The $450M Bet That Says AI's Next Frontier Isn't Software
When capital starts flowing into atoms instead of apps, it's worth paying attention.
Everyone wants to build the next foundation model. Fewer founders are asking a harder question: What if AI’s biggest impact isn’t generating text, but discovering entirely new materials?
That’s the bet investors just made on CuspAI.
Today, the Cambridge-based startup announced a $450 million Series B at a $2.6 billion valuation, led by Kleiner Perkins and NEA. The round included backing from the UK government, Bezos Expeditions, Lux Capital, AMD Ventures, and other major investors. At the same time, CuspAI launched its AI Materials Foundry, a consortium of more than 45 partners, including Nvidia and Meta, to accelerate the discovery of next generation materials for semiconductors, clean energy, and advanced manufacturing.
That’s a massive swing.
Most AI startups are competing to make knowledge work faster. CuspAI is trying to compress decades of chemistry into months by combining generative AI, simulation, synthesis planning, and experimental validation. If it works, the upside isn’t another productivity tool. It’s entirely new physical products.
The mental model that comes to mind is Second-Order Thinking.
First-order thinking says AI helps scientists work faster.
Second-order thinking asks what happens after scientists become dramatically more productive.
If the bottleneck shifts from discovery to manufacturing, supply chains, regulation, and commercialization suddenly matter a lot more than model quality. The companies that win won’t necessarily have the smartest algorithms. They’ll have the best ability to move discoveries from a GPU into the real world.
That’s also why this funding round stands out. Investors aren’t just writing a bigger check because AI is hot. They’re funding an ecosystem. Partnerships with hardware companies, research labs, and industrial players create compounding advantages that are much harder to replicate than another model release.
We’ve seen this pattern before in regulated industries. Building the technology is only part of the challenge. Building the infrastructure, compliance, and trust around it often creates the more durable business. At /mkt, that’s been true in bringing regulated athlete investment products to market. The technology matters, but so does everything around it.
Here’s the part I think many people will miss.
The headline isn’t the $450 million. It’s where the money is going.
For years, venture capital rewarded software that reduced friction. Increasingly, capital is flowing toward companies that can reshape the physical world using AI. That’s a much longer cycle, but it can also create much deeper moats.
So here’s my take.
My takeaway isn’t that every founder should pivot into materials science. Far from it.
It’s that the most interesting startups of the next decade may look less like software companies and more like vertically integrated businesses that happen to use AI exceptionally well. The model is important. The system built around it is what creates lasting value.
Nothing in this post should be interpreted as investment, legal, or financial advice. It’s commentary on a publicly reported funding announcement and broader startup trends.
If this was useful, share it with someone who builds things. And if you want the full toolkit of 50 mental models, you can grab my book, Mental Models: How to Think, Act, and Win, on Amazon right now.
If you want the mental models behind breakdowns like this, my book, Mental Models: How to Think, Act, and Win, is on Amazon now.
Informational and educational only. Not investment advice, and not an offer or solicitation of any security. Figures on volume, growth, and licenses are as reported by the company and haven't been independently verified; Cyclops is private and didn't disclose a valuation. /mkt is referenced solely as an operating example of building in regulated markets, not as an investment opportunity.
