DeepMind CEO: AI to Revolutionize Drug Discovery, Slashing Development Timelines from Years to Months

Cailian Press, September 12 (Editor Zhao Hao)Developing new drugs typically takes years with a very high failure rate, but this situation may soon see a reversal.

Demis Hassabis, CEO of DeepMind under Alphabet and winner of the 2024 Nobel Prize in Chemistry, recently stated that artificial intelligence (AI) technology will soon reduce the cycle of new drug development to less than one year.

Hassabis said in an interview: "In the coming years, we may see the R&D cycle shorten from several years to just a few months. I think that's possible, and it could even be faster."

Hassabis was born in London, first earning a double first-class degree in computer science from the University of Cambridge, then founding his own video game company, obtaining a PhD in cognitive neuroscience, and subsequently co-founding DeepMind, which was acquired by Google in 2014.

Currently, Hassabis is in charge of both Google DeepMind and Isomorphic Labs, an Alphabet subsidiary focused on drug discovery. Since its establishment in 2021, Isomorphic Labs has partnered with pharmaceutical giants Eli Lilly and Novartis.

In recent years, the pharmaceutical and biotechnology industries have been advocating for AI-powered new drug development, viewing this new technology as a means to accelerate the delivery of new therapies to patients, reduce development costs, and respond more quickly to medical crises.

There have been significant breakthroughs in computational power in this field, with algorithms capable of processing vast amounts of molecular data. However, to date, no drug designed by AI has completed clinical trials, meaning that no such drug has truly entered the human body.

In January this year, Hassabis said that the company would start clinical trials of AI-designed drugs by the end of the year. But so far, it has not been achieved.

Early September, Hassabis said at the London office that the company had demonstrated "the first set of validation results," proving that the drug candidates are promising to enter the trial stage, but did not provide a specific timeline.

Isomorphic Labs was established with the aim of commercializing DeepMind's AlphaFold system, which is capable of predicting protein behavior.

Hassabis stated that his research team is developing a "more advanced" new generation of AlphaFold models, which will not only be able to understand protein interactions but also comprehend more.

Hassabis previously stated that Isomorphic Labs has the potential to grow into a company valued at over $100 billion. Earlier this year, the company completed a $600 million funding round led by Thrive Capital.

This Alphabet subsidiary is currently focusing on the treatment of cancer and immune system diseases. Rebecca Paul, head of drug design, stated that these types of diseases offer a relatively more direct path for translating algorithmic models into clinical outcomes.

Paul said in another interview that AI-developed drugs would turn many cancers into controllable chronic diseases. She admitted, "It's hard to give an exact timeline at the moment, but we can already start thinking about how to solve this problem now."

Last year, Isomorphic Labs stated that it was collaborating with Novartis to develop new therapies for three targets. Hassabis mentioned that the current collaboration between the two parties has expanded to six targets.

Paul added that the cooperation between the two companies was "proceeding very smoothly," but did not disclose more details.