Bringing Extinct Species Back to Life: Ancient DNA, Genetic Rescue, and the Future of Conservation

Bringing back a woolly mammoth sounds like science fiction. Using the same technology to help an endangered animal survive sounds like conservation. What makes this area of biology so interesting is that these are not separate endeavors. They depend on many of the same tools.

Ancient DNA allows us to learn about animals that disappeared thousands of years ago. Genome editing gives us ways to recreate particular traits in living cells. Advances in reproductive biology can turn some of those cells into animals. Together, these technologies raise an extraordinary question: Can we use what we learn from the past to build a more biodiverse and resilient future?

My discussion with evolutionary biologist Beth Shapiro explored that question from several directions, including mammoths, dodos, engineered wolves, endangered ferrets, and human medicine. The central issue is not simply whether we can make an animal that resembles an extinct species. It is what we are trying to accomplish, how we evaluate the risks, and who gets to decide.

Table of Contents

• What Is a Species, Really?
• Ancient DNA Reveals a Branching, Interconnected Human History
• Why Some Distinct Species Can Still Interbreed
• How Do You Choose a Species for De-Extinction?
• What Bringing Back a Woolly Mammoth Actually Means
• Why the Dodo Requires a Different Reproductive Toolkit
• The Dire Wolf Project: Recreating Traits in a Gray Wolf Genome
• Conservation Is Already an Intervention
• Preventing Extinction With the Same Tools
• Black-Footed Ferrets: A Case Study in Genetic Rescue
• Mosquitoes, Gene Drives, and the Risk of Doing Nothing
• Who Decides What Gets Released?
• Human Genetic Selection Raises Related but Distinct Questions
• Artificial Wombs: An Ambitious Goal, Not a Finished Technology
• Public Trust Requires More Than Enthusiasm
• Why Scientific Curiosity and Conservation Can Reinforce Each Other
• The Goal Is a More Resilient Future, Not a Perfect Copy of the Past

What Is a Species, Really?

Before discussing de-extinction, we need to examine a word that seems much more straightforward than it is: species.

Biology does not organize itself into perfectly labeled boxes. People create those boxes to describe evolutionary relationships, communicate clearly, and make conservation decisions. That does not make species meaningless. It means that different definitions answer different questions.

Taxonomy is the naming and classification system. For example, the American buffalo is a bison, and the plains bison has the wonderfully repetitive scientific name Bison bison bison. Common names can obscure relationships, whereas formal classification tries to make those relationships clearer.

Species concepts address a different problem: How do we determine whether two organisms belong to the same species?

• The biological species concept emphasizes whether organisms can reproduce together and produce fertile offspring.
• Genetic species concepts classify organisms using differences or similarities in their DNA.
• Geographic approaches distinguish populations according to where they live, which can be useful in conservation.

The biological species concept is a useful starting point, but hybridization complicates it. Some lineages that we give different species names can interbreed. Others become reproductively incompatible relatively quickly.

The important distinction is between the label and the biology. A label can help us ask a question. It should not substitute for understanding the organism.

Ancient DNA Reveals a Branching, Interconnected Human History

Human evolution is an excellent example of why simple categories can be misleading.

The familiar image of an ape gradually becoming an upright human captures broad changes in anatomy, but it can suggest a single, orderly progression. The actual history includes branching lineages, extinctions, migrations, and interbreeding.

Fossils provide essential evidence, but a fragment of jaw or tooth does not automatically tell us whether its owner belonged to a direct ancestral population or a lineage that left no descendants. Ancient DNA adds a different kind of evidence. A genetic sequence can be compared with sequences from ancient individuals and living people.

That approach helped establish that ancestors of modern humans interbred with Neanderthals. It also revealed Denisovans, initially identified genetically from a small bone found in Denisova Cave in Russia. These were not simply isolated branches that never interacted.

For additional background, the Smithsonian’s human origins genetics resources explain how genetic evidence contributes to our understanding of human ancestry.

Neanderthal ancestry is not the same in everyone

Beth emphasized a subtle but important point: Two people can carry similar amounts of Neanderthal ancestry without carrying the same pieces of Neanderthal DNA.

She described estimates of roughly 2% to 5% in the populations under discussion. Those percentages should not be interpreted as a universal value for every person or population. The broader lesson is that ancient ancestry is distributed unevenly, both among people and across the genome.

Collectively, living people retain far more of the Neanderthal genetic legacy than any one individual does. Conversely, regions where Neanderthal ancestry is absent or unusually scarce can help researchers investigate which genetic differences mattered for reproduction, development, and survival.

None of this means that people with different ancestry percentages belong to different living human species. It tells us that the history of Homo sapiens includes interbreeding with related ancient lineages.

A closer comparison helps identify meaningful differences

Before researchers had Neanderthal genomes, investigating what genetically distinguishes humans often required comparisons with chimpanzees or bonobos. Those comparisons span millions of years of evolutionary change.

A Neanderthal genome provides a much closer comparison. It reduces the evolutionary distance and helps narrow the set of genetic differences that might be particularly important to modern human biology.

That matters because most differences in DNA are not a simple explanation for a distinctive trait. Mutations continually arise through copying errors. Some affect function; many do not. Ancient genomes help researchers separate potentially meaningful changes from the enormous background of ordinary genetic variation.

Ancient variants can have benefits and costs

Neanderthal ancestry is sometimes presented as a collection of “vigor genes.” That is much too simple.

Beth discussed ancient variants associated with immune function, pain sensation, and susceptibility to conditions such as type 2 diabetes. During the COVID-19 pandemic, researchers also identified Neanderthal-derived variants associated with different disease outcomes, including increased risk and protection.

A variant that helped people survive one pathogen in the past might create vulnerability to another pathogen today. Evolution does not optimize for every possible future environment.

Pigmentation offers another illustration of what ancient DNA can reveal. Variants in genes such as MC1R provide clues about hair color in ancient populations. But those findings do not justify sweeping claims that red hair, Neanderthal ancestry, pain tolerance, and general toughness are all part of one package.

Our discussion also touched on eye color and the possibility that mate preferences contributed to its diversity. That remains a different kind of explanation from showing that a particular eye color directly improved survival.

Why Some Distinct Species Can Still Interbreed

When populations become isolated, mutations accumulate independently. Different environments favor different variants, and chance can have a substantial effect, especially in small populations.

Eventually, some changes may prevent successful reproduction. A sperm might no longer fertilize an egg, or a hybrid embryo might fail to develop normally. But reproductive incompatibility does not appear on a fixed timetable.

Brown bears and polar bears illustrate this beautifully. They are adapted to very different environments, yet they can interbreed. Ancient DNA has revealed a history of genetic exchange between their lineages.

Beth described work involving an ancient Alaskan polar bear and the brown bears of southeastern Alaska’s ABC islands. These studies show that bear ancestry has crossed boundaries that species names can make seem absolute.

However, producing a living hybrid is not the same as producing an animal that can thrive in either parent’s habitat.

One hypothesis discussed was that hybrids with less suitable coat coloration might struggle to hunt seals successfully in polar bear habitat. Under that scenario, the barrier is ecological rather than a failure of fertilization. The animal can be born, but its traits may fit one way of life better than another.

Reproductive compatibility, genetic ancestry, and ecological identity are related, but they are not interchangeable. That distinction becomes especially important when discussing de-extinction.

How Do You Choose a Species for De-Extinction?

Choosing a de-extinction target requires more than identifying an impressive animal. Beth organizes the problem around technical, ecological, ethical, and social considerations.

1. Is usable DNA available?

DNA begins degrading after an organism dies. Ultraviolet light damages it. Freezing and thawing can contribute to fragmentation. Microbes break down tissues as they recycle the organism’s carbon and nutrients.

Living cells maintain and repair their DNA. After death, those energy-dependent repair processes stop.

Cold, stable conditions can preserve DNA much longer than warm, wet conditions. The comparison is familiar: Food deteriorates differently on a sunny counter than it does in a freezer.

Beth described mammoth DNA recovered from bones approximately one to two million years old. That is extraordinary, but it is nowhere near the more than 66 million years separating us from the extinction of non-avian dinosaurs.

We do not have recoverable dinosaur DNA suitable for this approach. A fossilized skeleton is not a preserved library of intact genetic instructions.

The dodo presents a different preservation problem. It lived on Mauritius, where heat and moisture are unfavorable for ancient DNA. Beth described testing many island bones without obtaining what she needed. A useful dodo genome instead came from a specimen taken to Europe and preserved in a museum collection.

Multiple genomes are especially valuable because they help distinguish traits shared across a lineage from differences unique to a particular individual.

2. Why did the animal disappear?

Recreating traits does not solve the original cause of extinction.

If the habitat is gone, the disease remains, or the same pressures would immediately eliminate a restored population, the project has not solved the conservation problem. It has simply produced an animal.

3. Is there an ecological role for it now?

Ecosystems have continued changing since an animal disappeared. They are not empty stages waiting for the original cast to return.

A credible project therefore asks whether restoring an ecological interaction could improve the resilience of a present-day ecosystem. That is a different goal from reproducing an exact historical scene.

4. Will the project create broadly useful tools?

A mammoth project develops techniques for placental mammals. A thylacine project addresses marsupial biology. A dodo project requires solutions for birds.

Those differences matter. A reproductive technique that works in one group may not work in another. Choosing diverse targets can expand the conservation toolkit rather than repeatedly solving variations of the same problem.

There is also the role of excitement. Mammoths and dodos generate curiosity, student engagement, and investment. Beth’s argument is that this enthusiasm can fund technologies that also benefit less famous endangered animals.

What Bringing Back a Woolly Mammoth Actually Means

The popular image of de-extinction owes a great deal to Jurassic Park: Recover ancient DNA, fill in missing pieces, and grow the extinct animal.

The actual approach Beth described is different.

Asian elephants are the closest living relatives of mammoths. Mammoths and Asian elephants already share highly similar genomes. Rather than building a mammoth genome from scratch, researchers compare ancient mammoth genomes with living elephant genomes to identify changes associated with mammoth traits.

The emphasis is on positions where mammoths consistently differ from elephants, particularly differences relevant to the characteristics researchers want to recreate.

For a cold-adapted elephant, a woolly coat is not simply a matter of switching on longer hair. Supporting that coat can require changes in skin structure, follicle density, glands, and other interacting systems.

And the animal has to remain healthy. Elephants have approximately 22-month pregnancies and take many years to reach reproductive maturity. Mistakes cannot be treated casually.

Beth’s goal is not an exact genetic copy of one historical mammoth. Individual mammoths themselves differed genetically. The stated objective is an elephant with mammoth-like traits that could function in appropriate habitats today and in the future.

That last point is critical. A living animal must contend with present-day food, pathogens, microbes, and developmental conditions. It cannot survive in a reconstructed genetic past while inhabiting an entirely different biological present.

Why large herbivores might matter in the Arctic

Large herbivores do more than consume plants. They move nutrients, disperse seeds, disturb soil, and change vegetation through their movement and feeding.

At Pleistocene Park in northeastern Siberia, researchers have been studying what happens when large herbivores are restored to a northern landscape. Animals discussed included bison, horses, deer, and musk oxen.

One proposed mechanism involves snow. Snow insulates the ground. Animals searching for food disturb and compact that snow, changing how the ground exchanges heat with the winter air.

Beth described a resulting mosaic of wetter and drier areas, different ground temperatures, and more varied vegetation. The larger idea is that animals can help create the conditions of their own ecosystem.

That does not establish that engineered mammoths will solve Arctic warming. It provides an ecological hypothesis to investigate, rather than a reason to make a spectacular animal with no defined purpose.

Why the Dodo Requires a Different Reproductive Toolkit

Birds present problems that mammalian cloning methods do not automatically solve.

The technique associated with Dolly the sheep, somatic cell nuclear transfer, uses the genetic material from a body cell and an egg capable of supporting early development. Accessing a bird’s egg at the equivalent useful stage is not straightforward.

That is one reason Beth wanted a bird program. Birds include many threatened species, and conservation needs technologies suited to their reproductive biology.

Bird genomes also contain technical challenges. They have small microchromosomes, and some birds possess germline-restricted chromosomes found in reproductive cells but not ordinary body cells. Reconstructing ancient bird genomes from short DNA fragments can therefore be particularly demanding.

What unusual reproduction teaches us

Our discussion included California condors that produced offspring without a male genetic contribution. This is parthenogenesis, not reproduction between two females.

Normally, meiosis produces eggs and sperm with one set of chromosomes each. Fertilization brings those sets together. In unusual circumstances, development can proceed with genetic material derived only from the mother.

That does not mean the same process would work normally in humans. Mammalian development includes genomic imprinting, in which the parental origin of particular genes matters. Having two maternal copies is not necessarily equivalent to having one maternal and one paternal copy.

Sex determination also varies across animals. Humans typically use an XX or XY system. Birds use a different chromosomal system, with females generally ZW and males ZZ. In some reptiles, incubation temperature influences sex.

The lesson is not that biology has no rules. It is that the rules must be understood for the organism in question.

The Dire Wolf Project: Recreating Traits in a Gray Wolf Genome

The engineered wolves have attracted enormous attention, along with disagreement about what they should be called.

Beth described animals created using 20 selected edits in a gray wolf genome, informed by ancient dire wolf DNA and intended to produce traits such as larger size, greater robustness, and a fuller, light-colored coat.

Her preferred definition emphasizes appearance and ecological function. Under a definition centered on evolutionary lineage or genome-wide identity, an edited gray wolf is not the same thing as recreating the extinct dire wolf lineage.

Being explicit about the method is more useful than allowing the name to carry the entire explanation: These animals were engineered from a living wolf genetic background to express selected extinct traits. They were not cloned from intact ancient dire wolf cells.

Animal welfare changed which edits were used

The coat-color decisions provide a concrete example of how the project approached safety.

Beth explained that ancient dire wolf sequences contained variants associated with light coloration. But changing corresponding regions in gray wolves raised concerns because related pigmentation pathways can also affect vision or hearing.

Rather than reproduce those ancient variants directly, the team used alternative changes associated with light coloration in living dogs and wolves, which they considered safer in that genetic background.

That is an important distinction. The aim was to recreate a trait, not to insist on an ancient sequence regardless of its effects in a different genome.

Living animals are not just collections of edits

The three wolves discussed were Romulus, Remus, and Khaleesi. Beth described them as large, wild animals, not oversized domestic dogs.

The two males differed in their willingness to approach people despite being genetically identical. Their behavior also reflected hand-rearing and social experience. Genetics matters, but it does not eliminate individual development or environment.

The animals were not being bred together because of their close relatedness. Beth described the possibility of another pack for further study, but no plan to release these wolves into the wild.

One reason is ecological: Gray wolves already face conservation pressures. Introducing another large predator that could compete with them would require a compelling justification, not merely the ability to produce it.

Conservation Is Already an Intervention

It is tempting to divide conservation into “leaving nature alone” and “interfering with nature.” In practice, that distinction often collapses.

Conservation programs decide which populations receive protection, where animals can live, whether they are vaccinated, how predators are managed, and which individuals reproduce in captivity. People have also moved plants and animals around the world for centuries.

Recognizing that history does not mean every intervention is wise. It means that nonintervention is not automatically a return to an untouched natural baseline.

Beth’s emphasis is resilience: Can an ecosystem withstand disease, environmental change, or the loss of one component? Biodiversity and overlapping ecological roles can provide redundancy, giving a system more ways to keep functioning when conditions change.

Predators can influence an entire ecosystem

The return of gray wolves to Yellowstone illustrates how restoring a predator can affect prey populations, vegetation, and other ecological relationships. Those effects can extend well beyond the predator itself.

Beth also discussed the thylacine, or Tasmanian tiger, formerly a top predator in Tasmania. She raised the hypothesis that restoring a missing predator could alter disease dynamics by removing vulnerable or sick animals.

This came up in relation to Tasmanian devil facial tumor disease, a transmissible cancer affecting devil populations. Whether a restored predator would meaningfully change that disease is a question for investigation, not an established outcome.

That distinction matters throughout de-extinction: A plausible ecological mechanism is the beginning of a research program, not proof that a release will work.

Preventing Extinction With the Same Tools

The strongest response to “Why not help living species instead?” is a concrete example of doing exactly that.

Northern quolls and toxic cane toads

Northern quolls are small carnivorous marsupials threatened by introduced cane toads. A quoll that eats a toxic toad can die.

Some mammals elsewhere possess genetic variants that make them resistant to related toxins. Beth described work by Australian collaborators introducing a targeted change associated with resistance and testing its effects in a laboratory system.

The proposed conservation pathway is straightforward in principle:

1. Identify the biological mechanism that makes another animal resistant.
2. Determine whether a corresponding change could protect the threatened species.
3. Make and verify the targeted edit.
4. Test function, animal health, and ecological consequences before considering deployment.

A promising laboratory result is not yet a restored wild population. But it shows how comparative genomics and precise editing could address a specific cause of extinction risk.

Red wolves and lost genetic diversity

Beth noted that Colossal announced red wolf cloning work alongside its engineered dire wolf announcement, although the extinct animal received much more attention.

The conservation problem involves a small endangered population and limited genetic diversity. Collaborators identified canids carrying substantial red wolf ancestry outside the main managed population. Cloning from such animals could provide an additional route to recovering genetic variation.

That work makes the overlap tangible: Techniques developed to edit and clone cells for a de-extinction project can also be applied to a living endangered lineage.

Black-Footed Ferrets: A Case Study in Genetic Rescue

Black-footed ferrets bring together nearly every major conservation issue in this discussion: habitat pressures, prey loss, disease, captive breeding, genetic bottlenecks, and cloning.

Efforts to control prairie dogs, their principal prey, contributed to the ferrets’ decline. The species was once thought extinct until a surviving population was discovered near Meeteetse, Wyoming, after a family dog killed a ferret.

That discovery led to renewed conservation work. When disease threatened the remaining wild population, animals were brought into captivity to establish a breeding program.

The program enabled releases into the wild, but the founding population was small. Over generations, breeding among relatives can reduce diversity and increase the chance that harmful recessive variants occur in two copies.

Frozen cells can preserve options

The genetic rescue effort involved the U.S. Fish and Wildlife Service, the San Diego Frozen Zoo, and the conservation organization Revive & Restore. Beth was clear that she was not personally responsible for the ferret project.

Stored tissue from an animal whose genes were not represented in the living breeding population provided an opportunity to restore otherwise missing diversity.

Using cloning, researchers produced Elizabeth Ann in 2020 from cells preserved decades earlier. Elizabeth Ann did not reproduce, but another clone from the same genetic line subsequently produced offspring.

In somatic cell nuclear transfer, an egg’s cellular machinery can reprogram the introduced genetic material to support development. It is not always necessary to first create a separate stem cell line. The National Human Genome Research Institute’s cloning fact sheet provides useful background on the process.

Biobanking can preserve genetic possibilities that a future population may urgently need.

Diversity alone does not solve disease

Black-footed ferrets remain highly vulnerable to plague. Restoring diversity does not automatically remove that threat.

Beth described research into the genetic basis of resistance in related animals. If researchers can identify a protective mechanism, editing might eventually complement cloning by helping ferrets survive the environment into which they are released.

These are different tools addressing different problems. Cloning can recover lost variation. Targeted editing might address a particular vulnerability. Habitat protection and disease management remain essential.

Low diversity does not always mean immediate failure

The Channel Island foxes offer an important complication. Some populations have extremely low genetic diversity yet remain viable.

Beth explained that a long history of small population size can expose harmful recessive variants to selection, allowing some to be purged. That does not make low diversity desirable. A genetically uniform population may still be vulnerable to a new disease or environmental change.

The useful lesson is that genetic diversity must be interpreted in the context of population history and present threats.

Mosquitoes, Gene Drives, and the Risk of Doing Nothing

Mosquitoes make the trade-offs especially clear. Some species transmit devastating diseases, but mosquitoes also participate in ecosystems. “Eliminate mosquitoes” is not a sufficiently precise conservation or public health plan.

Only certain species transmit particular human pathogens. Human-created standing water can also increase their abundance. In some settings, reducing an unusually large disease-carrying population is a different intervention from eradicating an entire ecological group.

Another possibility is to modify a mosquito’s ability to transmit a pathogen rather than eliminate the mosquito itself.

Gene drives are one potential tool for spreading a genetic change through a population. Beth discussed the possibility of designs with limited persistence and safety controls. She also emphasized an obstacle: If a drive reduces reproduction, variants that escape its effects can have a strong evolutionary advantage.

Gene drives are therefore neither magic switches nor inherently risk-free. Their behavior depends on design, population biology, resistance, and the surrounding ecosystem.

Could targeted suppression help restore native plants?

Invasive grasses provide another possible application. Beth discussed shallow-rooted grasses that dry rapidly and contribute to flammable landscapes in the western United States.

Temporarily suppressing an invasive population might give native plants a chance to recover. That is a proposed use for carefully controlled genetic tools, not evidence that a ready-to-deploy solution already exists.

Florida panthers show why stewardship can be ongoing

Florida panthers experienced severe inbreeding problems, including reproductive abnormalities. Introducing Texas panthers in the 1990s brought new genetic variation and improved the population’s condition.

But genetic rescue is not necessarily a one-time event. If a population remains small and isolated, the same pressures can return.

This is where Beth’s central ethical point becomes particularly important: Doing nothing is also a decision.

If habitats are changing faster than populations can adapt, rejecting translocation, assisted reproduction, or genetic tools does not preserve a neutral status quo. It can mean accepting a future with fewer species.

Who Decides What Gets Released?

The ability to make an animal is not permission to release it.

De-extinction projects operate within animal welfare requirements and regulatory systems. Any movement toward release also requires ecological assessment, staged testing, and participation from the people who will live with the consequences.

Beth described advisory groups that include conservation biologists, animal specialists, local stakeholders, and people with different economic and political interests. For the thylacine work, that includes a Tasmanian advisory panel.

She also described a moa project led by the Ngāi Tahu Research Centre in New Zealand, with decisions about species, numbers, and possible releases shaped by the people who would become long-term stewards.

That approach recognizes that technical expertise is necessary but insufficient. Scientific questions, cultural responsibilities, land use, and public trust all matter.

Models can inform decisions, not remove uncertainty

We discussed whether artificial intelligence could help construct digital twins of ecosystems: models that allow researchers to explore many possible interventions and outcomes.

Such models could integrate more variables than any one researcher could hold in mind. But their usefulness still depends on what is known, what is measured, and which interactions are missing.

Beth also described ecosystem assessment reports examining short-, medium-, and long-term consequences, including a report concerning the bluebuck. These assessments combine ecological evidence with conversations involving local stakeholders and specialists.

The aim is not perfect prediction. It is a better-informed decision, followed by monitoring and the ability to respond.

Human Genetic Selection Raises Related but Distinct Questions

The same conversation eventually reaches human biology.

IVF can involve testing embryos for particular genetic conditions. Some companies also offer predictions about complex traits. Those activities must be distinguished from editing DNA.

• Embryo testing and selection assess and choose among embryos that already exist.
• Somatic editing targets cells in an existing person to treat disease.
• Heritable editing changes genetic material in ways that could be passed to future generations.

These interventions differ in their uncertainty, ethical implications, and consequences for people who cannot consent.

Complex traits are not simple genetic settings

Height is heritable, and ancient DNA can help researchers study how variants associated with height moved through populations. Beth discussed the contribution of steppe ancestry to European population history and the interaction between genetic potential and environmental conditions.

Intelligence is much harder to reduce to a genetic prediction. What is being measured, how it is measured, and the environment all matter. Claims that an embryo can be confidently ranked for a future complex outcome should not be treated as equivalent to identifying a well-understood disease-causing variant.

Mate choice has influenced human genetic variation throughout our history. But that observation does not settle the ethics of technological selection. One concern Beth identified is whether one generation could constrain the opportunities or expectations of the next.

The edited babies controversy and medical gene editing

We discussed the widely condemned experiment in China involving embryo editing of CCR5, a gene relevant to HIV infection. It illustrated the gap between being technically able to intervene and having an adequate medical, scientific, and ethical justification.

A very different example was baby KJ, an infant with a severe urea cycle disorder. A collaboration developed a personalized CRISPR base-editing treatment designed to address the child’s specific mutation in the liver.

The early reported response was encouraging. It is more accurate to describe it as a landmark personalized treatment with ongoing follow-up than to assume that a short-term result proves a lifelong cure.

The contrast is useful: Treating a seriously ill child’s cells is not the same intervention as editing an embryo in a way that may affect future generations.

Artificial Wombs: An Ambitious Goal, Not a Finished Technology

Mammoth de-extinction also raises a practical reproductive question. Producing a large population by repeatedly using Asian elephants as surrogates would impose substantial demands on an endangered living species.

Beth described artificial womb research as one possible long-term response, with current work focused on model systems such as mice.

The aspiration is to support development without requiring an elephant pregnancy for every animal produced. That is not the same as already possessing a system capable of gestating a mammoth.

Potential medical applications help explain why the work could matter beyond conservation. A more advanced system for supporting fetal development might eventually help in circumstances involving extreme prematurity, fetal surgery, or a pregnant patient who urgently needs treatment incompatible with continuing the pregnancy.

Neonatal intensive care already demonstrates how technology can support very premature infants. An artificial womb would require additional capabilities and a much deeper understanding of development.

The promise is substantial. So are the biological and ethical questions.

Public Trust Requires More Than Enthusiasm

Fear of a new technology is not evidence that people are irrational. People are trying to understand consequences they cannot directly observe, often in areas where institutions have not earned universal trust.

We discussed how familiarity, delivery methods, and communication can change public acceptance. A technology that initially feels intrusive may become ordinary once people understand its purpose and see reliable benefits.

But familiarity is not proof of safety. Better communication must include limitations, uncertainties, and dissent, not simply reassurance.

Beth described a mix of peer-reviewed papers, preprints, public talks, documentaries, and shorter educational material. Her publication record offers a route into the underlying research.

There is also a genuine role for awe. A mammoth, a dodo, or an engineered wolf can prompt someone to ask a serious scientific question for the first time. Students may be excited, skeptical, or angry. In each case, they are engaging with decisions that could shape the world they inherit.

Why Scientific Curiosity and Conservation Can Reinforce Each Other

Beth’s path into this work was not a childhood plan to recreate extinct animals. She began at the University of Georgia studying broadcast journalism and working in radio.

A cross-country geology and archaeology course changed the direction of her interests. Seeing glacial landscapes, national parks, and evidence of human impacts made her want to tell stories about environmental change.

She moved into science, studied parasitoid wasps in Panama, and eventually arrived at Oxford on a Rhodes Scholarship. There, ancient DNA brought together geology, paleontology, ecology, and storytelling in a field that was still new.

That history helps explain the appeal of the work. Ancient DNA is not merely a technique for reading old sequences. It is a way to reconstruct how ecosystems changed when climates shifted and people arrived.

The field’s early history also puts current excitement in perspective. In 1984, researchers recovered a small DNA fragment from preserved quagga tissue. Showing that DNA could survive after death was itself a major result, even though confirming the extinct animal’s relationship to zebras was not especially surprising.

From there came increasingly powerful ways to investigate extinct populations, evolutionary relationships, and historical environments.

The recurring dinosaur question is understandable. But the more consequential questions may concern a ferret that can survive plague, a quoll that can tolerate a toxic toad, or an endangered population that can recover lost genetic diversity.

The Goal Is a More Resilient Future, Not a Perfect Copy of the Past

The most useful way to evaluate de-extinction is neither to celebrate every engineered animal nor to reject the entire field as interference with nature.

Instead, ask specific questions:

• What biological trait or ecological interaction is the project trying to restore?
• What evidence suggests the intervention could help?
• How are animal welfare and unintended effects being evaluated?
• Who participates in the decision, and who will provide long-term stewardship?
• What happens if the intervention is not attempted?

We already shape the future of other species through habitat change, domestication, conservation, and medicine. The challenge is to become more deliberate and responsible about that influence.

Ancient DNA and synthetic biology do not remove the need to protect habitats or address the causes of extinction. They may give us additional tools when those efforts alone are not enough.

That is the possibility worth taking seriously: not simply making the past walk again, but using what the past teaches us to give living ecosystems more ways to survive.