This moth, a member of the sphinx moth family, ranges from as far north as the Yukon to at least as far south as the Amazon. It's often mistaken for a hummingbird because it's wings beat so fast, and, like a hummingbird, it flits from flower to flower, hovering rather than resting on the flower while sipping nectar. Click on the 3rd photo and you'll see the curled tongue that the moth dips deep into the flower, as it hovers in the air taking aim.
The usual explanation for one species so closely resembling another -- and there are many many examples -- is that there's an adaptive advantage for the mimic in looking like the species it's mimicking, the model. Usually it's that the model species is poisonous, or tastes bad or has spines or something that makes a predator think twice about eating it, and thus by association about eating the mimic This is called "Batesian mimicry," after English naturalist Henry Walter Bates who explored the Amazon in the mid-1800's and was fascinated by all the mimics he saw, particularly among butterflies. Bates read his paper proposing his antipredator theory of mimicry to the Linnean Society in 1861 (the same society before which Darwin and Wallace's theories about evolution were first announced to the world in 1858). Cases where such suggested selective pairings have been identified have reinforced the generic explanation.
And indeed the hummingbird moth is often mistaken for a hummingbird -- Bates himself wrote in his book The Naturalist on the River Amazons in which he describes his decade of explorations along the Amazon, that he often had trouble distinguishing between a hummingbird moth and a hummingbird, often shooting the former when he thought he was shooting the latter. One can imagine scenarios whereby resembling a hummingbird is advantageous because hummingbirds have relatively few predators and so looking like one would perhaps be safer than looking like a tasty moth. However, hummingbirds do have predators, including other birds, snakes, even frogs and insects -- e.g., see this rather astonishing story of hummingbird as prey. Bates didn't think the moth's resemblance to the hummingbird was adaptive. He attributed it instead to the fact that they occupied the same niche, hovering before trumpet-like flowers to drink their nectar.
Maybe so, but then why don't bees or butterflies who sip at the same flowers resemble hummingbirds? And the moth survives in such a large geographic range that it clearly manages just fine in many different niches. It feeds on many flowers where hovering would not be necessary, and at least in those places there wouldn't necessarily be selective pressure either for hovering or mimicry. So, it's not clear that Bates was right that its ecological niche gave the moth its hummingbirdness.
The novelist Vladamir Nabokov was a very learned lepidopterist in his day and was himself intrigued by mimicry. He organized the butterfly collection at the Museum of Comparative Zoology at Harvard -- recently his theory that the Polyommatus blue butterflies came to the New World across the Bering Straits in five waves was confirmed genetically.
He, too, was interested in the similarity between hummingbird moths and hummingbirds, although like Bates he didn't believe there was a functional explanation. Nor did he believe the similarity was due to natural selection at all, but instead that it was incidental. Even accidental. And it's hard to argue with that.
Just because an adaptive scenario is plausible doesn't make it true. That is, just because an organism does something, and hence is adapted to the environment where it does it, is no proof that the behavior was the result of natural selection for that behavior -- nor, of course, is it an argument against such an hypothesis.
One might not think that the hovering these moths do was likely to arise by chance, but there are thousands of lepitdopteran species, and a lot of traits can arise even if rarely, by chance. Since the moths don't particularly look in terms of coloration like humingbirds, and in fact are variable, since all sorts of other color adaptations have been suggested for moths and butterflies, and since mimicry may be coincidental if hovering were the adaptive cause, we lack a sound explanation of their appearance.
At the same time, such doubts don't detract from the interest, much less the beauty, of this product of evolution, however it came about.