Episode Summary
Executive Summary: Jonathan Godwin explains how Orbital Materials uses AI-first R&D to accelerate advanced materials discovery, then vertically integrates into physical products for data centers. The company combines AI simulation, generative chemistry, and lab validation to develop sustainable materials—starting with a dual-use chiller/capture system—aiming to cut time, cost, and risk in industrial innovation.
Main Topics: Jonathan Godwin’s AI-to-materials founder journey (Priority: 5/5): Godwin traces his path from AI research at DeepMind to founding Orbital Materials after seeing that AI could do more than optimize software—it could drive real-world scientific discovery. How advanced materials were historically discovered (Priority: 5/5): The discussion contrasts traditional materials discovery—human intuition, trial-and-error, and sometimes chance—with Orbital’s design-first, AI-assisted approach. Orbital’s AI stack for materials discovery (Priority: 5/5): Orbital combines an AI-accelerated quantum simulation model (Orb), literature search, lab data, and generative chemistry agents to propose and validate hypotheses before physical experiments. Managing AI risk and hallucinations in physical R&D (Priority: 4/5): Because material design has real-world consequences, Orbital uses simulation, multiple AI systems, and human review to reduce hallucination risk and improve confidence before lab work. Why Orbital verticalized into products (Priority: 5/5): Instead of being only a discovery platform, Orbital chose to build and commercialize physical products because the founders wanted tangible impact and saw a rare market opportunity in data centers. First product strategy: sustainability for data centers (Priority: 5/5): Orbital’s entry point is a dual-use chiller aimed at hyperscalers, using waste heat for carbon capture or water capture, aligning with sustainability goals and data center needs. Capital-light commercialization model (Priority: 4/5): Orbital plans to fabricate pilots in-house, then secure purchase orders and partner with scaled manufacturers, avoiding the need to build its own plant.
Key Arguments: AI can dramatically compress materials R&D cycles by reducing brute-force experimentation and increasing the probability of successful experiments. The biggest shift is not just better software; it is creating AI-first industrial companies that can outperform incumbents on speed, cost, and quality. Advanced materials innovation has historically depended on universities, government funding, and corporate labs, but AI enables startups to participate more directly. Orbital’s open-source model helps the research community while keeping commercialization focus on internal product development. Data centers are a uniquely attractive market because hyperscalers have strong sustainability mandates and are already spending heavily on cooling infrastructure. A dual-use chiller that also enables carbon capture or water capture creates a compelling value proposition by leveraging waste heat that would otherwise be lost. Orbital aims to be asset-light by designing and piloting systems itself, then using those proven pilots to unlock manufacturing partners. Vertical integration is justified because materials companies can create generational value when they control both discovery and productization.
Data Points: Company age: 2.5 years - Godwin says Orbital Materials has been operating for about two and a half years. DeepMind tenure: 5 years - He spent about five years at DeepMind before founding Orbital. Series A timing: End of 2022 - Orbital raised its Series A round at the end of 2022 with Radical Ventures. Seed round timing: About a year before Series A - Godwin says the seed round was raised roughly a year before the Series A. Open-source release: Earlier this year - Orbital open sourced Orb, its AI model for simulating advanced materials, earlier in the year referenced in the interview. Purchase order / scale-up approach: First system in a shipping container - Orbital’s first-of-a-kind system is described as shipping-container-scale, designed to secure a purchase order for scaling. Target launch: End of this year - Godwin says the first shipping container is being sent to a data center toward the end of the year. Hyperscaler market share: Over 70% - He says hyperscalers account for over 70% of the data center market and are highly sustainability-focused. Product expansion outlook: 6–7 products - Godwin estimates Orbital may have around six or seven products over time, each potentially a large business. Long-term product count: Dozens - He suggests that in about 10 years the company could have dozens of products in the market.
Pivotal Quotes: "We want to flip that on its head. We're starting with design and using these new tools that give us an insight that just hasn't been possible before." — Jonathan Godwin: On how Orbital is reversing the traditional trial-and-error model of materials discovery. "AI is going to accelerate all of that. And we're just here, and there's one thing here, and that's incredible, but the ambition and the vision is so much wider." — Jonathan Godwin: On AI’s role across discovery, formulation, and manufacturing, not just simulation. "We need more businesses that are putting stuff out in the physical world and perhaps a few fewer software companies." — Jonathan Godwin: On why Orbital chose to build physical products rather than stay only a software platform.
Implications: AI-first industrial companies may shorten R&D cycles, lower capital needs, and create new startups in materials and climate tech. Orbital’s model suggests future winners may combine AI, product design, and manufacturing partnerships rather than software alone.