(Bloomberg) — Google DeepMind has released a new version of AlphaFold, a landmark tool for predicting protein structures, that puts the artificial intelligence software on the path to making advances in biological research and bolstering a business that Google’s AI chief says it could be worth it. north of US$100 billion.
Bloomberg’s Most Read
The AI system has the potential to revolutionize medicine and create “tremendous commercial value” for DeepMind subsidiary Isomorphic Labs, said Demis Hassabis, CEO of both subsidiaries, in an interview with Bloomberg Television. “I hope to do both with Isomorphic: building a hundreds of billions of dollars business – I think it has that potential – as well as being incredibly beneficial to society and humanity.” Isomorphic Labs, a unit of Alphabet Inc. ago, was built to commercialize DeepMind’s AI for drug discovery. DeepMind first launched AlphaFold in 2018 with advances in decoding the shape of proteins, a scientific problem often compared to mapping the human genome. Now in its third iteration, AlphaFold can model a range of molecular structures, including DNA and RNA, and predict how they interact with each other.
“To really understand biology, you need to think about the interactions between different biological molecules,” Hassabis said. “And that’s what AlphaFold 3 represents a big step toward.”
Hassabis described the update, which was also published in the scientific journal Nature, as “essential for drug discovery” because of its critical insight into the types of chemical compounds used in designing and testing new medicines, including vaccines. “I would expect, perhaps in the next few years, the first AI-designed medicines in the clinic,” Hassabis added.
AI in medicine is a hot field. It is driven by the belief that software algorithms can find and help develop new medicines at drastically lower costs and much faster than the decades it might take now. The craze has attracted pharmaceutical companies, venture capitalists and technology giants like Nvidia Corp. for a market valued at around US$50 billion. Investors have invested more than $18 billion in AI-first biotech companies over the past decade.
But even with more money and computer advances, companies working on this technology have not yet demonstrated great clinical success. Some biotech companies specializing in AI drug discovery, such as BenevolentAI and Exscientia Plc, have struggled in the public markets.
In January, Isomorphic Labs announced its first two commercial partners, Eli Lilly & Co. and Novartis — alliances that the technology company said “could be worth nearly $3 billion” if they succeed in a series of performance milestones.
Alphabet has already tried to turn its healthcare innovations into new business opportunities, with units like Verily, a “precision health” platform for managing medical data. But these have shown limited commercial success so far.
Hassabis has previously said that Google will spend more than $100 billion to develop its AI arsenal. “The benefits of generative AI models, things like drug discovery, will far outweigh these costs in the long run,” he said in the interview this week. DeepMind is responsible for some of Google’s biggest advances in AI research. Since merging with Google’s other AI units last year, DeepMind now leads the company’s efforts on Gemini, its foundational AI model that competes with OpenAI.
To create AlphaFold 3, Hassabis said his researchers rebuilt the program “from the ground up,” using methods that take into account the shortcomings of previous versions, such as a deficit in training data. The new version relies on an AI technique used in generative AI called the diffusion model, which allows computers to do things like transform text into hyper-realistic images. In the Nature paper, DeepMind researchers write that their model predicts certain molecular interactions – proteins with nucleic acid and antibodies with antigens – with greater accuracy than existing computer models. “It’s much more generalizable,” Hassabis said of the new model. “And it’s much more powerful.”
Along with the update, DeepMind is launching AlphaFold Server, a tool that gives external access to most of the program’s functions. Hassabis said millions of researchers have used AlphaFold so far, although they “need to be a little technical” to deploy the program effectively. “The server is even simpler,” he said. “If you’re a biologist, you don’t need to understand the model or the technology.”
DeepMind will offer the tool to scientists for “non-commercial use”. Although Hassabis added that Isommorphic Lab’s pharmaceutical partners will have access to the Server and other more advanced tools.
–With assistance from Naomi Kresge.
Bloomberg Businessweek Most Read
©2024 Bloomberg LP
