Colossal Biosciences, the company best known for its ambition to bring back the woolly mammoth, has announced a new AI platform intended to accelerate its de-extinction work, TechCrunch reports.
The logic is straightforward even if the goal is not. Rebuilding a lost species is not one problem but thousands: which stretches of DNA matter, which edits are compatible with one another, how a modified embryo develops, and which of the resulting traits survive contact with the real world. Colossal already depends on large-scale genomic comparison to identify the differences between an extinct species and its closest living relative. A machine-learning layer that prioritises candidate edits and predicts their downstream effects is a natural extension of that pipeline — and a plausible place for AI to earn its keep, since the search space is enormous and the ground truth is a laboratory result.
The company has not been shy about the commercial framing around the science. Alongside mammoth, thylacine and dodo programmes, it has built out work in synthetic biology, genetic engineering tools and conservation genetics, arguing that the technologies developed to resurrect one species have nearer-term uses for species that are still with us. That is a reasonable argument, and it is also how a long-horizon research programme pays its bills.
Scepticism remains warranted where the science meets the world. Even with excellent tooling, a self-sustaining population of a proxy species is years away, and the ethics of releasing one into a habitat that has changed since its ancestor disappeared are unresolved. A faster pipeline shortens the research loop. It does not answer the question of what should be built.
