A $2B company with no product, no revenue—just a goal: predict biology
before it evolves. The next frontier isn’t editing life, it’s forecasting it.
By Futurist Thomas Frey
In September 2025, two SEC filings showed up quietly in a database that tracks new company formations. A Delaware corporation called Astromech had raised $30 million. No press release. No announcement. No explanation of what it was building or why.
By March 2026, a second filing showed another $10.5 million had come in. Total funding: $40.5 million. Valuation: $2 billion. Still no revenue. Still almost no public information about what the company actually does.
The founders, it turned out, were Ben Lamm and George Church — the same two people who built Colossal Biosciences into a $10 billion de-extinction company. And when Lamm finally described what Astromech is trying to do, the ambition was staggering even by his standards.
He wants to build a machine that can predict how biology will change — before it changes.
What Astromech Is Building
Think about what a weather forecast actually does. It takes data about current conditions — temperature, pressure, humidity, wind patterns — feeds it through models built on decades of atmospheric science, and produces a prediction about what the atmosphere will do next. The forecast isn’t perfect. But it’s good enough to be genuinely useful. Good enough that we’ve built entire industries around it.
Astromech is trying to do something similar for biology.
The platform, as Lamm has described it, combines two capabilities. The first is deep learning algorithms that identify patterns across biological systems and species — patterns in how genes are expressed, how diseases spread, how organisms respond to environmental change, how vulnerabilities develop over time. The second is something called Bayesian ancestral reconstruction, which is a mathematical method for working backward through evolutionary history to model how a biological system got to where it is — and then forward, to project where it’s likely to go next.
Put those two things together and you get what Lamm calls a unified biological intelligence architecture. A system that doesn’t just describe biology as it is today, but predicts where it’s headed.
If it works, the applications are almost too broad to list. Disease risk. Pandemic early warning. Drug resistance forecasting. Agricultural vulnerability assessment. Conservation biology planning. Wildlife health monitoring. Livestock resilience modeling. The question of which specific pathogens are most likely to cause problems five years from now. The question of which ecosystems are most likely to collapse and why.
“If the model works the way we anticipate,” Lamm said, “it will be transformative for prediction modeling that will impact vulnerability and resilience applicable to microbes, human healthcare, disease, livestock, and wildlife.”
That is a sentence that covers almost every living system on Earth.
Where This Came From
Astromech did not appear from nowhere. It grew directly out of the work Colossal has been doing since 2021.
Think about what Colossal has actually built over the past five years. A genomic database containing the DNA of extinct and living species at a depth and breadth that has never existed before. Computational tools — many of them now commercialized through Form Bio — for analyzing massive biological datasets. A scientific team that thinks routinely about evolutionary timescales, about how species changed over thousands of years, about the genetic mechanisms that drive adaptation and vulnerability. A set of techniques for reading ancient DNA and comparing it to living genomes to identify what changed and when.
All of that is, at its core, the raw material for exactly what Astromech is trying to build. A model that has been trained on the history of biological change across deep time — one that can look at a living system and say: based on everything we know about how biology evolves, here is what we expect to happen next, and here is where the system is most vulnerable.
Astromech is hiring for genomic inference, synthesis design, ancestral modeling, gene regulation, sequence reconstruction, metabolic modeling, and protein folding. The job listings read like a map of the exact scientific capabilities that Colossal spent four years assembling. The spinout isn’t a departure from the main mission. It’s the main mission’s most powerful tool, built out as its own company.
Medicine reacts after damage begins. Astromech aims to predict threats before they emerge—turning biology from a crisis response system into an early-warning engine for what comes next.
The Problem It’s Solving
One of the persistent frustrations of modern medicine and public health is that we are almost always reactive. A new pathogen emerges, and we scramble to understand it. A disease becomes drug-resistant, and we scramble to find alternatives. An ecosystem begins to collapse, and we scramble to identify the cause. The scrambling is expensive, slow, and often too late.
The COVID-19 pandemic made this painfully visible to the entire world. The virus existed in animal populations long before it crossed into humans. The genetic tools to identify it were available. The computational power to model its likely behavior was available. What wasn’t available was a system sophisticated enough to put those pieces together and say: this is coming, and this is what it will do.
Astromech is a direct response to that gap. Not the only response — there are other early-warning and pandemic-preparedness initiatives working on related problems. But it may be the most ambitious one, because it’s not just trying to spot specific known threats earlier. It’s trying to build a general model of biological vulnerability — one that could flag a threat that nobody has identified yet, because the model has recognized the pattern that precedes it.
That’s the difference between a smoke detector and a system that predicts where fires are most likely to start.
Why This Valuation Makes Sense
Two billion dollars for a company with no revenue and no product yet in the market sounds, on the surface, like the kind of number that raises eyebrows. But the valuation logic is straightforward if you understand the market.
The global pandemic preparedness market alone is measured in hundreds of billions of dollars, and governments around the world spent the last five years being reminded, painfully, what under-investment in early warning systems actually costs. Drug discovery — which predictive biology could accelerate dramatically by identifying drug resistance patterns before they become treatment failures — is a multi-trillion dollar industry. Agricultural biotech, conservation biology, livestock health management: each of these is a substantial market in its own right.
A platform that works across all of them, built on some of the most sophisticated genomic and evolutionary data ever assembled, co-founded by the team that just built a $10 billion company from scratch in four years — investors have seen enough from Lamm and Church to know the ambition is real. The question isn’t whether the idea is valuable. It’s whether the science will hold.
Lamm thinks it’s undervalued. That’s the kind of thing founders say. But he said the same thing about de-extinction in 2021, and three dire wolf pups are living on a farm somewhere right now as evidence that he wasn’t wrong.
From revival to prediction—the tools keep expanding. Astromech’s bet isn’t fixing biology, but forecasting it, shifting humanity from reaction to anticipation at a planetary scale.
The Biggest Bet Yet
Each company in this series has been bigger than the one before it. Colossal brought back an extinct species. Form Bio built the operating system for a new era of biological research. Breaking developed a microbe that eats one of the most persistent pollutants in history. Each one started as a tool built to solve a specific problem, and became something larger than the problem that created it.
Astromech is the biggest bet in the portfolio. Not because the technology is further from reality — it’s actually built on real science with real precedents. But because the potential outcomes are the most consequential. A forecasting engine for biology, if it works the way Lamm describes, doesn’t just change one industry. It changes how humanity manages its relationship with the living world — from treating disease after it strikes to anticipating it before it forms.
That’s not a pharmaceutical company. That’s not a biotech company. That’s something new.
Up Next: The Colossal Foundation — the Noah’s Ark that Lamm is building at the cellular level, and what it means to preserve the genetics of every species before they’re gone.
Related Reading
The Next Pandemic Could Come From Anywhere. Here’s How Scientists Are Watching for It
Scientific American — How early-warning systems for biological threats actually work today, what their limitations are, and why predictive modeling is the frontier the field is racing toward
AlphaFold and the AI Revolution in Biology
Nature — The story of how AI cracked one of biology’s hardest problems and what it opened up — the clearest existing precedent for what a truly powerful predictive biology platform could accomplish
Can We Predict Evolution?
Quanta Magazine — A deep look at the science of evolutionary forecasting — what biologists have already shown is predictable about how living systems change, and where the real frontiers of the field lie