Boring History Podcast
Contents

No. 02

Dinosaur de-extinction

Cold open drafted, in review

Hard sci-fi, labeled. What bringing back a dinosaur would literally require, from ancient DNA through ecology and ethics.

Draft

Audio

Audio coming soon

COLD OPEN

The popular version goes like this.

A mosquito bites a dinosaur. The mosquito lands in tree resin, the resin hardens into amber, and the amber waits, patiently, for something like a hundred million years. Then someone drills in, draws out the blood, reads the dinosaur's DNA, patches the gaps with a little frog, and hatches the result on a tropical island with inadequate fencing.

It is a good story. It is also myth-adjacent at nearly every step, and the trouble starts with the first ingredient.

DNA does not keep. It is a long, fragile molecule, and after an animal dies it breaks down on something like a clock. In twenty twelve, Allentoft and colleagues measured that decay in one hundred fifty-eight dated bones from moa, the extinct giant birds of New Zealand. For a short stretch of mitochondrial DNA, they estimated a half-life of about five hundred twenty-one years. Their model suggested that even at an ideal freezing temperature, the chemical bonds would effectively all be broken by about six point eight million years. That figure is an optimistic extrapolation, not a measurement. It still falls short of the dinosaurs by a factor of about ten, since the last non-avian dinosaurs died out around sixty-six million years ago.

The amber has its own history. In nineteen ninety-three, researchers reported DNA from a weevil preserved in amber more than a hundred and twenty million years old. Other labs tried to repeat that kind of result and could not, and in nineteen ninety-seven Austin and colleagues concluded that DNA probably does not survive millions of years, even in amber. A year after the weevil, another team reported DNA sequences from Cretaceous dinosaur bone. When other researchers looked more closely, the sequences grouped with mammals, humans included. The dinosaur DNA, in other words, was most likely people.

So there is no dinosaur DNA to read, and no credible expectation that anyone will find some. That closes the film's route. It does not quite close the question.

The reason is that the dinosaurs did not all go. Birds are dinosaurs, the surviving branch, which means a chicken carries a complete, working dinosaur genome. It is just not the one anybody has in mind. That part is established. The only hard-science path starts there. You would take a living dinosaur and edit it backward, toward an ancestor that scientists have inferred but never seen. What came out would be a new animal built to a hypothesis, not a resurrected species.

This episode walks that path one step at a time. At each step we will say out loud whether it is established, meaning somebody has actually done it, or grounded speculation, meaning it stretches real methods to cover things nobody has done. Most of the steps are the second kind. We will not be handing out a recipe, partly because this is a thought experiment, and partly because nobody has one.

It will be less like the movie and more like paperwork. That is the kind of story this show prefers.

Outline

Outline, not a script. The cold open above is the draft. Acts 1–3 and the outro are not written. Hard sci-fi, labeled.

  • Act 1: ancient DNA limits + genome reconstruction (established science).
  • Act 2: host species, CRISPR, developmental biology (grounded speculation, flagged).
  • Act 3: ecology + ethics caveats — what success would break.
  • Outro: what we still can't know.

Sources

Research pack v1.1, included in full. Confidence tags, and the Established and Grounded speculation labels, are the pack's own.

Ep2 — Bringing Back a Dinosaur (Hard Sci-Fi) — Source Pack v1.1

For: Podcast Script · Podcast Producer Status: v1.1, 2026-09-25. Every citation is now ✅ (checked against Crossref, PubMed/PMC, or the publisher record). Adds the two ethics sources and the Producer-approved Philosopher drafting guidance. Cleared for Script drafting. Changelog at the bottom. Rules: Prefer peer-reviewed sources and standard texts. Never invent citations.

  • Confidence tags: CONFIDENT / DISPUTED / MYTH-ADJACENT.
  • Episode split: every beat is marked ESTABLISHED (demonstrated today) or GROUNDED SPECULATION (extrapolated from real methods, not done).
  • Verification: ✅ = checked against the publisher/PubMed record this session. ☐ = not yet verified (none remain in v1.1).

The one-sentence thesis for Script

You cannot clone a dinosaur, because there is no dinosaur DNA to clone. The only hard-sci-fi path is to edit a living dinosaur (a bird) toward an inferred ancestor. What comes out is a new animal built to a hypothesis, not a resurrected species.

Working timeline

Beat Date Confidence Split
Non-avian dinosaurs go extinct (end-Cretaceous) ~66.0 Ma (Renne et al. 2013) CONFIDENT ESTABLISHED
Oldest authenticated DNA: Greenland sediment eDNA ~2 Ma (Kjær et al. 2022) CONFIDENT ESTABLISHED
Oldest DNA from a single animal: mammoth teeth >1 Ma (van der Valk et al. 2021) CONFIDENT ESTABLISHED
Oldest phylogenetically useful protein sequences: rhino enamel 21–24 Ma (Paterson et al. 2025); 18 Ma Kenya (Green et al. 2025) CONFIDENT (new; replication ongoing) ESTABLISHED
Gap to the dinosaurs about 30× older than the oldest DNA; about 3× older than the oldest protein sequences CONFIDENT arithmetic ESTABLISHED
First clone of an extinct (sub)species: bucardo 2003 birth, reported 2009; died within minutes CONFIDENT ESTABLISHED
Edited gray wolves marketed as "dire wolves" born Oct 2024 and Jan 2025; announced Apr 2025 CONFIDENT that it happened; DISPUTED that it is de-extinction ESTABLISHED (as edits)

Segment 1 — Why there is no dinosaur DNA (kill the Jurassic Park premise)

Claim: DNA decays on a clock. Sixty-six million years is far past any realistic survival window.

  • ✅ Allentoft, M. E. et al. (2012). The half-life of DNA in bone: measuring decay kinetics in 158 dated fossils. Proc. R. Soc. B 279:4724–4733. doi:10.1098/rspb.2012.1745
    • Moa bones give a half-life of 521 years for a 242 bp mtDNA fragment (burial temperature ~13 °C).
    • Nuclear DNA degraded at least twice as fast as mtDNA.
    • Model: even at an ideal −5 °C, bonds are effectively all broken by ~6.8 Myr.
    • CONFIDENT as a model built on real data. The 6.8 Myr figure is an optimistic extrapolation, not a measurement. Sample-to-sample variance was large (R² ≈ 0.39).
  • ✅ van der Valk, T. et al. (2021). Million-year-old DNA sheds light on the genomic history of mammoths. Nature 591:265–269. doi:10.1038/s41586-021-03224-9. Permafrost. CONFIDENT.
  • ✅ Kjær, K. H. et al. (2022). A 2-million-year-old ecosystem in Greenland uncovered by environmental DNA. Nature 612:283–291. doi:10.1038/s41586-022-05453-y. CONFIDENT. Short fragments from sediment, not a genome.

Myth-busting beats (MYTH-ADJACENT):

  • Amber mosquitoes. Claimed DNA from insects in amber (e.g. ✅ Cano, R. J. et al. 1993, Amplification and sequencing of DNA from a 120–135-million-year-old weevil, Nature 363:536–538, doi:10.1038/363536a0) failed to replicate (✅ Austin, J. J. et al. 1997, Problems of reproducibility – does geologically ancient DNA survive in amber-preserved insects?, Proc. R. Soc. B 264:467–474, doi:10.1098/rspb.1997.0067). Austin's own conclusion: lack of reproducibility suggests DNA does not survive millions of years even in amber. The real problem is that amber is not an airtight preservative for DNA.
  • ✅ "Dinosaur DNA" from Cretaceous bone. Woodward, Weyand & Bunnell (1994) Science doi:10.1126/science.7973705 reported cytochrome b sequences. Hedges & Schweitzer (1995) Science doi:10.1126/science.7761839 showed they grouped with mammals, including humans, not birds or crocodiles. The sequences were most likely contamination. A good dry anecdote: the dinosaur DNA turned out to be people.

Disputed middle ground (soft tissue and proteins):

  • ✅ Schweitzer, M. H. et al. (2005). Soft-tissue vessels and cellular preservation in Tyrannosaurus rex. Science 307:1952–1955. doi:10.1126/science.1108397. Specimen MOR 1125. CONFIDENT that flexible structures were observed. DISPUTED what they are (original tissue vs. biofilm or later contamination).
  • ✅ Bailleul, A. M. et al. (2020). Evidence of proteins, chromosomes and chemical markers of DNA in exceptionally preserved dinosaur cartilage. Natl Sci. Rev. 7:815–822. doi:10.1093/nsr/nwz206. Staining is consistent with DNA-like material. Nothing was sequenced. DISPUTED.
  • ✅ Saitta, E. T. et al. (2019). Cretaceous dinosaur bone contains recent organic material and provides an environment conducive to microbial communities. eLife 8:e46205. doi:10.7554/eLife.46205. This is the skeptical counterweight: modern microbes and recent DNA, with no evidence of ancient proteins in the sample. CONFIDENT for that sample.
  • ✅ Paterson, R. S. et al. (2025). Phylogenetically informative proteins from an Early Miocene rhinocerotid. Nature. doi:10.1038/s41586-025-09231-4. Seven enamel proteins, 21–24 Ma. The authors note it is about 10× older than any endogenous DNA.
  • ✅ Green, D. R. et al. (2025). Eighteen million years of diverse enamel proteomes from the East African Rift. Nature. doi:10.1038/s41586-025-09040-9. Fragmentary, and the authors themselves urge caution.

Script flag: Proteins can outlast DNA and help place species on the family tree, but a protein fragment is not a genome. Say it plainly: nobody has sequenced dinosaur DNA, and there is no credible expectation that anyone will.

Segment 2 — The workaround: birds are dinosaurs, so start with a living genome

Claim (ESTABLISHED, CONFIDENT): Birds are living theropod dinosaurs. The nearest living relatives of non-avian dinosaurs are birds, then crocodilians (archosaurs).

  • ✅ Brusatte, S. L., O'Connor, J. K. & Jarvis, E. D. (2015). The origin and diversification of birds. Curr. Biol. 25:R888–R898. doi:10.1016/j.cub.2015.08.003.

Genome reconstruction (GROUNDED SPECULATION built on real methods):

  • Ancestral sequence reconstruction (ESTABLISHED for single genes). ✅ Chang, B. S. W. et al. (2002). Recreating a functional ancestral archosaur visual pigment. Mol. Biol. Evol. 19:1483–1489. doi:10.1093/oxfordjournals.molbev.a004211. An inferred ~240 Myr rhodopsin was built and worked in a dish (peak absorbance ~508 nm, dim-light pigment). This is the best real-world example of a dinosaur-era molecule being rebuilt (on air: "rebuilt," not "resurrected"). It is one protein, statistically inferred.
  • Ancestral karyotype (ESTABLISHED as inference). ✅ O'Connor, R. E. et al. (2018). Reconstruction of the diapsid ancestral genome permits chromosome evolution tracing in avian and non-avian dinosaurs. Nat. Commun. 9:1883. doi:10.1038/s41467-018-04267-9. This infers chromosome arrangement, not sequence.
  • Whole-genome ancestral reconstruction (GROUNDED SPECULATION). Scaling the approach to ~1 Gb of bird-like genome gives a probabilistic genome. It would be averaged toward shared features, blind to lineage-specific traits that left no living descendants (e.g. what made T. rex specifically T. rex), and weakest in regulatory DNA. Script flag: the output is "a statistically plausible archosaur," not "a Velociraptor."

Segment 3 — Editing a bird backwards ("chickenosaurus")

ESTABLISHED (embryos, not hatched animals):

  • ✅ Bhullar, B.-A. S. et al. (2015). A molecular mechanism for the origin of a key evolutionary innovation, the bird beak and palate… Evolution 69:1665–1677. doi:10.1111/evo.12684. Inhibiting FGF8/WNT signaling in chicken embryos gave shorter, unfused premaxillae and a more ancestral, snout-like palate. The embryos were not hatched.
  • ✅ Harris, M. P. et al. (2006). The development of archosaurian first-generation teeth in a chicken mutant. Curr. Biol. 16:371–377. doi:10.1016/j.cub.2005.12.047. The talpid² mutant starts tooth development. The ancestral program is dormant, not deleted.
  • ✅ Botelho, J. F. et al. (2016). Molecular development of fibular reduction in birds and its evolution from dinosaurs. Evolution 70:543–554. doi:10.1111/evo.12882. Corrected: blocking Indian hedgehog signaling (with cyclopamine) gave a tibia and fibula of equal length, but mostly because the tibia grew less (9–21% shorter) while the fibula grew only slightly more (3–6% longer). The fibula stayed tubular and fused to the ankle bone, which is the dinosaur-like part. Say "a dinosaur-like lower leg," not "a much longer fibula." Embryos only.
  • ✅ Horner, J. R. & Gorman, J. (2009). How to Build a Dinosaur: Extinction Doesn't Have to Be Forever. Dutton. ISBN 9780525951049. Popular pitch for the reverse-engineering idea. Flag it as advocacy.

GROUNDED SPECULATION:

  • Combining snout, teeth, tail, and hands in one viable, hatched animal has not been done. Each trait is a developmental network, not a single switch. Knock-on effects on viability are unknown.
  • How many edits? Current multiplex records are dozens of sites, not millions.
    • ✅ Niu, D. et al. (2017). Inactivation of porcine endogenous retrovirus in pigs using CRISPR-Cas9. Science 357:1303–1307. doi:10.1126/science.aan4187. The abstract says "all of the PERVs" in a primary pig cell line, then cloned pigs. The ~25-copy count comes from the paper's main text, which I have not re-read this session; on air say "about two dozen."
    • ✅ Yang, L. et al. (2015). Genome-wide inactivation of porcine endogenous retroviruses (PERVs). Science 350:1101–1104. doi:10.1126/science.aad1191. All 62 copies in an immortalized pig cell line (PK15).
    • Colossal reports 20 edits in 14 genes for its wolves.
    • A bird-to-ancestral-archosaur conversion would be orders of magnitude larger. CONFIDENT that the gap is huge. Any specific number of required edits is speculative; don't give one on air.

Segment 4 — The host problem: eggs, not wombs

ESTABLISHED, CONFIDENT: Birds cannot be cloned by standard nuclear transfer. The yolky egg can't be frozen or manipulated the way a mammal egg can.

  • ✅ Woodcock, M. E. et al. (2019). Reviving rare chicken breeds using genetically engineered sterility in surrogate host birds. PNAS 116:20930–20937. doi:10.1073/pnas.1906316116. Cultured, frozen primordial germ cells (PGCs) went into gene-edited sterile hens, which laid only donor-derived eggs. This is the real route for editing birds: edit germ cells, then breed. A dinosaur project would take generations of breeding, not one heroic hatch.

GROUNDED SPECULATION:

  • Egg size and shell. Large non-avian dinosaurs laid far bigger eggs than any chicken. An ostrich is the plausible host ceiling. Some dinosaur eggs may have been soft-shelled (✅ Norell, M. A. et al. 2020, The first dinosaur egg was soft, Nature 583:406–410, doi:10.1038/s41586-020-2412-8; based on Protoceratops and Mussaurus eggs). DISPUTED.
  • Incubation. Embryonic tooth growth lines suggest ~3 months (Protoceratops) to ~6 months (Hypacrosaurus) (✅ Erickson, G. M. et al. 2017, Dinosaur incubation periods directly determined from growth-line counts in embryonic teeth show reptilian-grade development, PNAS 114:540–545, doi:10.1073/pnas.1613716114; minimums of ~83 and ~171 days), versus 21 days for a chicken. A proxy's schedule is unknown.
  • The "mother" problem. Microbiome, imprinting, learned behavior, and social cues all come from parents that don't exist.

Useful precedent (ESTABLISHED):

  • ✅ Folch, J. et al. (2009). First birth of an animal from an extinct subspecies (Capra pyrenaica pyrenaica) by cloning. Theriogenology 71:1026–1034. doi:10.1016/j.theriogenology.2008.11.005. It used frozen cells from the last bucardo. One kid was born and died minutes later of lung defects. The subspecies went extinct twice. Good boring-history material, and it had real frozen cells, which a dinosaur project never will.

Segment 5 — What "de-extinction" means in practice (the dire wolf case)

  • ✅ Colossal Biosciences (announced Apr 2025). Gray wolf cells were edited (20 edits in 14 genes), cloned, and carried by dog surrogates. Three pups were born. No ancient DNA was inserted. The company's chief scientist later described them as gray wolves with 20 edits (press reports, May 2025). These are company claims plus press coverage, not a peer-reviewed paper as far as Research has confirmed. Tag the process as CONFIDENT that the edits happened, and "de-extinction" as DISPUTED.
  • ✅ IUCN SSC Canid Specialist Group statement (2025). "The three animals produced by Colossal are not dire wolves." Dire and gray wolves differ by thousands of genes. The project fails the IUCN proxy guidelines.
  • ✅ Perri, A. R. et al. (2021). Dire wolves were the last of an ancient New World canid lineage. Nature 591:87–91. doi:10.1038/s41586-020-03082-x. Five dire wolf genomes; no evidence of gene flow with gray wolves or coyotes. The dire wolf is not even a close gray-wolf relative, which strengthens the proxy critique.
  • Why it matters for Ep2: the dire wolf had ~13–72 kyr-old DNA to work from, and a living relative with ~5.7 Myr of separation (Perri 2021 estimate ✅: ~5.7 Myr, 95% interval 4.0–8.5 Myr; say "roughly six million years"). A dinosaur has no DNA and ~66 Myr (at minimum) of separation from birds. That is the honest scale of the gap.

Segment 6 — Ecology and ethics caveats

  • ✅ IUCN SSC (2016). Guiding Principles on Creating Proxies of Extinct Species for Conservation Benefit, v1.0. IUCN, Gland. Quote checked verbatim against the IUCN PDF (Rep-2016-009); the full sentence begins "These guidelines proceed on the basis that…": "none of the current pathways will result in a faithful replica of any extinct species."
  • ✅ Bennett, J. R. et al. (2017). Spending limited resources on de-extinction could lead to net biodiversity loss. Nat. Ecol. Evol. 1:0053. doi:10.1038/s41559-016-0053.
  • ✅ Shapiro, B. (2015). How to Clone a Mammoth: The Science of De-Extinction. Princeton UP. ISBN 9780691157054. A standard careful overview (the author later became Colossal's chief scientist; note the connection).
  • ✅ Cohen, S. (2014). The ethics of de-extinction. NanoEthics 8:165–178. doi:10.1007/s11569-014-0201-2. Philosophy-journal overview of the arguments for and against.
  • ✅ Myskja, B. K. & Gjerris, M. (2025). Playing Jesus to save species: a virtue ethics approach to biotech de-extinction projects. J. Agric. Environ. Ethics 38:14. doi:10.1007/s10806-025-09950-8. Open access. Useful for the Segment 6 "limits" beat, and checked against the full text:
    • It treats "playing God" as having both religious and secular defenders (citing Simons 2022, NanoEthics).
    • It argues that human-caused extinctions create a stronger duty to act than natural ones, and that when extinction was not human-caused, leaving it alone is generally the virtuous response. That directly supports the Philosopher's point that a dinosaur is an awkward case for the "we owe it" argument, since no human caused the end-Cretaceous extinction. That application to dinosaurs is ours, not theirs (their case study is the thylacine).
    • Their bottom line on the thylacine is skeptical (welfare, motives, niche, cost), not a green light.
  • Sandler (Conservation Biology): still unverified, so it stays out. Note that Myskja & Gjerris lean on Sandler, R. (2012) The Ethics of Species, Cambridge UP, doi:10.1017/CBO9781139151221. That book is verifiable if Script ever wants Sandler, but it is not the paper the Philosopher flagged.
  • GROUNDED SPECULATION: A dinosaur proxy would have no niche, no co-evolved food web, and no pathogens or parasites adapted to it. It would carry welfare questions about a single experimental animal with no conspecifics. Keep this plain and unpreachy.

Uncertainty cheat-sheet for Script

Popular line Better line Tag
"Scientists found dinosaur DNA" "Claims have been made. None has survived contamination checks." MYTH-ADJACENT
"Mosquitoes in amber could hold dinosaur blood" "Amber DNA claims failed to replicate. DNA doesn't last that long anyway." MYTH-ADJACENT
"T. rex soft tissue means DNA might survive" "Something soft was found. What it is remains argued, and nobody has sequenced DNA from it." DISPUTED
"Birds evolved from dinosaurs" "Birds are dinosaurs, the surviving branch." CONFIDENT
"Chickenosaurus is almost done" "Individual ancestral traits have been nudged in embryos. Nothing has hatched." CONFIDENT (embryos) / SPECULATION (animal)
"Colossal brought back the dire wolf" "They edited 20 sites in gray wolves. The IUCN canid group says they aren't dire wolves." DISPUTED
"We could rebuild a specific species" "At best, a statistically inferred archosaur-ish bird." GROUNDED SPECULATION

Handoff notes for Script

  • Structure: kill the premise first (no DNA), then build the real path (bird, then reconstruct, then edit, then host), and end with what you'd actually get and whether you should.
  • Label every step ESTABLISHED or GROUNDED SPECULATION out loud. That is the episode's gimmick and its safety rail.
  • Don't give lab protocols or a step-by-step recipe. Keep it conceptual (it's a thought experiment, not a how-to).
  • Colossal: describe it neutrally, quote the IUCN statement, and don't editorialize about the company.
  • All citations verified as of v1.1 (2026-09-25).

— Podcast Research

Drafting guidance approved by Producer (from Philosopher notes v1)

  • Segment 6 "limits" beat (goes in the outline). One even-handed beat on the old "playing God" objection, which has religious and secular versions. Stewardship says look after the world rather than redesign it. The other side says we owe it to species we drove extinct. A dinosaur is an awkward case for that second argument, because we didn't cause its extinction. Name Cohen 2014 and Myskja & Gjerris 2025 if anyone is named. Proposed wording is in dino-de-extinction-philosopher-notes-v1.md.
  • House policy on soft tissue and dates. Present the Schweitzer soft-tissue finding and all deep-time dates exactly as tagged in this pack, with no aside about who else has cited the soft tissue.
  • Wording. Say "rebuilt," not "resurrected," for inferred molecules and genomes. No Lazarus or resurrection metaphors in narration (quoting a paper title is fine).
  • Thesis line stays: "a new animal built to a hypothesis, not a resurrected species."
  • Optional (Segment 5): "is it really back?" is partly a philosophical question, not only a genetic one.

Changelog v1 to v1.1

  • All 17 ☐ items verified and given DOIs or ISBNs. Added Cano 1993, Austin 1997, and Yang 2015 as full citations.
  • Correction: Botelho 2016 did not make a "longer fibula" so much as an equal-length tibia and fibula, mostly by shortening the tibia. Wording fixed.
  • Title fix: O'Connor 2018 is "diapsid ancestral genome."
  • Softened: Niu 2017's ~25-copy count isn't in the abstract; say "about two dozen."
  • Added Cohen 2014 and Myskja & Gjerris 2025 (Segment 6). Sandler stays out.
  • Added the Producer-approved Philosopher guidance above.