But apart from that obvious one, there are:
* Child prodigies. They are irrelevant in fields like literature, world conquest, philosophy, etc. In highly "technical" areas like mathematics, physics, engineering, technical art, they are overrated in the following sense: most of those "child prodigies" would have been better off developing in a broader way (that is, gaining interest outside of their specialty).
* IQ, or just "tests" in general.
First: outside of the formal sciences or highly technical arts -- say in poetry, novel-writing or other literature -- tests don't mean anything (and grades mean very little). One of the reasons for the decline of fields like "philosophy" is the excess importance attached to academic credentials in the subject itself, that usually consists of being a ideologically-compliant kitten to the instructors. If someone who seeks a career in "philosophy" wishes to demonstrate objective *skill*, then take additional courses or a minor in a technical subject. (UC Berkeley has in fact such a graduate-level program, its " Logic and the Methodology of Science " degree.)
In technical fields such as mathematics or engineering, grades mean something for sure, but not what most people think it is. Tests establish a minimal level of competence - and yet the nature of basic exercises (even graduate-level ones) is that they use existing methods. Another way to state this is: basic "exercises" all involve combinations of existing constructions (if you pay attention to what they are), rather than any new constructions. This requires a minimal level of mental discipline (to force your thoughts to focus on certain things, and no other), and power of analyzing objects into their components and "organizing" it in an efficient way (and furthermore analyzing all possible kinds of visual resemblance into only a few kinds of resemblance, in order to "use" things), but certain *not*, say, the ability to make a completely new construction that is also relevant to a given problem.
For e.g., see this guy , who is an example of someone only minimally competent in the first (even if it was just laziness, the laziness itself would be a form of mental indiscipline, since he can't easily "force" his thoughts into a certain track and no other ). (Other examples of people who lacked technical skill, but were deep analysts of particular problems: Newton at least in the beginning of his career, Leibniz in 'traditional' mathematics (basic things like geometric proofs he found difficult even to the end -- he was only the master of his own methods, using his own notation), Hermite, Sophus Lie, Louis Bachelier, Einstein, Grothendieck [he went to the French equivalent of a "community college", which is one of the low-level Universities instead of the Grandes Écoles.) But clearly, there are people who get perfect GPAs in MIT or Berkeley, and then go on to achieve.. nothing, despite their best efforts to become "the next Einshtein" (many of them become quants instead).
And of course, IQ itself is distinct from test-taking ability of the academic sort. This is because "IQ" still does not test the power of analytic thinking, i.e. no questions that require extremely high, continuous concentration on a visually complicated object without any interruptions. That is, none of these problems require you to decompose an extremely complex object and carefully picture *all of the components at once* as you re-assemble and summarize it, not completely losing sight of anything. In that sense, chess is a better test for analytic power (if not cleverness in summarizing things).
Seemingly, what IQ tests is your ability to cleverly manipulate (not simply analyze) things; to wiggle around in a clever way until the problem itself is simpler, to re-frame a problem, to distinguish objects for the purpose of analysis, not to actually conduct such analysis. I don't believe my mind is that highly focused during IQ testing, but I have maxed out the scores in most of the official ones (e.g., the Stanford-Binet), when I took it during high school. Seriously, alot of people I know consider IQ test questions mostly trivial since it's quite easy to max out the scores by simply analyzing a few simple heuristics (the ability to "shift around" and "re-frame" a problem mostly comes down to optimizing the "search space" for pure possibilities, as well as possible analogies and resemblances).
* Human intelligence, or even "talent".
[ Edit: I heavily revise this portion, based on more recent views of mine.]
Human intelligence is complex enough so that there are many ways of thinking about it, other than IQ. But then.. why even think of *success* as a matter of intelligence, i.e. greater capacity? In most cases it seems to be a compatibility between one's style of thinking and how this "hits the weak spots" of current thinking. Or -- if we think "success" is based largely on "creativity" -- then if "creativity" is analogous to some kind of chaos, or balance between chaotic ideas and order, then creativity is not a question of capacity but of being balanced. I suspect that poetic ability is largely based not on any additional mental capacity (say of visualizing things in more detail), but is largely a question of being "well-balanced", being able to generated chaotic combinations, but then not going too far in this.
But it goes further than this.. I furthermore suspect that a very *vigorous* and active sort of stupidity, which generates ideas that are subtly false, is also an essential ingredient in problem-solving ability. 'Mistakes' must be well-controlled, but there's an entire field (" biomimicry ") that attempts to simulate organic forms of error and mistakes.
Or instead of "intelligence" and skill, it is possible to view things in terms of "insight". In this sense, "perfection of skill" is a micro-insight and nearly insensible, acting upon the "details" of analysis in the smallest intervals in time. (That is, every time you "analyze" an object into components and put it together, there are observational details -- spotting this or that possible pattern or shortcut -- while putting things together. There are of course details of the exact where and how, of manipulation.) This leads to elegance in construction, but not to any single big insight. On the other hand, "chaotic" combinations of imagery, often initially false, are much bigger single insights and suddenly cast a great deal of illumination on a subject. Significant illumination (much like a light cast about a dark room) is always "inelegant", while an "elegant" solution in fact throws no light upon anything.
Another way of seeing this is that "chaos" is the opposite of skill (or mental stability), whereas intelligence is just one among other forms of mental stability. *Instability* is also required. There are also many other ways of seeing it.