Stay informed with free updates
Just enter the Artificial intelligence myFT Digest — delivered straight to your inbox.
Isomorphic Labs, the four-year-old drug discovery firm owned by Google parent Alphabet, will have an artificial intelligence drug in trials by the end of this year, said its founder Sir Demis Hassabis.
“We are looking at oncology, cardiovascular, neurodegeneration, all major areas of the disease, and I think that by the end of this year, we will have our first drug,” he said in an interview with the Financial Times in the World Economic. Forum.
“It usually takes an average of five to 10 years (to discover) one drug. And maybe we could speed it up 10 times, which would be a dramatic change in human health,” he said. Hassabiswho received the Nobel Prize in chemistry with his colleague John Jumper and biochemist David Baker in October.
It is isomorphic was spun off from Google's AI research arm Google DeepMind in 2021, but remains a subsidiary of its parent, Alphabet. The startup's potential has attracted major pharmaceutical partners, who are interested in reducing costs and increasing the efficiency of the high-cost drug development process.
Hassabis previously told the FT his team was working on six drug development programs with Eli Lilly and Novartis.
In a wide-ranging interview, Hassabis, who is also the chief executive of Google DeepMind, said that the prototype of the search giant's AI assistant, known as Project Astra, will probably be announced to customers later this year. He described the near future, in three years, when there will be “billions” of AI agents, “negotiating with each other on behalf of the seller and the buyer” and said that it will require a rethinking of the web itself.
He also called for greater prudence and coordination among leaders AI developers competing to build universal artificial intelligence. He warned that this technology could threaten human civilization if left unchecked or re-used by “bad actors . . . for dangerous purposes”.
The main goal of Google DeepMind is to create a general artificial intelligence, or “a system capable of demonstrating all the cognitive abilities that people have”, according to Hassabis, who said that despite the media “hype” about its nearness, real AGI is still five to 10 years away.
“If something is possible and important to do, people will do it,” Hassabis said. “We're past that point now with AI, the genie can't be put back in the bottle. . . so we have to try and make sure we look at this on the ground as safely as possible. “