The wonder and controversy of bringing back the dire wolf from extinction | Colossal Biosciences interview

As soon as Colossal Biosciences declared that it brought the dire wolf back from extinction, everyone felt a sense of wonder. George R.R. Martin himself, who popularized the dire wolf as Jon Snow’s Ghost in A Song of Ice and Fire, posed for a picture with the pups to be part of a historic scientific achievement. He wept at the sight of the snowy fur of the white wolves.

To me, along with Colossal Biosciences’ other work of making mice with the hair of wooly mammoths, the work was a reminder of what science can achieve when it’s supported with an amazing amount of talent and capital — Colossal Biosciences has raised $435 million at a valuation of $10.2 billion. It is the stuff of science fiction, as Michael Crichton wrote Jurassic Park based on the notion that genetic material could be recovered for dinosaurs and they could be brought back to life. Colossal’s chief aim is to bring back the wooly mammoth, the dodo and the thylacine.

The computing power and the genetic tools finally exist to make that science fiction into reality — which is one of my favorite topics in the world and why I’m straying from games to write about this. After all, the only thing better than the intersection of science fiction, tech and games is the intersection of science fiction, tech and Game of Thrones. The scientists extracted DNA from recovered fossils a tooth from Sheridan Pit, Ohio, where the fossil was 13,000 years old, and an inner ear bone from American Falls, Idaho, that was 72,000 years old.

Some critics felt like attacking the company for pulling some kind of scam for dressing up dogs as an extinct species because they used too little of the original dire wolf DNA. More seriously, some say that it was the creation of a brand new species by humans, not the bringing back of a vanished species.

Of course, some people had to rain on the parade. Colossal Biosciences saw the objections that some scientists had about how these were not really dire wolves and that they were more like dogs, and that the amount of DNA they had to work with was insufficient. I can’t say, but them wolves sure do look like Ghost, Jon Snow’s dire wolf in Game of Thrones. Kidding aside, we’ll see how their work will stand up to scientific scrutiny soon enough.

Here’s the statement that the company issued after the criticism:

We understand that some scientists are not comfortable calling these dire wolves because they feel like the wolves are not sufficiently genetically similar to an extinct individual to merit that name. That’s OK with us. We can disagree about what makes a dire wolf qualify as a dire wolf, or what makes a mammoth qualify as a mammoth.

Colossal has 500 times more data than anyone has ever had on a dire wolf. We have had a small army of people doing comparative genomics to wolves and other canids for the last 18 months with this proprietary data set. We know what makes a dire wolf a dire wolf including that it is not closer to a jackal. We will be submitting that data next week for peer review. Colossal has always said that we are doing functional de-extinction where are looking to de-extinct the core genes that make a species a species as it relates to their phenotypes or physical attributes.

All animals on this planet are ad mixtures. Just like a polar bear is a white adaptive bear compared to a brown bear a dire wolf is a plasticine wolf when compared to a grey wolf – it is 20% to 25% percent larger, more muscle mass, has an arctic white coat, is stronger and bigger, and has cranial facial structure. We have identified the genes that drive those phenotypes and de-extincted them as we are doing on all our projects.

The scientific community does not agree on how to classify species because it is a man-made construct that does not apply to nature and that is why there are so many variants of it. Under several of the variants the dire world would be classified as a dire wolf. We are calling it a dire wolf because it is a dire wolf. If you do not want to call it a dire wolf you can always call it Colossal’s dire wolf.

Our interview

In my interview, Colossal Biosciences’ leaders told me that they had abided by ethical procedures in their recovery work. They say that half the species on earth are in danger of going extinct in the next 50 years, and that would be part of an ecosystem collapse like we have never seen. Saving species from extinction is also applauded by indigenous peoples that Colossal Biosciences consulted as it proceeded. It’s interesting that these people can see their myths come back to life, and it conjures thoughts on what it means to play God with science. But rather than play God, the team believes it is saving animals — including some that are gone because of humans.

After the announcement yesterday, I interviewed Colossal Biosciences’ CEO Ben Lamm and Beth Shapiro, an evolutionary biologist who specializes in the genetics of ice age animals and plants. She is also chief scientist at Colossal.

As professor of ecology and evolutionary biology at UC Santa Cruz and HHMI Investigator, Shapiro was instrumental in the work. She uses DNA recovered from bones and other remains to study how species evolved through time and how human activities have affected and continue to affect this dynamic process.

Here’s an edited transcript of our interview. I don’t think we’ll get to the bottom of who is right about whether this is de-extinction or a false claim for a while. But the subject is fascinating and I hope it will make us think about de-funding science in the current administration.

VentureBeat: How much time has it taken you to get to this point?

Ben Lamm: We’ve been building the de-extinction toolkit since 2021, when we launched the company. Obviously we announced the woolly mouse a month and a half ago, which you’re aware of. We started the dire wolf project about 18 months before we birthed the first dire wolves in October 2024.

VentureBeat: How do you extract DNA successfully from such an old fossil?

Beth Shapiro: This is the science I’ve been doing in my academic career since 1999. The first time anyone successfully extracted DNA from something that wasn’t alive was in 1984. It inspired Michael Crichton to write Jurassic Park. The field has gotten a lot better since then. We’re much better at extracting DNA from older remains. It used to be that we could only get things from remains preserved in the Arctic, where they were frozen. But we’re much better at extracting from things in warmer places. The record for the oldest DNA that’s been recovered from a bone is a mammoth bone that’s between 1 million and 2 million years old. Probably closer to 2 million. But most of the DNA we’ve recovered dates to the last several tens of thousands of years. One of our dire wolves is actually 72,000 years old, making it one of the oldest genomes that’s ever been recovered and extracted to date.

We did the DNA extraction with some academic collaborators. It was done at my lab at the University of California Santa Cruz. We’re the Paleogenomics Lab. We’ve been pioneering new technologies for getting DNA out of old remains, things like rootless hairs and all sorts of cool stuff that’s going on. The DNA is not in good condition. If I were to extract DNA from a cheek swab of my own face, I could get strands of DNA that were tens to hundreds of millions of letters long. The DNA we get out of the dire wolf bones is maybe 35 letters long. We have to extract millions to hundreds of millions to billions of fragments and then use a computer to figure out how they line up against the genome of something that’s still alive, where we have a good quality genome, to piece it together.

That was step one. Get the DNA out of old bones, line them up to reconstruct the dire wolf genome, and then compare that – again, with a computer – to a grey wolf genome, a coyote, jackals and all sorts of other canids, wolf-like animals, to figure out what its closest living relative is, and then what genetic variations make it uniquely dire wolf. We knew that dire wolves are most closely related to grey wolves. They’re about 99.5% genetically identical to grey wolves. They look a lot like grey wolves except they’re larger and more muscular. They have different hair patterns, different length and thickness and color. We learned by looking at our grey wolf genomes that dire wolves are light colored in code, which is cool. Obviously we didn’t know that looking at bones. But from the genome we could learn that..

VentureBeat: It seems like George R. R. Martin had done his homework.

Lamm: He actually cried when he first saw them. He was so excited.

VentureBeat: I remember taking a class at Berkeley in anthropology. They were telling us way back then that a very small percentage of DNA separated us from the apes. It sounds like that works in your favor. You can rebuild it by finding a small percentage of DNA.

Shapiro: We focused specifically on DNA variants that were most likely to bring back these key traits – the size, the hair patterns, the musculature of dire wolves. It’s not possible to re-create something that’s 100% genetically identical to something that used to be alive, but that’s not the goal of de-extinction. Our goal is to re-create these phenotypes, these key traits, so that we can put these animals back into ecosystems and restore missing components of those ecosystems.

VentureBeat: How do you know that the DNA sequence is not somehow messed up? How do you make sure you won’t get some weird variant of a dire wolf instead of the real thing?

Shapiro: We’re focusing specifically on DNA variants that are in both of our dire wolf fossils, and we know what they do. That’s one of the other really good things about working with dire wolves. We know a lot about grey wolves. Everyone has their own favorite grey wolf. Mine is right here. Because of that we have lots of information about DNA sequence variants and what they do. We know a lot about what causes eye color, hair texture and thickness and density. When we see particular variants in the dire wolf genome, we can predict with confidence what they’re going to do.

One of the benefits of starting with the dire wolf project is that–it’s not easy at all, but it’s simpler than some of our other announced species as far as getting to a predictable phenotype. Obviously we want a healthy animal that expresses these traits that have been extinct. Finding a way to get there using DNA and genome editing, and then cloning, is going to be hard with every animal, but with some animals there are steps that we’re better at already.

VentureBeat: With a woolly mammoth being so different from an elephant today, would that task be bigger?

Lamm: They’re really not. Asian elephants are 99.6% the same genetically as woolly mammoths. They’re actually more closely related to mammoths than they are to African elephants.

Shapiro: The challenge with elephants is that we know a lot less about elephants than we do about grey wolves. We know a lot less about elephant animal reproductive biology than we do about grey wolf reproductive biology. We have partners in elephant sanctuaries, elephant conservation organizations on the ground. We’re developing new tools and protocols that will benefit elephant conservation as we learn about what we need to do to make our mammoths. But with grey wolves, a lot of that was already known. We could ride on the shoulders of scientific research that’s happened over the last several decades. For the animal reproductive biology part, that is, not the ancient DNA part, which is brand new. We had to do that ourselves.

VentureBeat: That reproductive biology seems pretty interesting in itself, the idea of interspecies gestation. Is that not rocket science so much?

Shapiro: It’s hard. But the surrogates for our dire wolves were large domestic dogs, hounds. Domestic dogs have, in the past, birthed grey wolves. They’ve never birthed dire wolves before. But because they’re so genetically similar to each other, we predicted that this would be–it’s not one of the many challenges. We’ll have challenges like that when we move to, for example, the dunnart and the thylacine. They’re more distantly related. We’ll get there. Interspecies cloning has happened before, including things as distantly related as the two different camel species, the one-humped camel and the two-humped camel. It’s just harder.

As this technology gets better, it’s also benefiting conservation more broadly. We want to develop technologies to have common species be able to be surrogates for rare species or more endangered species. Again, this is another way that Colossal’s work is contributing to developments that have real utility for conservation.

VentureBeat: From here you do go to enable them to breed among themselves, to repopulate?

Lamm: We’ve made three. We’re probably going to make another three to five more so we get the right pack dynamics. They live in a 2,000-acre ecological preserve with 10 full-time care providers. They live a seemingly wild life. Six and a half acres of that is a sub-preserve where we have animal husbandry, an animal hospital, storm shelters, natural dens and whatnot, feeding, all that stuff. That’s where they live today.

Long term, we’re in talks with MHA Nation and other indigenous people groups that relate them to great wolf in their mythology, in their cultures and their oral traditions. They want them potentially back on their own land. We’re working with them on a potential long-term rewilding plan back to, once again, very secure expansive ecological preserves.

VentureBeat: This is a funny question for a serious subject, but I can think of some sheep that would rather not see the dire wolf back. How do you think about that?

Lamm: We keep them very separate. There’s a whole lore about wolves in general. But interesting enough, only .02% of wolves ever attack anything livestock-related. It’s very rare. If that ever happens, it’s subsidized by the government. Not that it’s a good thing, but the wolves are–our goal is to never put the wolves back near ranching-type communities.

VentureBeat: I don’t know if this is an ethical issue or just an ecosystem issue, but bringing things back that are gone–does that carry with it some decisions about what you ought to do? If you brought a tyrannosaurus back, there are parts of the ecosystem that wouldn’t appreciate it. That’s an extreme example, but I do wonder how you think about these things as you go about the work.

Shapiro: Our goal for de-extinction is to create technologies, and that includes species that are able to help ecosystems that are threatened because of things that people have done to them today, whether extinction or continued decline, to be able to stabilize. When we make a decision about what species to bring back, we have to understand that there’s a place for them to go, that we’ve corrected whatever wrong it was that caused them to become extinct – like rats with the dodo on Mauritius – and that there are communities of people that want these species here. The dire wolf project was launched after long collaboration and conversation with our indigenous partners, who see this as a culturally important species. They’re willing and would like to become stewards of the species in the long term. There’s a lot of thought that goes into choosing a species for de-extinction.

As we do this, we’re developing technologies that we will be able to apply directly to protect and preserve species that are still alive. Some people are always going to be fearful of any sort of new technology. Not even a biotechnology, but any new technology. It’s important that we remember that if we decide not to reach into the trenches of human ingenuity and come up with these new tools and apply them, that’s also an active choice. It’s not just passively saying, “This is too scary. I won’t do it.” It’s an active choice that has consequences. We know what those consequences are. Half the species that are alive today are in danger of becoming extinct in the next 50 years. Habitats around the world are changing at a rate faster than evolution can keep up. If we don’t develop and deploy these technologies, the future will be much less biodiverse than it is today. That’s not a risk I’m willing to take.

VentureBeat: Were you all into science fiction when you were younger?

Lamm: I was always into science fiction as a kid. I’ve always loved building technologies that are in the future, for sure. I’ve always been fascinated with the concept of de-extinction.

VentureBeat: I’ve been fascinated lately with the intersection of things like entertainment, technology, and science fiction.

Lamm: We need wins, right? We need people to be excited about technology. We live in a 24-hour news cycle. We’re not reaching kids that much. If we can inspire the next generation, that’s great.

VentureBeat: Is there a way that you can also profit from this? What is the ultimate business you think Colossal Biosciences can enable so that you can keep on doing this?

Lamm: From a technology perspective, we’ve already spun out three companies. Two of them we’ve announced. One is Breaking, our plastic degradation company. Our first was Form Bio, our computational biology firm for human health care. We’ve done a good job of monetizing the technologies. We’re solving very complicated genome engineering tools and building very complicated solutions to reading ancient DNA. DNA is simply comparative genomics. Embryology and multiplex editing, being able to edit a lot of the genome at the same time. It’s a very powerful thing that we’re working on. It has hundreds of millions of dollars of economic value, in the technologies alone, for human health care. We can easily subsidize our work for conservation.

VentureBeat: Conservation gets the benefit of those profitable technologies.

Lamm: Right. All the technologies we make on the path to de-extinction, we make them available for free for conservation. In addition to that, we also launched the Colossal Foundation, colossalfoundation.org. In addition to the $435 million we raised for Colossal, we also raised $50 million for our foundation.

VentureBeat: At this point, then, does it feel like you’re in a self-sustaining enterprise?

Lamm: We have no problem raising capital right now. That’s a good thing. As long as that continues, we’ll be in a good spot.

VentureBeat: Have you gotten any interesting feedback? Whether it’s kids or–

Lamm: Every single week we get kids sending stories to us.

Shapiro: Lots of drawings.

Lamm: We get drawings from kids and parents. We get letters from teachers thanking us. They’re inspired. We get a lot of great feedback. We’re bringing more awareness to conservation. Any time you can inspire kids, bring back species that have cultural importance to indigenous people groups, and make technologies to save species from this mass extinction we’re currently in, where we’re going to lose up to 50% of biodiversity, it’s a massive win. I’m inspired by the combination of all that.

Shapiro: I’m just excited to finally be able to tell people about this.

VentureBeat: How long was it a secret for you?

Lamm: It was 18 months to October, and then another six months.

Shapiro: A two-year secret.

VentureBeat: Was there ever any government crossover with this?

Lamm: The Secretary of the Interior, Doug Burgum, just put out a statement on X about us. They’re endorsing our work on dire wolves, as well as acknowledging that de-extinction is critical technology for conservation. We’ve been educating the Department of the Interior on the power of these technologies for conservation.

VentureBeat: How do you keep the dire wolves surviving from here, and not going extinct again?

Lamm: We have them in a managed facility, that 2,000-acre expanse of ecological preserve. They live there with 10 full-time caretakers. Beyond that, any rewilding project–if you look at Yellowstone wolf rewilding, or some of the stuff that’s happening in Europe, or some of the world we’re doing with the Tasmanian tiger in southern Australia and Tasmania, it’s a very stage-gated process. They will be under some form of managed care. It may take five to 10 years to fully reintroduce them back into the wild in a managed way.

VentureBeat: What else looks promising, or what else is coming next for you?

Lamm: We’ll continue to work on our three flagship projects: the woolly mammoth, the Tasmanian tiger, and the dodo. We’re making tremendous progress. Based on the success of the dire wolf, we’ll probably look to add additional species on the avian side, as well as on the mammalian side, over time.

VentureBeat: How are we doing at finding fossils of everything now?

Lamm: They’re not really fossils, because fossils are rocks. But there’s actually a decent amount of ancient DNA, in various forms. It’s constantly a roll of the dice to see what we get. Occasionally we get nothing. Occasionally we get a lot more than we expected.

VentureBeat: Do you ever find anything in amber, like in Jurassic Park?

Lamm: Amber is actually a terrible material for preserving DNA. It’s very porous.

VentureBeat: Did you have to attempt more than one generation to get to the three dire wolves?

Lamm: We were done in one generation.

VentureBeat: So it’s not as if any attempts started and died.

Lamm: No, no. We spent a lot of time on the computational analysis to identify the edits. We spent a lot of time on genetic engineering. We spent a lot of time on quality control and sequencing on the back end.

VentureBeat: What kind of hardware was behind the computation?

Lamm: We use a bunch of computational analysis with a bunch of external partners for compute, some cloud and some internal. We’ve built a lot of AI models. Without the intersection of these synthetic biology tools, AI, and access to compute in the cloud, this would be an impossible project. A lot of it’s human thinking, though.

One thing we have, which is in the pictures, is a laser-assisted–this is some of the technology we had to build. We built a laser-assisted somatic cell nuclear transfer system. We use lasers to drill holes in the outer shell of the embryo so that it’s less hard on the DNA when we do DNA extraction and the insertion in the somatic cell nuclear transfer process.

VentureBeat: When you think of some of the hardest problems that are out there, where would you put this as far as the scale of difficulty?

Lamm: I’d argue that this is the moon landing of genetics. This is insanely hard. We took a 13,000-year-old tooth and a 73,000-year-old skull and we made puppies. We did it in 18 months. When you think about that–we understand what genes made a dire wolf a dire wolf. I said this in my quote, so I’m not trying to paraphrase, but it’s magic. It really is.

What’s funny is, we’re only a couple of years old. We just launched the woolly mouse, which was the precision germline edited animal, multiplex animal in the world until now. We’re scaling the technology up quickly. We’ll continue to hopefully make advancements that keep the world happy.

VentureBeat: Is there a species you can think of that, if you brought it back, could save the world?

Shapiro: The technology is there to save the world. There’s not a single species out there that would save the world, unless humans went extinct.

Shapiro: Ecosystem restoration. Stopping the loss of biodiversity, or at least slowing it. Using evolutionary innovation–we have a project with collaborators in Australia where we’ve taken a cell line from a quoll, which is an endangered little carnivorous marsupial. It’ll probably become extinct within the next 10 years without this technology. They eat cane toads, which were introduced to Australia, and they die from the cane toad toxin. We and our collaborators have made a version of the quoll that includes a single change to a protein sequence that evolved in an animal that lives on the other side of the planet that eats toxic cane toads. We put it in the quoll and they can eat the toxic cane toad and not die. It’s that sort of technology, that innovation–we’re discovering what a gene does, transferring that to a different species, and making an animal out of that.

Lamm: That one single nucleotide change, one letter in the genome, conferred 5,000 times the resistance to cane toad toxin.

VentureBeat: There was Dan Brown’s book Origin, which was about AI concluding that humans were too much of a threat to the Earth to allow them to survive. Maybe if the AI knew about you guys it might not have felt the same way.

Shapiro: I think we have a pretty poor track record, as far as being good or bad to the Earth as a species. Maybe we need more training data. This is the entry of some new training data into the model. Eventually we might not be the bad guys.