Race to Save Humanity: Inside $12B Colossal Biosciences & $3.8B Astromech

Summary

• Colossal’s woolly mice were a pipeline proof, not a publicity stunt — 100% healthy mice born with exactly the predicted phenotypes, carrying “the most multiplex edits delivered at one time ever,” engineered in 30 days. Lamm’s frustration is that “the 30-day part got lost” amid the cuteness discourse; the point was validating mammoth-target edits in a 21–22-day-gestation animal instead of a 22-month endangered elephant.
• Colossal has raised $635M total, with the last round supposed to be $200M and oversubscribed to $400M — and every round has been more than 2x oversubscribed. Lamm’s read: “we’ve completely mispriced every single round… we’re a cheap date and we should raise our rates on capital,” and he calls Colossal “one of the most undervalued companies in the world” given the ability to “preserve, engineer and direct life on a planetary scale.”
• The business model is not selling animals — Molly frames the broader opportunity as “civilization tech,” while Lamm describes government-scale use cases: gene drives against invasive species (a “half a trillion dollar problem globally”), artificial wombs, and human-health applications. The screwworm case is the tradeable math: taxpayers are spending ~$1.5B over 8 years to slow the problem, while Colossal’s already-built solution “could eradicate the entire problem in every model that we’ve done in six to seven months” — pending government permits it does not yet have.
• Spinout Astromec (raised recently, “may or may not be raising money now soonish”) uses focused evolutionary-genomics AI for drug discovery — “we don’t have to go build a $500 billion data center.” The mechanism: elephants get cancer at 2–3% vs. humans’ 24–25% via p53 overexpression, and Colossal holds the world’s largest elephant and multi-species herpesvirus datasets, informing prediction models applicable to cancer evolution and bird flu.
• The episode’s extinction setup is a forecast of up to 50% biodiversity loss in the next 25 years, and Lamm says conservation “just doesn’t move at the speed at which we’re eradicating species.” His hedge against critics: de-extinction is “net-new money, net-new technology,” not a replacement — all technologies are open-sourced free for conservation use — and “wouldn’t you rather have a de-extinction toolkit and not need it than need a de-extinction toolkit and not have it?”
• Lamm’s hottest take: AI will converge into “a digital twin of nature” — real-time prediction models like those for missile defense and tsunamis, showing the ecosystem consequences of every data center or diverted river. He calls it “incredibly enlightening and terrifyingly scary,” and sees it enabling productionized species to offset adverse effects of human decisions.
• Editing capability is compounding faster than promised, and the next milestone is teased but unnamed. Investors said 5–7 edits within five years would be great; Lamm says Colossal is now doing 100 edits, while elsewhere saying “hundreds,” and is pushing a new construct and delivery mechanism for thousands of edits at once. In the next 12 months: “Colossal potentially having the next extinct species birthed” — which one, “I’m not saying.”

Deep dive

1. The woolly mouse was an end-to-end pipeline demo — and the 30 days mattered more than the cuteness

• The Gen 2 mice Lamm unveils on camera are offspring of the original cohort — “the second-most genetically modified multicellular organisms on the planet after our dire wolves.” His core complaint about the launch coverage: “we had 100% healthy mice born with exactly the predicted phenotypes with the most multiplex edits delivered at one time ever, and we did it in 30 days… the 30-day part got lost.”
• The logic of the mouse: elephants gestate 22 months and are endangered, so Colossal identified the mouse equivalents of its mammoth edits and tested phenotype predictions in a 21–22-day-gestation animal. “It showed that our end-to-end pipeline worked.”
• The aftermath was surreal: superfans showing up at the non-public office, “legalized woolly mice” Etsy shirts (“which is weird because they’re not illegal”), the scientist in China associated with the CRISPR babies posting that they were bioweapons, and a Smithsonian request—recounted by Molly—to taxidermy the mice when they die. Lamm’s line on purchase requests: “We’re not in the selling-animals business.”

2. $635M raised, every round 2x+ oversubscribed — built on a deck Lamm admits he’d have passed on

• The last round was supposed to be $200M and was oversubscribed to $400M; Colossal has raised $635M total, and “we’re just getting started. I’m sure we’ll raise more capital.” Since every round has been more than 2x oversubscribed, “what I hear from that is we’re cheap… we’re a cheap date and we should raise our rates on capital.”
• Investor curation was deliberate: long-horizon technology and impact investors — Jim Breyer, Thomas Tull, and Mark Walter — not “fund return in four to five years” money, and not the Masa/SoftBank route they initially considered but did not pursue. Investors had to accept from day one that “inspiration needs to be as important as value creation.”
• His honest origin story: the original deck was “we have no scientists, we have no partnership with Harvard, we have no technologies… and George Church thinks he can make a woolly mammoth.” His verdict: “I would have passed on this. Like in all reality, I would have passed on this.” What fixed it — trust via setting and exceeding expectations, with the dire wolves as the signature “surprise-and-delight moment.”

3. The real business: gene drives, artificial wombs, and demand Colossal didn’t know existed

• Molly introduces the “civilization tech” label; Lamm describes it as a category of technologies with effects broader than their immediate products or services. Invasive species alone is a “half-trillion-dollar problem globally” — rats on islands, screwworm in Texas, and ticks spreading Lyme disease and harming moose and deer in the Northeast — where the alternative today is spraying chemicals that poison ecosystems.
• Inbound demand keeps surprising him: a buyer of billions of dollars of an ethically harvested animal fabric asked Colossal to 5x domestic supply — possibly via cloning, possibly via the lab-grown route Colossal already demonstrated with mammoth hair follicles grown from elephant stem cells, “plates and plates of that… full warehouses… without even touching the animals.” Separately, interests in the Middle East want an extinct or semi-extinct bird and a critically endangered bird revived after billions spent on unsuccessful artificial-egg programs.
• Human-health adjacencies he claims leadership in: artificial-womb components for IVF and embryo health, through germline editing of genetic disease states — “Colossal is the world leader in that.” The named risk: “making sure that we stay focused on our core mission and don’t deviate too much for some of these large economic opportunities.”

4. Astromec: drug discovery from evolutionary trends, not brute-force compute

• The spinout Astromec “raised money recently and may or may not be raising money now soonish” and is deliberately narrow: “we don’t have to go build a $500 billion data center… we don’t have to fight for water.” Instead it tracks why evolution conserved certain genome regions across species.
• The carrying example: elephant p53 overexpression is associated with cancer rates of 2–3% versus humans’ 24–25% — and Colossal holds the world’s largest elephant dataset plus the largest dataset of multiple species of herpesviruses (EEHV kills elephants), informing prediction models “applicable not only to cancer evolution, but also things like bird flu.”
• Lamm’s macro call on AI-bio convergence: “you’re going to see this pre-Cambrian explosion of new drugs and new therapies that all come from AI-assisted drug design and discovery.”

5. The dark math: 50% biodiversity loss in 25 years, and the toolkit argument against conservation critics

• Molly sets up the statistic: “It’s currently forecasted that we’re going to lose up to 50% of all biodiversity in the next 25 years.” Lamm calls it dark but hopeful: “that’s why we have to do stuff.” Conservation isn’t a replacement issue, he says; “no matter how good conservation works, it just doesn’t move at the speed at which we’re eradicating species.”
• The conservation community was historically among the most vocal critics — feedback went from “negative to neutral” and is now becoming more positive. His counter: ask whether humanity is winning the fight to save species and conservationists “will all tell you no.” Colossal is “net-new ideas, net-new money, net-new technology”; all its technologies are open-sourced free for conservation, and elephant conservation partners will tell you “their funding has gone up since Colossal was invented.”
• The signature framing, complete with the Star Trek blue-whale time-travel reference: “wouldn’t you rather have a de-extinction toolkit and not need it than need a de-extinction toolkit and not have it?” — especially since “we know we’re in” a mass-scale extinction event right now. On misuse: bad actors are inevitable, but “that doesn’t mean that we shouldn’t innovate the technologies for positive use cases.”
• His hottest take ties it together: humanity will build “a digital twin of nature” — prediction models like missile defense or tsunami warning, but for ecosystems — so we understand the ripple effects of eradicating a species or “building XYZ data center” on water purity, food security, and zoonotic disease. “Incredibly enlightening and terrifyingly scary.”

6. Rewilding is real but government-gated — and the screwworm shows the stakes

• Everything — extinct species, gene drives, and cloned endangered animals — is intended to go back into the wild, but only with governments, Indigenous groups, and private landowners aligned: “they have to give us the green light.” Colossal is currently educating the USDA, the intelligence community, and the Department of War on the technology.
• The screwworm arithmetic, sharpened after Bluesky critics noted Colossal lacks permits (“yes, we do not have permits, but…”): the government is spending $1.5B in taxes over 8 years to slow or stop a problem threatening the Texas cattle industry, while Colossal’s already-made solution “could eradicate the entire problem in every model that we’ve done in six to seven months” — lower beef-price pressure, less animal suffering, and taxpayer savings without chemical spraying.
• The 12-month tease: what he’s most looking forward to is “Colossal potentially having the next extinct species birthed.” Which one? “I’m not saying.”

7. Org design: buy the whole stack, then reprogram the scientists

• Lamm’s self-deprecating credentialing — “having zero background in biology and biotech, I couldn’t be more qualified to run this company” — masks a first-principles vertical-integration thesis: “extinction is a systems model,” so Colossal built, licensed, or bought the supply chain. Buying ViaGen was the exemplar: its cloning efficiencies run at 80% “where most people are at 2%.”
• Scale: 260+ people, 17 academic partners, 80 people in academic labs that Colossal funds, and about 95 tiered advisers — scientific, Indigenous, executive, and a youth advisory board, because kids “are also going to inherit our solutions” and he doesn’t want to hand them “tech debt” like what he calls “dumb” direct-air-capture carbon infrastructure.
• The AI stack had to be bespoke: “it turns out that Opus doesn’t do an ancestral-state reconstruction of a genome.” Custom laser/cloning hardware with full-time screen recording and computer vision feeds efficiency tracking, eventually to be fully automated with robotics — and “the hardest thing we’ve done is reprogram not stem cells, but scientists” to work in Jira and Smartsheet, now aggregated through an AI layer for near-real-time lab visibility.

8. “Harvard meets MTV,” the no-go list, and programming redwoods into houses

• The brand thesis from day one: old-school TRL-era MTV energy mashed with Harvard’s scientific rigor — “if I said Harvard meets MTV, you’d be like, that’s just dumb. But… that actually describes exactly what you guys have built.” It’s also a recruiting weapon: “go work back in academia after you work here… your labs are held up by pizza boxes.”
• The boundary is explicit: no human work in-house (technologies can be spun out and licensed), and no work on nonhuman primates in vivo. Lamm says he has been asked about Neanderthals but does not separately elaborate on that question. Gigantopithecus “would be amazing to see, but we’re not going to do that.” He also says there is no megalodon DNA, despite dinosaurs and megalodon being requests one and two. He flatly denies the conspiracy fodder: no secret mammoth, no aliens behind the genome-engineering advances — “just crazy talk.”
• The long-horizon vision he and George Church share, weird enough that “it was even too weird for Adam Draper”: synthetic biology eventually programming life itself — “can you build a redwood that actually grows in the form of a house?” His summary of the field: “we’re advancing it, I think, further than anyone, but at the same time, we haven’t even scratched the surface.”