The thing it was eating — or one very like it — is the Giant African Snail.
If you garden in this city you know the snail, even if you never knew its name: Lissachatina fulica (the older books call it Achatina fulica). It is not a Chennai problem, or even an Indian one. It is one of the planet's most successful invaders, established in more than fifty countries — and almost none of that travel was its own doing. It is native to a thin strip of coastal East Africa. Everything after that was us. Naturalists carried it to India around 1847. It moved on through cargo holds and sea routes into Southeast Asia; Japanese merchants and soldiers spread it across the Pacific during the Second World War; the pet trade, the food trade, and accidental stowing-away in freight did the rest. Geneticists who trace its lineages find the heaviest diversity sitting on the heaviest trade routes. The snail didn't conquer the tropics. It hitchhiked on globalisation.
And where it lands, it wins — because it arrives somewhere with almost no natural predator. It eats over five hundred plant species, strips seedlings overnight, breeds prodigiously, and faces nothing that has learned to eat it. That last clause is the whole story. An invasive species is, almost by definition, a thing the local system has not yet worked out how to handle.
So when a coucal works out how to handle it, that is not a small thing. And it raises a question my daughters — one a microbiologist, one a wildlife scientist — made me ask properly, instead of romantically: do birds actually learn to eat new things over time, or did I just see a coincidence?
They do. It has a name — foraging innovation — and a serious literature behind it. Ornithologists have spent decades cataloguing cases of birds inventing new foods and new handling techniques, to the point where the rate of such innovations is used as a field measure of a species' flexibility and intelligence. The textbook case is British tits learning to pierce the foil caps of doorstep milk bottles from the 1920s, the trick spreading bird to bird across the country. Green herons drop bait on water to lure fish. And the pattern of who innovates is remarkably consistent: the bold, exploratory, less fearful individuals try the strange new thing first; once one cracks it, others learn by watching, and the behaviour spreads. Crucially, the species that can do this — fold a new food into an old skill set — are measurably better at surviving habitat change and at colonising new places.
In other words, my coucal was not a fluke. It was doing the most well-documented clever thing its kind does.
There is an even sharper parallel, and it is almost too on-the-nose. In Florida, a specialist raptor called the snail kite ate exactly one thing: a native apple snail. Habitat loss crashed the snail, and the kite collapsed with it — down to a few hundred birds, nearly extinct. Then an invasive apple snail arrived from South America, two to five times bigger. At first the kites fumbled it; researchers watched them drop the oversized snails, unable to handle them. Within about a decade the birds had adapted — bill size up roughly 8%, body size with it, the better-equipped birds surviving and passing the trait on, the population recovering into the thousands. A study in Nature Ecology & Evolution documented it as one of the fastest such changes ever measured in a long-lived predator: first through behavioural plasticity, then with the genetic machinery for real evolution falling into place.
The biologists who found this were careful about one thing, and I want to borrow their care. They were explicit that this is not a story about invasives being good. The snail is still a pest; the long-term effects on the wetland are unknown. What they had documented was narrower and stranger — that a system under pressure can find, fast, a use for the very thing that disrupted it.
I cannot stop seeing AI in this.
The dominant story we tell about AI is top-down and sudden: a capability arrives, leadership mandates it, transformation happens on a quarterly clock. That is not how the coucal learned, and not how the kite adapted. In both cases the capability already existed — a strong bill, an opportunistic mind. What was new was a resource sitting unused in the environment, and the slow, edge-first discovery of the match between the two. It began with a few bold individuals, low on fear, willing to try the strange thing. It spread because it worked.
The AI adoption that lasts will look far more like this than like the press release. The real value won't come from the transformation slide; it will come from a claims processor, a field agronomist, a junior analyst, a panchayat clerk — someone already holding hard-won judgement — quietly pointing an existing skill at a new substrate and finding the alien armour comes off.
And it is worth separating the two timescales the kite showed us. There is fast behavioural learning — individuals adapting within their working lives, which is what adoption looks like right now. And there is slow structural change — the organisational equivalent of evolving a bigger bill, where the institution itself reshapes around the new resource. Confusing the two is how leaders manage to both over-promise the first year and under-prepare for the tenth.
So if you own AI in your organisation, the instruction is not "deploy faster." It is closer to this: find your bold, low-neophobia foragers, put the capability within their reach, let them try things that aren't on the menu, and watch what spreads on its own. That emergent spread is your real signal — far more reliable than any mandate.
But here is the part that matters most to anyone who builds systems, and it is the part the feel-good version leaves out.
The Giant African Snail carries rat lungworm — Angiostrongylus cantonensis — a parasite that can pass to whatever eats it and, in humans, cause a form of meningitis. The coucal's clever new meal is genuinely useful and genuinely a little dangerous, at the same time, and the bird has no way of knowing which snail carries what. Emergent behaviour is powerful precisely because nobody had to authorise it. That is also exactly why it can carry something nobody signed off on.
That is the whole governance argument in a single image. The same unplanned, bottom-up adoption that lets AI fill gaps we never mapped is the adoption that can quietly ingest a hallucinated fact, a leaked record, an injected instruction, a dependency no one chose on purpose. "Nature found a way" is not the same as "nature found a safe way." Both are true of the coucal. Both will be true of us.
So I am not arguing for caution instead of emergence, or emergence instead of caution. I am arguing for the posture of a good naturalist — which turns out to be the posture of a good architect: enable the adaptation, watch it closely, and always know what the bird is actually eating.
The coucal will keep teaching my garden how to absorb the thing that invaded it. Our job is to make sure that when our organisations learn the same trick, we are paying enough attention to catch the parasite — and humble enough to admit the bird thought of it first.
A note on sources, because claims deserve provenance — and because the two kinds of claim here deserve different kinds of evidence.
The mechanisms rest on peer-reviewed work: foraging innovation and behavioural flexibility in birds (Lefebvre, Sol, Overington and colleagues); the rapid adaptation of the Florida snail kite to invasive apple snails (Fletcher et al., Nature Ecology & Evolution, 2018); and the human-trade-driven global spread and parasitology of Lissachatina fulica (genetic-haplotype studies; CABI Invasive Species Compendium; the rat lungworm, Angiostrongylus cantonensis*, for which the snail is a confirmed intermediate host).*
The specific sighting — a coucal eating this snail — is documented as field observations rather than a learning study: recorded in Rajasthan in 2000 (via CABI) and photographed in Southeast Asia in 2018 (Bird Ecology Study Group), with the snail's spread across Chennai and Tamil Nadu reported by Bombay Natural History Society-linked field surveys and regional press. The years-long arc I describe in my own garden is firsthand observation — the same humble unit of evidence from which the science of foraging innovation is built.
Comments