We have often been surprised by lots of intelligent birds with their ability to fashion tools and solve complex problems
With bird intelligence while there is no easy defintion of animal intelligence that allows one to make direct comparisons, the consistent (seeming) over-performance of birds in both problem-solving and, in the case of New Caldonian crows, tool-using and tool-making, throws up some intriguing challenges to how we have traditionally understood brain size and organization to "intelligence". "Big brains" (especially large brain-to-body-mass ratios, what is known as the Encephalization Quotient (EQ)), are often a general basis for this comparison. Humans, for example, have a brain mass about 7.5 times what would be expected for an animal our size, comparable to dolphins, whales and elephants, animals that we generally acknowledge to have some kind of "higher" intelligence.
The evolutionary price that we pay for relying the big heads to hold these brains, though, is an stiff one: we are one of the very few mammals who routinely die in childbirth. This "cost" makes sense to us, though, as we feel that intelligence was subject to strong selction pressures for various early hominid groups. Birds, on the other hand, have brains only 2-3 times larger than what we would expect, based on their body mass, yet they consistently perform feats in the range of Chimpanzees (having substantially larger EQs) and by some measures do things that seem beyond Chimp capabilities. Human intelligence is probably based in very large brains for other than simple "problem-solving" reasons (most likely our complex social life, including being able to master group processes like empathy, politics and social sex). At the same time, very big-brained birds have an easier path to evolutionary emergence as egg-laying allows nearly any amount of head growth outside of the parent. Why doesn't such an obviously big-brained bird exist (if "intelligence" is so evolutionarily valuable) is an intriguing question. We have a lot more theoretical thinking to do about how tied is any idea of intelligence into the extended phenotype of the animal involved. It's good to see serious research moving beyond mammals in pursuit of this project.