Q&A with CuspAI's Max Welling on its AI Materials Foundry, partnerships with Nvidia and others, Geoff Hinton and Yann LeCun joining its advisory board, and more
The co-founder of the UK-based science start-up explains how AI can help create new materials to address some of the world's most complex challengesLinkedIn:Max WellingLinkedIn:Max Welling:Thank you John Thornhill for the interesting conversation. — https://lnkd.in/...
Context & Ripple Effects
CuspAI has moved from a $30M seed round focused on AI-designed materials to a reported $100M Series A and then a $450M Series B tied to reducing rare-metal use in chipmaking. This interview adds operating detail around the company’s Materials Foundry and its ecosystem strategy.
The company had already attached Geoffrey Hinton to its advisory effort at seed stage; highlighting Hinton, Yann LeCun and Nvidia now reinforces the blend of scientific credibility, model development and compute partnerships behind its materials-discovery pitch.
First-order effects
- CuspAI gains a more visible platform to explain its AI Materials Foundry to prospective industrial users and partners, while its Nvidia relationship is positioned as part of the company’s route to building and running materials models.
- The public emphasis on Hinton and LeCun strengthens CuspAI’s scientific signaling alongside the capital and leadership expansion reported in its recent Series B financing.
Second-order effects
- Rival materials-AI companies face a higher bar to demonstrate both credible scientific expertise and access to the compute and partner ecosystem needed to turn models into usable discovery workflows.
- For potential chipmaking customers, CuspAI’s focus on cutting rare-metal use makes materials discovery more directly relevant to input-cost and supply-chain constraints, rather than a purely research-oriented AI application.
Third-order effects
- If AI materials platforms can repeatedly connect model outputs to industrial deployment, competition will shift from standalone discovery models toward integrated stacks combining data, compute, domain expertise and customer validation.
- The pattern suggests that AI infrastructure providers can become consequential partners in scientific-software markets, though the durability of that role depends on whether model-led discoveries translate into commercially adopted materials.
The trend: AI is moving from general-purpose model development into vertically integrated scientific discovery platforms that pair specialized models with compute partnerships and industrial use cases.