From Code to Cabinets: Lila Sciences Raises the Bar on AI-Powered Science

The startup world lives for big rounds but every once in a while you get one that feels like a signal: Lila Sciences, founded just two years ago, has raised an additional $115 million in an extension of its Series A round, pushing its valuation past $1.3 billion.
That’s no small feat. It means investors aren’t just betting on new apps, they’re backing the idea of AI not just on the screen, but in the lab. Reuters

What’s going on

• The funding extension brings Lila’s Series A total to around $350 million, and the company’s total capital raised to about $550 million.
• Notable backer: Nvidia’s venture arm enters the round, signalling alignment between top-tier AI hardware and next-gen science infrastructure.
• Lila is building what it calls “AI Science Factories”, essentially highly-automated labs where AI models propose experiments, robots carry them out, results are fed back into models, and the loop repeats.
• The company has leased a 235,500 sq-ft lab space in Cambridge, Massachusetts — one of the largest lab deals in the region this year — scaling physical infrastructure to match its digital ambition.
• Lila says its focus spans materials science, energy, semiconductors, drug discovery and more. Industries hungry for speed, automation, and novel molecules.

Why this matters

• For founders and tech builders: This is proof that hardware-adjacent, deep-tech bets are still attracting major funding, if you solve a “real thing” (not just another app), the money is there.
• For investors: The direction is shifting. Training models on existing web data? Maybe done. Building new data via experiments? That’s where some capital sees value.
• For startups in adjacent fields: If your business sits at the intersection of AI + robotics + scientific discovery (bio, materials, energy), this is a signal that you’re in the same league—as long as you show credibility.
• For the broader ecosystem: The scientific method itself might be evolving. Instead of separate silos (hypothesis → experiment → analysis) we’re seeing closed-loop systems where machines do more of the cycle. That could change timelines, costs and outcomes across many industries.

Key nuances & things to watch

• Despite the big claim-making, Lila hasn’t publicly released detailed data for all its discoveries. Some claims remain unverified in the public domain.
• Scaling physical labs is expensive and messy. Recruiting, compliance, hardware costs, robotics maintenance — the risk and overhead are real.
• Commercialisation matters: It’s one thing to automate experiments; it’s another to turn that into products, partnerships or revenues. Lila’s timeline and business model will be important.
• While the lab is in the U.S., true global scale depends on multiple geographies, regulatory contexts, talent, and ecosystems.
• Competitors (AI for science / biotech + automation startups) are emerging rapidly. Execution speed and moats (hardware, data, IP) will count.

Closing takeaway

Lila Sciences isn’t just raising money—it’s raising the possibility that science can be done differently: faster, more autonomously, more data-driven. For anyone building in tech, innovation or startup land, the lesson is: solve foundational problems, build novel infrastructure, and you might just attract the kind of capital that turns a startup into a science revolution.