Dave Hone: T-Rex, Dinosaurs, Extinction, Evolution, and Jurassic Park | Lex Fridman Podcast #480
Summary
T. rex’s durable advantage was an extreme package of scale, force and locomotor efficiency, not cinematic sprinting. At roughly 12 meters long, seven-plus metric tons and perhaps 40 kph only at the upper estimate, it was “an orca on land,” power-walking on spring-efficient feet while a 2.5–3-meter tail muscle drove each stride. Hone said some contemporaneous carnivores were only a little bigger than Velociraptor; Lex compared that ecological gap to lions alongside a weasel, while its likely prey base contained abundant juvenile megaherbivores.
The best-supported hunting thesis is lower-risk endurance predation on juveniles, combined with opportunistic scavenging. Modern and fossil evidence suggests carnivores typically take prey around 5–20% of their own mass; young dinosaurs were smaller, poorly armored, inexperienced and pushed into inferior feeding grounds. Healed tyrannosaur bites establish active attacks, while bites laid over weathered bone establish scavenging: “They’re obviously both.”
Paleontology’s core model risk is mistaking one preserved event for a species-wide strategy. A Microraptor containing one fish does not establish fish specialization, three parallel tyrannosaur trackways do not establish coordinated hunting, and several bodies in one quarry may have washed together. Hone’s preferred standard is converging, independent evidence—mechanics, isotopes, stomach contents, bite marks, tracks and living-animal analogues—paired with explicit alternative explanations.
The commercial fossil market is pricing celebrity and scarcity far more aggressively than scientific utility. Stan sold for $31.8 million, while the less complete Stegosaurus Apex later reached roughly $40 million; yet Sophie, described as by far the most complete known Stegosaurus, sold about a decade earlier for roughly £250,000, or $400,000. Hone’s reaction is essentially art-market logic: “It’s worth what people will pay for it,” with the 100-fold repricing difficult to explain through specimen quality alone.
The highest-return research capital may be infrastructure and repeated sampling rather than trophy acquisitions. Given $10 billion, Hone would spend perhaps half a billion on a first-rate museum, fund excavations from a durable endowment, and buy productive quarries so trained workers could send every find directly to public collections. His sharper point is that paleontological data are perishable—“our science is literally disappearing”—while drones, seismic sensing, radioactivity detection and medical scanners have not displaced people walking exposed rock.
Jurassic Park demonstrates how scientific authenticity can compound—or erode—an entertainment franchise’s cultural value. The first film broadly captured T. rex’s size and power-walking gait, but the franchise embedded inventions such as motion blindness, giant featherless Velociraptors and assumed pack hunting; later films even regressed from accurate models despite accuracy requiring “no more hassle.” The offsetting legacy is tangible: the Jurassic Foundation has funded dinosaur research and helped young scientists enter the field.
Evolution rewards scale and specialization during normal conditions, then reverses those gains under regime change. Bigger animals gain efficiency, dominance and dietary reach, but require more land and food, reproduce slowly and carry less population-level genetic diversity; the asteroid-driven climate shock therefore hit large terrestrial dinosaurs especially hard. Birds—10,500–11,000 living dinosaur species—show that survival came through small-bodied lineages, while T. rex shows why the locally dominant incumbent can still be the most exposed when the environment resets.
Deep dive
1. T. rex was an orca-sized predator operating on land
Hone begins with scale that museum mounts flatten into abstraction: T. rex was roughly 12 meters long, 4.5–5 meters to the head and seven-plus metric tons. His downstairs skull cast is not unusually large, yet he could fit through its mouth; colleague Tom Holtz’s compression is exact: “an orca on land.”
That makes a large savanna elephant, at five or six tons, lighter than the biggest T. rex—and the dinosaur was both bipedal and carnivorous. A 200-kilogram lion is not remotely comparable; the operative impression would be “just how absolutely vast they are.”
Hone’s zoo analogy explains why public intuition stays weak. Elephant houses, doors, barriers and food are all built to elephant scale, while mounted dinosaurs sit beyond rails and above platforms; only underneath one does the reference point click: “the foot finishes at my knee.”
Its recognition is similarly oversized. Alongside lions, elephants, giraffes, tigers, hippos and rhinos, T. rex is among the few animals almost everyone can name, so it “just dominates conversations” for dinosaur and tyrannosaur researchers.
2. The skull was built around crushing force, while the arms were fading away
T. rex carried a boxy, heavily reinforced skull with forward-facing, tennis-ball-sized eyes and teeth thick enough to withstand enormous loads. A similarly long Carcharodontosaurus tooth is only about one-third as wide; T. rex combined that thickness with “this uber-powerful bite” capable of driving through bone.
The arms tell the opposite story. Their deltopectoral crests were modest, radius and ulna thinner than a human’s, claws less curved than those of many theropods, and ligament pits exceptionally small for a tyrannosaur—all signals that the two functional fingers were “not doing very much at all.”
Scale resolves the visual confusion: the arm looks dangerous beside a person, but not on a seven-ton animal whose individual teeth approach finger size. A sometimes-preserved extra metacarpal is not evidence of a working third digit.
3. Locked feet and a massive tail made bipedal gigantism efficient
Tyrannosaur metatarsals formed an interlocking arrangement in which the middle bone narrowed and was braced by its neighbors. The same feature evolved independently in ornithomimosaurs and other groups; by preventing the foot from splaying, it redirected each landing into ligaments and cartilage that returned a little energy “every single step.”
The real engine sat in the tail. A huge muscle across its first third to half—perhaps 2.5–3 meters long in T. rex—attached to the femur; when it contracted, the planted foot stayed still and the body moved forward. Humans use the butt for comparable work, while dinosaurs used “a giant set of muscles” behind the hips.
Detailed speed estimates have moved repeatedly, but Hone cites roughly 25 mph, or 40 kph, as a recent upper bound and expects somewhat less. T. rex probably power-walked rather than entering a gait with both feet airborne, yet a four- or five-meter stride still covered ground rapidly.
Lex’s robotics analogy lands because the problem is energy management, not merely top speed. Hone adds that cinema itself began with biomechanics: Eadweard Muybridge’s triggered photographs resolved a bet about whether a galloping horse became fully airborne, then became a moving photographic sequence.
4. A human might survive because T. rex was cautious—or because we were uneconomical prey
A T. rex had no learned category for a primate-like animal. Hone notes that animals isolated from humans sometimes approach them without a response; a tyrannosaur might find people unfamiliar, threatening or simply “off-putting,” while basic animal caution could delay an attack.
Prey economics offers a second possibility. Hone’s work on Microraptor and modern carnivores places typical prey around 5–20% of predator mass; at the lower end, a human may be beneath the worthwhile range of a seven-ton adult, though a one-ton subadult would present a very different calculation.
Motionlessness would not make someone invisible: “It’s nonsense. They can see really well.” If one approached in open ground, Hone’s desperate option would be an accurately thrown rock; in woodland, a person could pivot around a large tree faster until the animal lost interest.
5. Juvenile herbivores offered the best risk-adjusted prey
Documentaries favor an adult Triceratops showdown, but a predator injured by meter-long horns, a crushed foot or a body slam may never hunt again. Even killing a five-ton hadrosaur is inefficient because one T. rex cannot consume the carcass before much of it rots.
Fossils align with lower-risk targeting: swallowed remains and healed bite wounds from large theropods repeatedly involve young animals. “Juvenile” need not mean hatchling—a one-ton Triceratops could still be rhino-sized—yet it remained smaller and less formidable than the adult.
Young animals were vulnerable along several dimensions at once: incomplete horns, frills or armor; limited knowledge of predators; inefficient foraging; and higher food requirements relative to size. Adults could also displace them from the best patches into marginal areas or the forest edge “where the T. rexes hide.”
Hone treats this as a broad ecological rule of thumb, not dinosaur exceptionalism. Comparable patterns run from fish, starfish and mantises through crocodiles and large cats: juveniles are “dumb, but they’re inexperienced,” and direct dinosaur evidence fits that expectation.
6. Bite marks only become behavioral evidence after reconstructing what happened to the corpse
Hone’s governing warning is that fossils “basically can’t” be taken at face value. Several theropods in one quarry might have lived together, but fish scales, debris and bone orientation may instead show that water transported separate bodies into one channel.
A Mongolian hadrosaur provided unusually clean sequencing. Erosion marks lay beneath tyrannosaur bites, meaning the carcass died, washed downstream, weathered on a sandbank for days or perhaps weeks, and was bitten later: “It pretty much can’t have happened any other way” than scavenging.
Every bite occurred on the humerus. Deep punctures at the ends indicate the large side teeth removing bone, while close parallel scratches clustered along the muscle-bearing deltopectoral crest match the small, flattened front incisiform teeth pulling tissue away.
Hone’s analogy is scraping cream from an Oreo after taking off the top. The large teeth killed and dismembered; the small front teeth selectively fed—a distinction recoverable because the pristine remainder of the skeleton carried no comparable damage.
7. T. rex probably pursued prey over distance and scavenged whenever the opportunity arose
The energy-efficient foot and long stride point toward persistence rather than cheetah-like acceleration. Hone pictures something more wolf- or hyena-like: approach a juvenile prey animal, begin the chase with distance already closed, and exploit endurance even if the quarry is initially faster.
Open habitats make hiding a four-meter-tall predator difficult. Hone therefore suspects some combination of nocturnal activity, excellent smell and large eyes optimized for low light; darkness would let T. rex “sneak”—with the word heavily qualified—close enough to make pursuit viable.
Its hunting success need not have approached 100%. A one-ton Triceratops might escape often, but a tyrannosaur already near top speed after covering the first few hundred meters could keep enough pressure on an inexperienced animal to win sometimes.
The hunter-versus-scavenger debate is false: weathered bones bitten after death prove scavenging, while healed bones containing tyrannosaur teeth prove attacks on living prey. Hone still guesses active predation predominated because obligate scavengers require the ultra-efficient, long-range travel seen in soaring vultures and condors.
8. Dinosaur diversity remains radically undersampled
Hone places T. rex finds from Alberta down toward New Mexico; some tyrannosaurine teeth farther south may or may not be T. rex. Mongolia instead had Tarbosaurus, its exceptionally close relative; dinosaurs more broadly reached Antarctica even near the extinction boundary.
Paleontologists have named roughly 1,500–1,600 valid dinosaur species, acknowledging disputes around individual names. The field has nevertheless added around 40–50 species annually for at least 10–12 years, with “no signs of slowing down.”
Large geological inventories remain lightly explored. Hone cites India, newly recognized beds in Ecuador, fossil-rich Argentina and Australia as places where known potential greatly exceeds excavation effort.
The true historical species count is much harder. Preservation, rock exposure, species duration and sampling push estimates in opposite directions, leaving a range where a numerical guess might be “very accurate or very far out.”
9. Most fossil discovery still means walking the right rocks and noticing millimeters
Exceptional feather-bearing deposits are often commercially quarried lithographic limestones with extremely fine layers. Because a good fossil may occur only after hundreds of tons are split, researchers depend on quarry workers already processing the stone rather than assigning paleontologists to shift it all.
Elsewhere, the method is largely unchanged since the 18th century: identify rocks of the right age and depositional type, walk exposed surfaces and look. Surface skeletons rarely endure; wind-blown sand, moisture, freezing and porous bone rapidly reduce them to fragments.
Hone’s ideal specimen began as less than a centimeter of claw protruding from a northern Chinese hill. Excavation revealed a more than 90% complete Velociraptor relative, later named Linheraptor: “Every so often behind that is a whole skeleton.”
10. Excavation is an engineering problem governed by rock, access and transport
Technique begins with the relative strength of fossil and matrix. Soft Chinese or Mongolian sandstone can sometimes be cleared by hand, while fragile bone in hard rock rules out vibration-heavy equipment and may require slow chipping around every surface.
Preparators stabilize porous fossils with Paraloid, a consolidant that penetrates bone but can later be dissolved with acetone. That reversibility lets teams strengthen a specimen in the field without permanently locking future researchers into the treatment.
After tracing enough bones to infer the skeleton’s orientation, crews remove overburden and isolate a block. That can mean ten people with pickaxes, a backhoe in the desert or 20–30 feet of hill removed across a large area before the fossil layer is even reached.
Burlap, plaster and sometimes wooden beams turn the block into a transportable jacket. Where protected land bans wheeled vehicles, teams must subdivide it or shave away every safe gram to meet helicopter limits; a Stan-sized excavation can span months across three or four years.
11. Documentation converts an impressive skeleton into historical evidence
Grids, photographs and now drone photogrammetry preserve each bone’s original position. Long bones aligned with one another suggest current flow; absent small bones imply sorting; one anomalous element may belong to another animal washed into the same deposit.
Stan required more than 30,000 hours of cleaning, preservation, restoration and documentation. It is roughly 70% complete by bulk and 63% by bone count, with an exceptionally preserved skull that became a standard reference; discovered in 1987, it remained with the Black Hills Institute until about 2020.
Borealopelta, a roughly 112-million-year-old armored herbivore, preserved armor, skin, keratin sheaths and stomach contents despite being deposited miles out in an ancient sea. Preparator Mark Mitchell spent close to two years exposing it; color-pattern evidence tentatively supports a darker back and lighter underside, “give or take some very large uncertainties.”
Its adults resemble “nearly armored pine cones,” while juveniles appear much less armored. Hone infers that defense was real but probably not the armor’s original primary driver: if preventing predation alone explained it, the most vulnerable babies should have possessed it too.
12. Trophy-fossil prices have detached from completeness
Stan’s $31.8 million auction price shocked a market accustomed to major T. rexes selling for a few million. Lex noted that the result required more than one serious bidder to reach that level; the buyer was later identified as Abu Dhabi’s Department of Culture and Tourism.
Hone compares the market with art: beyond small fossils, a dinosaur “is worth what people will pay for it.” Apex, the Stegosaurus that subsequently broke Stan’s record at roughly $40 million, strikes him as large and good but not obviously singular enough to explain the premium.
Sophie in London is a young, smaller specimen, but Hone calls it “by far” the most complete Stegosaurus, missing only a limited set of plates and bones. It reportedly sold for around £250,000—perhaps $400,000—about a decade earlier, versus roughly 100 times more for the less complete Apex.
Size and public recognition deserve some premium, and T. rex sits above even famous names such as Stegosaurus, Triceratops and Diplodocus. Lex framed Apex as roughly 50% more expensive than Stan, but Apex was not a T. rex, making that comparison even harder for Hone to reconcile with specimen quality.
13. Research capital would compound through museums, land and open access
Given $10 billion restricted to dinosaurs, Hone would put perhaps half a billion into “the best museum you’d ever seen.” The purpose would be dual: communicate research publicly and provide stable, suitable storage so fragile fossils remain available for study.
A further billion could fund excavations and embedded researchers almost indefinitely if invested and spent through its returns. The remainder could acquire productive land and commercial quarries, including exceptional-preservation sites now controlled by miners or passive private owners.
His operating model would train every worker, reward discoveries and transfer specimens directly to museums. Researchers would no longer pay for access, and institutions would not need a new donor whenever an important fossil might otherwise end up in “some Silicon Valley billionaire’s foyer.”
The emphasis is not merely possession. Owning the source terrain would align search incentives, public preservation and scientific access—turning extractive sites into a durable sampling network.
14. Fossils are a perishable data stream that remote sensing has not rescued
Hone jokingly argues that all science funding should temporarily go to paleontology because black holes or panda genetics can still be studied later, whereas exposed fossils erode away. The serious kernel is stark: “Our science is literally disappearing.”
A unique giant could emerge from a hillside, weather for six months and vanish before anyone recognizes it. Researchers would then lose not only a marketable object but a biological observation that may never have been preserved anywhere else.
Technology has delivered less than Jurassic Park promised. Seismic “thumpers” do not reveal clean skeletons; drones struggle to match binocular human vision and head movement; unusually radioactive bones are detectable but not reliably discoverable from afar.
MRI and related imaging also falter because fossil bone absorbs minerals from its matrix and can approach the surrounding rock’s density. Paleontology lacks the funding to customize every promising technology, so for now “humans are quite incredible” at the essential task of looking.
15. Tyrannosaurs spent 100 million years turning a modest design into T. rex
Tyrannosaurs extend from roughly 160–165 million years ago to the extinction around 66.5 million years ago. The earliest known forms were two or three meters long, chest-high, long-armed and superficially similar to other small carnivorous dinosaurs.
Important traits arrived early: paired nasal bones fused into a rigid snout, flattened front teeth resisted pulling forces, and feathers are directly preserved. By the Early Cretaceous, Yutyrannus reached roughly six to seven meters and perhaps half a ton to a ton while remaining definitively feathered.
Later evolution split into long, narrow-snouted alioramins built for faster, lighter bites and robust tyrannosaurines. The latter line progressively enlarged the head and teeth, reduced tooth count and intensified bone-crushing power through Albertosaurus, Gorgosaurus, Daspletosaurus, Tarbosaurus and Tyrannosaurus.
16. Size created daily advantages and catastrophic tail risk
Cope’s Rule captures the broad tendency for lineages to grow. Some of it is diffusion from a small starting point, but larger animals can take more prey, travel more efficiently, win contests and dominate territories or mating opportunities.
The same strategy fails under extreme stress. Large animals need more food and land, occur in smaller populations and reproduce slowly; a mouse population breeding every eight weeks offers selection far more genetic variation than elephants breeding perhaps once every five years.
Dinosaur ecosystems supported several 10-, 20- or 30-ton herbivore species and terrestrial carnivores above one ton—far beyond modern land systems. T. rex was probably the largest expression, but the normal advantages of gigantism lasted only until the environment crossed a threshold it could not service.
17. T. rex was an arch predator in an ecosystem with almost no carnivore peer
Hone resists “apex predator” because ecology often uses it for predators of other predators, such as great white sharks consuming tuna or sea lions. Darren Naish’s alternative, “arch predator,” better describes an enormous top carnivore primarily eating herbivores.
Hone’s point was that T. rex’s niche was exceptionally lopsided even among giant tyrannosaurs: some contemporaneous carnivores were only a little bigger than the Velociraptor skull. Lex’s analogy was Africa containing lions, then “a weasel about this big,” with nothing substantial between them.
That gap does not mean T. rex routinely ate smaller carnivores. An adult Velociraptor-sized animal would be mouse-scale prey to it and probably not worth pursuing, while juvenile Triceratops, Edmontosaurus, Parasaurolophus and even sauropods offered far more return.
Against comparable land predators, Hone favors the heavier animal with the stronger bite: T. rex. Carcharodontosaurs likely grappled and slashed prey until it weakened; aggression might upset predictions—“the size of the fight in the dog”—but Velociraptors attacking T. rex would resemble “meerkats killing a lion.”
18. Jurassic Park’s inventions became public default settings
The motion-blind T. rex appears to be Michael Crichton’s fiction; Hone searched for the research hinted at in The Lost World and found none. The franchise similarly turned Velociraptor into a tall, cheetah-fast, highly intelligent, featherless pack hunter despite the real animal standing roughly thigh-high.
Some fundamentals were strong. The first film made T. rex approximately the right size and shape, and its Jeep pursuit shows a power-walking gait with one foot grounded; Jurassic Park III’s Pteranodons and its reconstruction of Spinosaurus also impress Hone.
Later entries regressed. Jurassic World’s Pteranodons could not grasp and fly away with people using their real feet, while its Gallimimus model was actively given teeth despite the animal having a beak—an edit barely visible during its brief appearance.
A Hollywood creature designer told Hone accurate revisions require about the same work as inaccurate ones: “It’s no harder to make it accurate.” The first film’s Jurassic Foundation partly repaid paleontology through research grants, but audiences should still treat the franchise like James Bond, not a documentary.
19. Group fossils do not establish coordinated pack hunting
The famous Deinonychus–Tenontosaurus association was interpreted as hunters beside their kill, but Hone asks the load-bearing question: “Well, why did they all die there?” Lions do not routinely die beside wildebeest; alternatives include transport, a predator trap or toxins accumulating carnivore bodies around prey.
A reported tyrannosaur “group trackway” amounts to roughly four or five prints from three similarly sized animals moving in similar directions. Game trails or males following a female hours apart can create that pattern without social contact, much less cooperative hunting.
Even true sociality does not answer the hunting question. Spotted hyenas live in complex clans and appear in dramatic group hunts, yet the scientific literature says they mostly hunt alone; fossil co-occurrence could show shared living without shared pursuit.
20. The gold standard for pack hunting would preserve interaction through time
Hone’s hypothetical decisive trackway would contain overlapping sequences: early prints from animal A lie over B, then later B lies over A. That reversal requires contemporaneous movement rather than animals passing minutes or hours apart.
If seven or eight such trackways converged on one herbivore and the surface then became chaotic, coordinated hunting would become highly persuasive. Repeated bite marks from differently sized attackers on disproportionately large prey could provide another supporting line, though tooth similarity complicates identification.
Deinonychus isotope and feeding studies may eventually strengthen its case; Hone is “not anti the idea.” His objection is transferring even credible behavior between relatives: lions hunt socially while leopards and tigers do not, and spotted hyenas differ sharply from other hyena species.
Crocodilians, iguanas and ground hornbills demonstrate that reptile- or bird-like brains do not preclude cooperation. Hone expects some dinosaurs were social and some pack-hunted; his honest conclusion about which ones is: “No idea.”
21. Sexing dinosaurs exposes how little a skeleton can certify
Eggs inside a body identify a female immediately, but only a few such specimens exist. Medullary bone offers a broader signal: egg-laying females temporarily grow highly vascular tissue in large bones so calcium can be mobilized rapidly into shells.
Its absence proves little. Males, juveniles, females outside breeding season, sick females and those that already laid eggs all lack it; the method identifies “laying female from everything else,” not female versus male.
Human pelvic sex differences are poor dinosaur analogues because they reflect childbirth through an unusually demanding birth canal. Horns, crests and body size can also mislead: maned female lions, maneless males and winter-antlered female reindeer show how variable living species remain.
Tyrannosaurs nevertheless carried nasal ornament and small horns without obvious mechanical roles. Hone groups likely mating, rivalry and communication functions under “socio-sexual selection,” because signals that attract mates may simultaneously intimidate competitors.
22. Honest signals convert biological cost into credible information
Dark-maned lions illustrate the trade: females prefer darker males and rivals find them intimidating, but black hair increases overheating risk near the equator. The signal remains honest because carrying the cost while surviving demonstrates condition.
Black swans provide a mutual version. Both sexes prefer curlier wing feathers, and females with curlier feathers tend to win nesting disputes; because the modified flight feathers make flying harder, they effectively announce, “Look how tough I am.”
Theropod crests may have worked similarly. Bright, oversized ornaments reveal a predator to prey, but an animal that still hunts successfully despite conspicuous red-and-yellow structures displays its quality; giant herd-forming herbivores face less need to hide and can push the same logic into enormous horns and frills.
23. Beauty can run away once preference and inheritance reinforce each other
The “sexy sons” mechanism does not require an initial survival benefit. If females arbitrarily prefer a novel color, daughters can inherit the preference while sons inherit the color, creating a feedback loop in which each generation becomes both redder and more attracted to red.
Wild female mollies preferentially choose male swordtails, which look familiar enough to register as mates but different enough to be exciting. Hone’s interpretation is that variation itself may be attractive because it carries genetic diversity useful against environmental change, parasites and disease.
Peacock tails probably mix both mechanisms. Their pigment, size and effect on flight impose a handicap, but extreme form and eye spots likely include arbitrary preference; “pure beauty” is therefore not fully separated from condition, inheritance and historical accident.
Sexual reproduction scales this option value across organisms producing dozens, hundreds or thousands of offspring. Humans naturally reason from few children, while many animals can afford vastly more variation and failed experiments.
24. Population samples matter more than isolated spectacular fossils
Signals in both sexes may hint at mutual sexual selection and cooperation in reproduction. Puffins, penguins, starlings and parrots need substantial contributions from both parents, so males as well as females benefit from choosing the strongest available partner.
Yet more than 90% of dinosaur species are known from one specimen, sometimes only a few bones or a tooth. Hone estimates fewer than ten species—perhaps five or six—have enough good skeletons for serious within-species comparison.
Protoceratops is the exceptional dataset: more than 100 good skeletons, perhaps 70–80 in accessible museums, spanning embryos and hatchlings through adults, mostly from one Mongolian locality and a window of roughly 100,000 years. That permits population-level questions about growth, ornament and dimorphism.
Hone’s 110-specimen gharial study showed why even that may fail: large females overlap small and medium males, leaving only the biggest males obvious. Drilling a Protoceratops mass-death assemblage for medullary bone might finally separate breeding females; scientifically, “another hundred Protoceratops” could beat 50 novel species.
25. Independent evidence is the antidote to seductive one-off stories
Microraptor specimens preserve a mammal, bird, lizard and fish in different stomachs. Each isolated paper could announce a specialist, but together they more plausibly describe generalists—while still leaving open that individuals learned different diets or scavenged unusual meals.
The strongest reconstructions ask whether mechanically and historically independent signals converge. One striking fossil is not a “silver bullet”; bone mechanics, habitat, isotopes, gut contents, bite marks and living analogues become powerful when they reach the same answer without depending on one another.
Lex’s vanished-human thought experiment clarifies the method. A healed compound fracture would imply months of care and therefore society; widespread treatment of major trauma and bone cancer would imply technology. As Hone puts it, dozens of ordinary skeletons may say little, while “the right one” reorganizes the inference.
26. Spinosaurus was probably a giant wader, not an underwater pursuit predator
Spinosaurs combine independent aquatic signals: crocodile-like skull mechanics, near-circular conical teeth suited to gripping wriggling prey, repeated association with watery settings, isotope signatures resembling fish and crocodiles, and fish scales inside the British Baryonyx. Hone cautions that water association is common in the fossil record because burial often occurs in aquatic settings, but argues the spinosaur pattern is stronger than that background effect.
Hone does not infer an exclusive fish diet—terrestrial dinosaurs and pterosaurs also carry relevant feeding evidence—but concludes the group lived differently from ordinary theropods. His shorthand for Spinosaurus is “a very weird, giant stork” or heron: water-associated and wading, but “not very” aquatic.
Spinosaurus may have reached 15 meters linearly while remaining lighter than T. rex, with a narrow body, elongated rosette-tipped jaws, recessed nostrils, sail, large arms, short legs and a thin paddle-like tail. Related Baryonyx and Suchomimus retained the head and arms without the full exaggerated package.
Hone and Tom Holtz “savaged” the strong-swimmer interpretation, and some original proponents later moved toward a non-swimming position. Against T. rex, Spinosaurus’s long weak jaw and neck poorly suited to rotation meet the tyrannosaur’s exceptionally strong neck and crushing bite: “No, I don’t buy it.”
27. Brains, cannibalism and extinction all reward conservative inference
Endocasts reveal relative olfactory and optic regions, while CT scans of the bony inner ear can estimate favored sound frequencies. Intelligence is harder: both brain volume and dinosaur body mass may be wrong by 20–30%, letting assumptions manufacture either a genius or a dullard.
A controversial estimate placed T. rex near three billion neurons; a rebuttal produced roughly 250 million–1.7 billion, closer to crocodilian expectations. Hone’s scaling check is blunt: most of the T. rex brain cast is smaller than a chimp’s, while the animal weighed seven tons—crocs are capable, “but they’re sure as hell not monkeys.”
Cannibalism rests on firmer evidence. Tyrannosaur bones carry tyrannosaur teeth or repeated feeding scrapes without healing; in a T. rex ecosystem, no other carnivore was large enough to make those marks. It was probably occasional because eating one’s own parasite-loaded species is riskier than ordinary prey.
The asteroid around 66 million years ago supplied the ultimate reminder. A roughly Everest-sized object traveling about ten times the speed of sound triggered rapid climate and ecosystem collapse; large terrestrial dinosaurs had maximum exposure, while some isolated populations may have persisted briefly—Hone says he would be amazed if none did—but were extraordinarily unlikely to rebuild global dominance.
28. Birds survived because dinosaur evolution had already diversified beyond giants
Birds are not merely descended from dinosaurs in a loose sense: “They literally are” dinosaurs, just as humans are apes and mammals. Roughly 10,500–11,000 living species remain, and their lineage coexisted with non-avian dinosaurs for perhaps 100 million years before the impact.
Feathers predated birds and occur in tyrannosaurs, dromaeosaurs, troodontids, ornithomimosaurs and other groups. Their likely early functions were dual: insulating warm-bodied animals and enabling communication through colors, erection, seasonal replacement and age-specific plumage.
Evolution’s scale is the mechanism. A female ocean sunfish may release up to 300 million eggs; across huge populations and millions of years, rare variations become inevitable. Evolution is not merely rolling dice—it “gets to keep the sixes.”
Yet the result is modification, compromise and repurposing rather than perfection. Swim bladders precede lungs, display tusks become digging tools, and one structure can fight rivals, deter predators and acquire food; organisms are “bodge jobs,” while selection keeps editing an inherited plan.