Former OpenAI Science Chief Wants Robots to Automate Drug Discovery

A former OpenAI executive is betting that robots can automate drug discovery. Kevin Weil, who once ran the artificial intelligence company's science unit, launched a new startup called Warp Sciences from stealth this week, backed by two hundred fifteen million dollars in fresh funding to train machines that make and test promising drug compounds around the clock.

The round was led by Lightspeed Venture Partners, with IVP, Mayfield, Lux Capital, Premji Invest and Menlo Ventures joining in, according to TechCrunch. Weil announced the launch and the financing on the social platform X, disclosing a valuation of eight hundred sixty-five million dollars for a company that has yet to ship a public product.

Warp Sciences is targeting one of the slowest stretches of drug development: the grinding laboratory work of synthesizing and screening thousands of variations of a molecule before a compound ever reaches a human trial. That stage can take years and cost hundreds of millions of dollars, according to Dealroom.

Instead of outsourcing that labor, the startup wants robots to do it in-house. Human chemists wearing cameras and sensor gloves will demonstrate laboratory techniques so the machines can learn to repeat them, eventually running experiments through the night without tiring.

Bringing the wet lab back home

For years, biotech companies have shipped the most labor-intensive chemistry work to contractors in China, including the giant WuXi AppTec. Weil wants to reshore that work, TechCrunch reported, insulating the drug development pipeline from geopolitical supply-chain shocks.

The pitch lands at a moment when drugmakers have learned how fragile far-flung supply chains can be. A domestic lab network built to automate drug discovery would let American startups iterate on compounds without waiting on overseas contractors, and without handing sensitive early-stage research to foreign firms.

The bet joins a crowded field. GenZ NewZ recently reported on a newly described form of cell death, baciptosis, that researchers say could boost cancer therapy — the kind of discovery that still needs years of compound testing before it helps patients. Automating that testing is exactly the bottleneck the new venture wants to break.

From chatbots to chemistry

Weil spent nearly two years at OpenAI, first as chief product officer overseeing the teams behind ChatGPT and the company's developer tools, then as head of a unit called OpenAI for Science. That group built a model trained on chemistry, protein engineering and genomics, which Weil called an attempt to build "the next great scientific instrument," according to Dealroom.

He left the company in April alongside Sora lead Bill Peebles, when OpenAI folded its science work into its core model, product and infrastructure teams. An OpenAI spokesperson described the move as consolidation ahead of a possible IPO rather than a verdict on the science bet, Dealroom reported.

Earlier reporting had suggested Weil was seeking around one hundred fifty million dollars at a valuation above seven hundred fifty million. The final round came in heavier on both counts, a sign that investor demand outran his original ask, Dealroom noted.

The bet and the skeptics

Skeptics will note the unusual shape of the deal: a very large round for a company emerging from stealth, with no public product and no disclosed path to validated drug candidates. Reporting on the raise named no product and no launch date, and the scientific focus beyond the plan to automate drug discovery remains undisclosed.

Still, the logic is straightforward. Drug discovery is slow because the physical work is slow. If machines can automate drug discovery at industrial scale, running experiments through the night and learning from every failed compound, promising medicines could reach human trials years sooner. For patients waiting on new treatments, that speed is the whole story.

For readers, the stakes are measured in medicine cabinets, not valuations. Faster, cheaper compound testing could mean new therapies arriving sooner, built on supply chains closer to home. The robotics bet also echoes beyond biotech: Tesla has begun producing its Optimus humanoid robot even as its profits fell, a sign that automation is having its moment across industries. And the immune-system puzzles these robots might one day help solve are getting clearer too — scientists recently found a hidden switch in killer immune cells that could reshape how infections are fought.