The best catalyst for the electrochemical reaction required to produce green hydrogen is a rare metal called iridium. As one of the scarcest elements on earth, it’s considered more valuable than gold, silver, or platinum — so finding a cheaper substitute could dramatically impact prices of the fuel.
That metal, said Lila Sciences senior VP of physical sciences John Gregoire, is, “a good exemplar problem” for his company, which is building a scientific superintelligence platform with an autonomous lab.
“There’s only one critical material to be discovered — an oxygen evolution catalyst that doesn’t contain iridium or comparably scarce and expensive materials,” Gregoire told Latitude Media. “That’s a tractable problem with an immediate pathway to have an impact on the world, and where each discovery is a sizable advancement of the field.”
Lila Sciences launched in 2023 with the goal to combine artificial intelligence, robotics, and human research to increase the speed and scale of climate breakthroughs. At the start, the team understood how much depended on tangibly demonstrating the potential of the autonomous lab and machine learning algorithms; if the company managed to make even one or two discoveries in its crucial early years, it would be clear that they were onto something, Gregoire said.
And so they did. Within a couple of months of operating their autonomous lab, the scientists at Lila made a non-iridium catalyst discovery. Last week, Lila announced its official launch, armed with $200 million in seed funding from investors including Flagship Pioneering and General Catalyst.
Molly Gibson, who’s co-founder and president at Lila, as well as origination partner at Flagship Pioneering, explains that the idea behind Lila was to do “closed loop science.”
“We’re exploring… being able to connect the lab to computation, and having that tight coupling between an AI making a hypothesis and going into the lab and running the experiment, and then learning from that experiment,” she said. “We can run the wheel of science autonomously. And with each turn of the wheel of science, we increase intelligence.”
The appeal to investors, according to Gibson, lies in the approach to research and discovery, rather than any single discovery in itself. For its non-iridium catalyst discovery, for example, “it remains to be seen whether or not the greatest value of that discovery is the intellectual property of the catalyst itself or the intelligence that the model gained by discovering that catalyst,” Gregoire noted. “Because that intelligence can be leveraged for a bunch of other problems that we’re moving into, and the echoes of that intelligence will ring for a long time.”
Lila Sciences also announced the discovery of new carbon capture sorbent materials, which it claims have “better capacity, thermal stability, and kinetic binding than leading products.” But the approach is meant to be applicable across scientific domains.
Flagship Pioneering, the “venture creation firm” out of which Lila was founded, and its largest investor to date, has historically focused mostly on the biotech space; Moderna Therapeutics, the biotechnology company known for developing one of the first COVID-19 vaccines, is one of its products. Gibson herself has a background in biotechnology, having previously founded a company called Generate: Biomedicines, which uses generative AI for protein therapies.Lila announced the discovery of antibodies and peptides to be used for various therapeutic targets alongside its climate- and energy-related materials findings.
“The vision only keeps getting broader in the last year,” said Gregoire. “We haven’t scaled down anything.”
That said, the company is still working to pin down its revenue model.It will likely rely on partnerships, at least in the short term, Gibson said, with “anyone who wants to solve a hard scientific challenge.”
“To us, one of the most important realizations is that, as intelligence increases the value of and the number of things that we can do, the number of business models that open up is really kind of unknowable today,” she said.
