A recent study has shown that bumblebees, just like dogs and parrots, can be trained to perform complex "tricks"-- if you train them one step at a time. This sort of learning, where you build upon previously learned behaviors, is called "scaffolding," and bees are stellar scaffolders.

Scaffolding can make certain individuals seem much smarter than others. If one dog, Archie, has been trained over time to perform a series of complex manoeuvres and the other, Violet, has barely learned to "sit" (if you're lucky), it's tempting to brand Archie as a genius and bemoan the lack of smarts in Violet. But experience and past training play in heavily, and it's possible that the Violet is a secret Canine Einstein, with heaps of potential just waiting to be unlocked. When judging the cognitive capabilities of animals-- and humans-- we must remember that experience matters, and that potential is very difficult to isolate.

It is likely that humans have reached the extreme levels we have (building rockets to the moon and studying particle physics) due largely to scaffolding! And we are much better at scaffolding than other animals. Instead of just building upon our own experiences, we erect scaffolds resting upon the works of generations, using a clever innovation that I am making use of right now: written language. Most of us couldn't build a toaster, let alone a rocketship, without reliance on extensive written sources (and the advice of others). Remove our scaffolding, and just how smart are we?

Humans clearly have greater raw cognitive potential than any other animal on earth; there is no question about it. (And part of that potential is an inclination towards highly effective, inter-generational scaffolding). But if we are interested in the comparative cognitive potential between humans and crows, it would be a mistake to compare a rocketship to a bent piece of wire in a crow's beak and proclaim "we are INFINITELY smarter!" The story is far more nuanced-- and far more interesting-- than that.