A proof of existence inspires and sustains the greatest ambitions. The Imitation Game, a test of whether a machine could exhibit human-like intelligence through conversation, gave us a target to chase in advanced machine intelligence. The sun has been producing clean energy at scale, through fusion, for billions of years. Nature makes energy and intelligence on its own, but not in the quantities we need. In chasing the discoveries embedded within these proofs, engineers, researchers, and entrepreneurs can make scarce things abundant.

One year ago, Matt Osman and Fabio Boniolo set out to build Polyphron, a human tissue foundry inspired by the proof of existence each of us carries. Human development shows that a single cell can give rise to a liver, kidney, and heart. Biology runs this process billions of times over, but does so once per person and never on demand. Polyphron is building the machinery to produce human tissue at scale.

The reason to pursue this mission has changed in the last few years. AI models trained on biological data can now generate novel hypotheses at a rate downstream methods can no longer keep up with. The bottleneck is no longer in generating interventions but in verification: an environment to test each candidate molecule against real human biology. The methods used today, shown in the figure below, each trade away something in time, cost, or fidelity. Polyphron’s bet is that engineered human tissue is the right abstraction to shrink the verification gap. It is complex enough to reflect actual biological function, and, with modern robotics and AI, tractable enough to manufacture and experiment on at scale. Tissue, produced abundantly, becomes the verification layer for biology.

Figure: The Current Approaches to Validating Biological Interventions

In the past year, the company has made tremendous progress towards its goal. Polyphron has automated the production of its micro tissue units within its facilities and has engineered state-of-the-art cortical and cardiac functional tissues across multiple donor lines. Initial experiments have confirmed the tissues respond like in vivo tissue, reproducing known drug effects across a panel of compounds. Earlier this month Vinh Tran, previously a researcher in post-training and self-improvement at Google DeepMind, joined Polyphron as an additional Co-Founder and Chief AI Scientist. Polyphron is now scaling its manufacturing facilities and continuing to improve the predictive power of its system.

We’re excited to announce our investment in Polyphron’s seed round with Quiet Capital and Haystack. The round brings the company to $20M in funding across its pre-seed and seed rounds with Compound, Humba, and Boost also participating alongside the founders of Chai Discovery, and Loyal. We encourage you to reach out to the team to learn more.