Colossal Biosciences has announced a new artificial intelligence platform designed to accelerate its de-extinction work, and while the company's headline ambition—resurrecting the woolly mammoth—still reads like science fiction, the underlying technology deserves serious attention. The Dallas-based startup is applying machine learning to one of biology's thorniest problems: predicting how genetic edits will actually behave in living organisms.
The platform, which Colossal is calling its "species restoration engine," uses large language models trained on genomic data to simulate the effects of thousands of potential gene modifications before any wet-lab work begins. In practical terms, this means researchers can computationally test which combination of cold-resistance genes from mammoth DNA might successfully integrate into Asian elephant cells—without burning through years of trial-and-error experimentation.
The synthetic biology bottleneck
De-extinction has always been less about sequencing ancient DNA (we have had woolly mammoth genomes since 2015) and more about the painstaking work of editing living cells to express extinct traits. Traditional approaches required scientists to make educated guesses, introduce changes, grow cells, and observe results—a cycle that could take months per iteration. Colossal claims its AI can evaluate millions of potential edits in the time a human team might test dozens.
This is not unique to de-extinction. The broader synthetic biology industry faces the same constraint: the gap between designing genetic modifications on a computer and knowing whether they will work in practice. Companies engineering microbes to produce pharmaceuticals or sustainable materials confront identical challenges. Colossal's platform, whatever one thinks of mammoths, represents a genuine attempt to close that gap.
The hype-to-substance ratio
Colossal has raised over $225 million since its 2021 founding, backed by investors including Thomas Tull and Paris Hilton. The company has promised not just mammoths but also the resurrection of the thylacine and the dodo—a menagerie of charismatic extinct species that generates considerable press coverage and, critics argue, distracts from more urgent conservation priorities.
The AI announcement arrives as the company reportedly prepares for a funding round that would value it north of $1.5 billion. Skeptics will note the timing. Yet even researchers who doubt Colossal will ever produce a living mammoth acknowledge that the computational tools being developed have legitimate applications. Predicting gene expression, modeling protein folding in novel contexts, and simulating cellular behavior are problems that matter well beyond de-extinction.
Our take
Colossal Biosciences operates in the uncanny valley between serious science and Silicon Valley spectacle. The mammoth is a marketing hook—a furry, trunk-nosed embodiment of technological optimism that photographs well and opens wallets. But beneath the circus tent, the company is building infrastructure that could genuinely advance synthetic biology. Whether that justifies the valuation is another question. The AI platform is real; the mammoth remains, for now, a very expensive thought experiment.
