There’s something almost embarrassingly primal about the appeal. You take an elephant, you tinker with its genome using DNA pulled from a 4,000-year-old carcass frozen in Siberian permafrost, and out comes something that looks and acts like a woolly mammoth. Not a perfect replica — more of a cold-resistant, tusk-bearing Asian elephant with mammoth traits stitched in — but close enough to make you reconsider what “extinct” actually means. This isn’t a thought experiment anymore. It’s a business plan. And it’s attracting serious money.
Colossal Biosciences, founded in 2021 by serial entrepreneur Ben Lamm and Harvard geneticist George Church, has raised over $200 million to pursue de-extinction across three flagship species: the woolly mammoth, the thylacine (Tasmanian tiger), and most recently, the dodo. The company isn’t alone in the space, but it has become the undisputed poster child — a venture-backed bet that synthetic biology has matured enough to reverse what human activity has destroyed. The pitch blends scientific audacity with ecological sincerity: bring back keystone species, restore degraded ecosystems, buy humanity some redemption for its role in the sixth mass extinction.
Whether that’s visionary or delusional depends on which questions you ask.
The Science: CRISPR, Proxy Genomes, and Honest Limits
The mechanics of de-extinction are less Jurassic Park and more sophisticated cut-and-paste. You can’t clone a mammoth the way you’d clone a sheep — there’s no intact mammoth cell with a viable nucleus. Instead, researchers sequence fragmented ancient DNA, compare it against the genome of the closest living relative, and identify the key genetic differences that defined the extinct species. Then they use CRISPR and other gene-editing tools to edit those traits into the living relative’s genome.
For the mammoth, that means editing Asian elephant cells to express mammoth hemoglobin (for cold tolerance), subcutaneous fat, smaller ears, and that iconic shaggy coat. The resulting animal wouldn’t be Mammuthus primigenius. It would be an elephant with roughly 40 to 50 mammoth genes expressed — a “proxy” creature designed to fill the mammoth’s ecological role, not to replicate its genome base pair for base pair.
The thylacine project, led by evolutionary biologist Andrew Pask at the University of Melbourne in partnership with Colossal, has a slight advantage: the thylacine went extinct in 1936, meaning preserved tissue samples yield far higher-quality DNA than permafrost bones. Pask’s team has already sequenced the thylacine genome and identified the fat-tailed dunnart as the closest living relative. The dodo project, announced in 2023, uses the Nicobar pigeon as its genetic template.
But here’s the honest technical reality:
- Editing is not gestation. Even if you produce a viable edited genome, you still need to turn that into a living organism. For the mammoth, Colossal’s plan involves artificial wombs — technology that doesn’t yet exist at the scale required. Using a surrogate Asian elephant raises enormous ethical and practical problems, given the species is endangered.
- Epigenetics matter. DNA is not destiny. Gene expression is shaped by environmental and developmental cues that can’t be reconstructed from sequence data alone. A proxy mammoth raised in a lab would not behave like a mammoth.
- Microbiomes are invisible co-pilots. Every animal carries an ecosystem of gut bacteria, skin flora, and symbiotic organisms that shape its digestion, immunity, and behavior. We can’t resurrect what we haven’t even catalogued.
The Conservation Argument: Is This the Best Use of Resources?
This is where the debate gets sharp. Colossal frames de-extinction as a conservation tool, and there’s genuine logic to parts of the argument. The mammoth’s proposed habitat — restoring Arctic tundra to grassland — has climate implications. Pleistocene Park, Sergey Zimov’s long-running experiment in northeastern Siberia, has demonstrated that large grazing animals can transform permafrost ecosystems, potentially slowing carbon release. A proxy mammoth could, in theory, be the keystone grazer that makes such restoration scalable.
The thylacine case is even more emotionally resonant. Humans hunted it to extinction. The last one died in a Hobart zoo in 1936, neglected, exposed to the elements. If any species deserves a second chance, it’s the one we killed out of fear and ignorance.
But critics — including many working conservationists — raise a uncomfortable economic point. Every dollar spent engineering a proxy mammoth is a dollar not spent protecting the roughly 41,000 species currently threatened with extinction, many of which could be saved with comparatively modest funding. The Sumatran rhino, the vaquita, the northern white rhino — these are animals that still exist, barely, and could be pulled back from the brink with habitat protection, anti-poaching enforcement, and captive breeding. They don’t require artificial wombs or CRISPR arrays.
“De-extinction gets the funding and the headlines because it’s technologically thrilling. Protecting wetlands for the whooping crane doesn’t go viral. But one of these strategies will actually prevent extinction in our lifetime, and the other might produce a single animal that can’t survive in the wild.” — Dr. Beth Shapiro, evolutionary biologist and author of How to Clone a Mammoth
Shapiro, notably, has contributed to Colossal’s science. Her critique isn’t anti-de-extinction. It’s about priority and honesty.
The Ethics: Playing God or Cleaning Up?
The theological framing — “playing God” — is lazy and unhelpful. Humans have been reshaping ecosystems since we learned to control fire. The relevant ethical question is narrower and more practical: what happens to a proxy mammoth after it’s born?
There is no ecosystem waiting for it. The Arctic tundra it would theoretically restore is fragmenting under climate change faster than any de-extinction timeline can match. A single mammoth — or even a small herd — would essentially live in a managed reserve, a curiosity visited by donors and photographed for press releases. Is that conservation, or is it a very expensive zoo exhibit with better branding?
The thylacine faces a different problem. Tasmania still has habitat, but the ecological niche the thylacine occupied has shifted in its 87-year absence. Predators don’t slot back into ecosystems like batteries into a device. Reintroduction would require years of monitoring, conflict management with livestock farmers (the same conflict that doomed the species originally), and public buy-in that hasn’t been tested.
And the dodo — a flightless bird that evolved on an island with no mammalian predators — would be released into what, exactly? Mauritius is not the ecosystem it was in the 17th century. The dodo’s return would be symbolic, not ecological.
The Real Product Isn’t the Animal
Here’s the thing that gets lost in the spectacle: Colossal’s most valuable output may not be a mammoth at all. The company is pushing the boundaries of synthetic biology, gene editing, and reproductive technology in ways that have applications far beyond de-extinction. The artificial womb research could advance premature infant care. The genetic rescue techniques being developed for the thylacine could help preserve genetic diversity in endangered populations of living species. Colossal has already spun out a subsidiary, Form Bio, focused on computational biology for drug discovery.
De-extinction is the narrative. The technology is the business. And that’s fine — venture-backed science often works this way, funding breakthroughs through compelling stories that capture imagination and capital. The Apollo program gave us Velcro and integrated circuits alongside moon rocks.
But we should be clear-eyed about what we’re buying. A proxy mammoth will not save the Arctic. A thylacine in a Tasmanian reserve will not undo the colonial violence that erased it. The dodo will not restore the forests of Mauritius. What these projects will do is prove that humans can reverse extinction — technically, partially, at enormous cost — and that capacity, once demonstrated, will be impossible to un-invent.
The question isn’t whether we can bring back the dead. We probably can, in some form. The question is whether we’ve earned the right to, given how poorly we’ve managed the living.
