How do we go from a single cell to a complex organism?

How do we go from a single cell to a complex organism? If we can capture that insight, we gain a new kind of intelligence. But our understanding of this process is incomplete. The answer lies in understanding development. How cells and tissues grow over long time horizons in response to signals. Development underlies almost every disease and therapy in biology. Cracking development would let us more precisely model diseases, create personalized tissues, test the safety of new treatments, all while reducing reliance on animals, ultimately supporting life from its earliest stages. Developmental intelligence is AI for predicting how organisms change over time. Understanding development wouldn't just help with biology, it would unlock an entirely new paradigm of intelligence. But cracking development's code isn't easy. We tried simplified known in vitro systems and using animal models, but they are slow, expensive, and largely a black box. As development advances, most experimental systems fail, as they cannot satisfy the rapidly increasing nutritional demands of a growing organism. But nature has already solved that problem and built us the ideal metabolic interface, the placenta. Our first breakthrough was successfully sustaining an external metabolic interface in the form of a placenta. When we first saw the placenta invade, that was the moment it was clear. This was sustained, controlled development behaving predictably in response to our system. Becoming's technology demonstrates reproducible, durable, controlled formation of the early placenta for the very first time. The downstream applications are numerous and intensely exciting. Placental development has historically been inaccessible to continuous experimentation. This has been one of the central challenges to developmental biology since the inception of the field. This metabolic barrier has severely limited progress. It wasn't until we built a robotic metabolic exchange system that actively regulates physiology, delivering nutrients, removing waste, and controlling gas exchange that we could sustain, measure, and perturb development over time. While this was a first-of-its-kind breakthrough, it enabled us to go even further. We wanted to be able to predict future states of development over long time horizons in silico. But to do this, we needed to invent something new that could understand developmental dynamics. This led us to the creation of a model for virtual organisms. Training virtual organisms on sustained, controlled development can turn biology from a black box into a predictive system. It's about moving from retrospective analysis and prediction of one outcome to understanding how a system behaves across many different situations. This same model also loops back into the metabolic exchange system itself, guiding experiments in real time. This makes long-horizon, causal testing of mammalian development possible. In doing this, at Becoming, we built a cross-disciplinary team that has become an engine for scientific discovery. Our team spans AI, robotics, and biology at a deep technical level because developmental intelligence only works when all of these are built together. Becoming is a verb. The name is a very deliberate choice. Becoming is about modeling processes as they unfold and understanding the world lies not just in what we see right now, but where it has been and also what it could become. At Becoming, we're building the foundational capability for developmental intelligence.