Smart people actively seek evidence that they are wrong - finding evidence that you're right doesn't prove anything. A single piece of evidence that you're wrong shows everything.
There was a great Veritasium video on this while back. The schtick was that Derek Muller was posing a puzzle to people as a series of numbers. They had to figure out the rule governing the series by proposing their own series.
The point was that finding other sequences with matched the rule didn't do much. Finding a sequence which failed the rule, that is, one that disproved what they thought the rule was, was much more informative.
If you've got lots of brainpower but never push the envelope of your beliefs, that brainpower's not going to be of much use actually expanding your understanding.
Which is why really smart people try to break things: software, hardware, systems of understanding. By pushing them beyond what they feel their limits are and seeing what happens.
Intelligence and education have the effect of greatly increasing the number and variety of "things you can think", exponentially so. Unfortunately the exponentially-vast-bulk of such new thoughts are also wrong. Many of them grossly and obviously so (the vast bulk are new ideas that are syntactically-correct gibbering nonsense), but there's a much more dangerous and still very large fringe of wrong ideas that require one to dig into reality to determine are wrong. And what I often find myself calling something like the Prime Danger of Intellectualism is that our education system is happy to introduce you into that vast space of new ideas, but while flirting for a while with ways of determining their truth, a lot of intellectual tradition in the last 50 years or so has given up on it (in many cases wearing this surrender as a badge of pride!), leaving many modern intellectuals adrift in a massive sea of ideas with no compass, no sextant, and no map. It's a sad and frankly dangerous state of affairs, and it is the default in the intellectual world right now, unfortunately. I know that's a bold statement but once you learn what to look for you, alas, see it everywhere.
I'll condition my definition as "those exhibiting smartness", not merely those possessed of an innate mental capacity and facility.
Knowledge is a container. You can be knowledgeable (full of information) and either smart or not, though I'd consider someone filled with facts and the lack of capacity to connect and relate them sensibly not particularly knowledgeable.
Intelligence is somewhere between a talent and a skill. It's trainable, but requires an innate foundation. #INCLUDE GLADWELL_10000_HOURS
Wise is a mix of both. You need the information (much of it acquired through experience, or at least direct observation), and the capability to use it. And it's the wise who will almost always exhibit the trait of questioning their most closely held facts, beliefs, understandings, and mechanisms.
I've been reading up on Hyman G. Rickover, father of the US nuclear navy. His approach to technology, organizations, training, and more, are very much what I consider to be the workings of an absolutely first-rate mind. I'd like to find a good biography of him.
http://www.reddit.com/r/dredmorbius/comments/28uxu6/more_hym...
And: accepting ideas without testing them is a rapid path to failure. It's anti-evolutionary (variance without selection).
You get to define what you mean by "smart", but others get to reject it.
This only holds if you define "smart people" to be this way. Which is kind of a tautology.
I you use the common meaning of the word, or even just check what people other people consider "smart", you'll find out that they don't always fit your description.
I think to be practical it's best to see who people label as smart, and then note what attributes they have. Instead of starting from the attributes that you think a smart person should have (because then you just made your own definition, and it doesn't describe what most people see).