Science
Two genes, 120 million years, and why "evolution is just luck" gets it half right
Picture two insects 120 million years apart. A butterfly flitting through a Cretaceous fern forest, and a moth in some Eocene meadow long after the dinosaurs were gone — each wearing nearly the same loud warning colors, each painted by the very same two genes. Not similar genes. The same two: a pair called ivory and optix. Everyone "knows" evolution is blind luck, a lottery of lucky accidents. But the most striking thing about this story is that the luckiest-looking part isn't luck at all.
The same two brushes, reached for again and again
In 2026, researchers at the University of York and the Wellcome Sanger Institute published a finding in PLOS Biology that should unsettle anyone who thinks of evolution as a slot machine. Across butterflies and moths — lineages that split more than 120 million years ago — nature did not keep inventing fresh genetic machinery to make warning patterns and mimicry. It kept reaching for the same two tools. ivory and optix, plus the little on/off switches that govern them, show up over and over, producing strikingly similar colors in creatures that have not shared a recent ancestor since before flowers ruled the world. The genome, it turns out, behaves less like a casino and more like a toolbox — and evolution keeps opening the same drawer.
What evolution actually changed (and what it left alone)
Here is the detail that flips the cliché. Evolution did not rewrite ivory and optix each time it needed a new pattern. The genes themselves stayed remarkably constant. What changed was the wiring around them — when and where each gene gets switched on. This is regulatory evolution: same proven tool, new instructions for using it. It's the difference between forging a new hammer for every nail and simply learning fresh ways to swing the one good hammer you already own. And it isn't a quirk of insects. Dolphins and fish, separated by a vast gulf of ancestry, independently arrived at the same streamlined torpedo body. Birds and bats both built wings without a winged common ancestor. This is convergent evolution, and it hints at something heretical to the "pure chance" crowd: evolution may be more predictable than a lottery has any right to be.
The half people mistake for the whole
So is the popular story simply wrong? Not entirely — and that's exactly where it goes off the rails. Mutation, the raw supply of variation, genuinely is random. Copying errors fall where they fall. But isn't mutation random? Yes — and that's precisely the half people mistake for the whole. Because what happens to those mutations next is the opposite of random. Natural selection is a filter, and a ruthless one. Think of it this way: mutation deals you a random hand of cards, but selection is a very non-random poker player who keeps folding the junk and holding what works, hand after hand, for millions of years. The chance is in the dealing. The discipline is in the keeping. To call the whole game "luck" is to watch only the shuffle and ignore the player.
Dawkins, and the gene that outlives the butterfly
Half a century ago, Richard Dawkins gave us the lens that makes this study legible. In The Selfish Gene, his deepest move isn't the provocative title — it's a quieter line buried in an early chapter: stable things survive. Darwin's "survival of the fittest," Dawkins argued, is a special case of something more basic — the non-random persistence of whatever is more stable, more successful at lasting. Natural selection, in this framing, is not a dice-roll but a stabilizing filter, retaining the replicators that endure. And the unit that endures, he insisted, is not the animal. The butterfly dies; you will die; bodies are temporary "survival machines" a gene builds and then discards. The gene is the near-immortal replicator that rides on across the eons.
Read the 2026 result through that lens and it stops being a curiosity and becomes the gene's-eye view made literal. Ask Dawkins's signature question — who benefits? — and the answer is the gene. ivory and optix are exactly the kind of stable, successful replicators his theory predicts: tools so good they get selected again and again, persisting across 120 million years and countless bodies, while the vehicles they build flicker in and out of existence like sparks. The butterfly is the vehicle. The gene is what lasts. (A caution Dawkins himself would press: this is a metaphor, not a wish. The gene does not "want" anything, and "it evolved" never means "it is therefore good." Evolution describes; it does not prescribe.)
Selection is the opposite of random
Mutation supplies a random hand; selection is the non-random player who keeps folding the junk — which is why the same two genes can survive, unbeaten, for 120 million years.
What this means for you, the next time someone says "it's all chance"
You will hear, often, that evolution is "just random." Here's a quick field test for whether that claim deserves a nod or a raised eyebrow. First: are they talking about mutation or selection? Mutation is random; selection emphatically is not — collapsing the two is the single most common error. Second: are they arguing about how evolution happens or whether it happens? Working scientists argue fiercely about the how — the relative weight of regulatory tweaks, the predictability of convergence, the pace of change. They do not argue about the whether. Antievolutionists routinely smuggle the second debate in under cover of the first. And third, the Dawkins move: ask who benefits? and watch whether the answer points to a stable replicator quietly persisting. Do that, and a story that sounded like a casino resolves into something far stranger and more orderly — two genes, two brushes, painting warning colors across 120 million years, kept not by luck but by the patient, non-random hand that keeps what works.
Sources and honest caveats: the science draws on the 2026 PLOS Biology study led by the University of York and the Wellcome Sanger Institute, reporting that the genes ivory and optix (and their regulatory switches) have been repeatedly reused for over 120 million years to produce warning colors and mimicry in butterflies and moths — an instance of regulatory ("toolkit") evolution and of convergent evolution seen elsewhere (dolphins/fish, birds/bats). The framing leans on Richard Dawkins, The Selfish Gene — the gene's-eye view ("who benefits?"), "stable things survive," and genes as near-immortal replicators. The core correction — random mutation versus non-random selection, and the conflation of "how" with "whether" — is standard popular-science commentary. Note the caveats: avoid the naturalistic fallacy ("evolved" ≠ "good") and gene-determinism, and read teleological phrasing ("the gene wants") as metaphor, never literal intent.