelerate specific phases of drug development—especially at the candidate discovery stage. But "solving all diseases" is not a technological task. It's a question of how complex biology is, how strict regulatory requirements are, and how much money to invest in each direction. AI will be a good assistant to scientists. But it won't be a magician. Hassabis's promises are, likely, just a speech for Google I/O. A useful speech, but it's worth staying realistic.