Fast crows aren't smarter than us: they're a different design

Are crows smarter than us? The headlines flirt with yes: two birds, a touch screen and a decision time about twice as fast as ours. It is a great hook. It is also the wrong question, and the study itself points to a more interesting one.

Two crows and a touch screen produced a real, but narrow, result

Researchers at the University of Tübingen, Andreas Nieder and Philipp Schmidbauer, trained two male carrion crows, Iggy and Lenny, on a motion task. As the New York Post summarizes, dots on a screen drifted left or right among random decoys, and the birds pecked the matching side to earn a mealworm. Each choice took about 235 milliseconds on average, roughly twice as fast as humans and monkeys on the same kind of test, with perfect accuracy when the motion was clear.

The team also recorded 248 neurons in the nidopallium caudolaterale, the bird region that plays a role comparable to the primate prefrontal cortex. From that activity they could predict the choice about 170 milliseconds before the bird made it. The work is reported in Science, according to The Transmitter.

Speed is a property of the pipeline, not of the mind

Here is where we would slow the celebration down. Two crows is a tiny sample, and the task is one narrow kind of perceptual decision: which way is the cloud moving? Being fast at that is not the same as being good at deciding in general. This study does not show that a crow would beat a person choosing a career, or even planning a route around an obstacle it has never seen.

Nor is there yet a satisfying mechanism. Ars Technica notes that the team can't pin the speed on any single fast process, and Nieder's leading explanation is that nearly every biological process runs faster in birds than in mammals. If that holds, the result looks less like superior reasoning and more like a faster clock speed. The researchers point to the demands of avian life, where rapid visual processing and motor response matter for flight, quick beak movement and predator evasion. A flying animal that must dodge predators in real time is under strong pressure to react quickly.

That reading is an interpretation, and a plausible one, not a proven account. But it fits the pattern: speed is cheap to select for when your survival depends on it.

The real headline is the architecture

What we find most valuable is Nieder's remark that a brain does not need a mammalian neocortex to perform sophisticated computations. The Transmitter adds that crows last shared an ancestor with mammals more than 300 million years ago and have distinct neural structures. Two lineages, very different wiring, and yet comparable decision principles emerge. That is the interesting finding: convergence, not a leaderboard.

Intelligence is not a single dial with humans at the top and crows slightly ahead on one axis. Our brains buy us things this experiment does not measure, such as language, abstract planning and cumulative culture. We suspect those may come with some cost in speed, but that is speculation the study does not test. Crows seem to have taken a different trade, and whether their speed costs them anything elsewhere, we simply don't know from this study.

How to read the next "animal beats humans" headline

Our recommendation is simple. Ask what was measured, in how many animals, and whether the result was speed or quality. Here the answer is one visual task, two birds, and a striking difference in reaction time. That is good science, and an excellent reason to stop treating the primate cortex as the template for thought.

It is not a reason to crown crows. The better lesson is humbler and more fun: evolution has built at least two quite different machines that arrive at similar computations, one of them running at twice our speed. Smart is not a single thing, and the next experiments, with more birds and more varied tasks, should tell us how far the crow advantage really goes.