So much to unpack here. 1) Intelligence isn't linear. How could it be. Problems in the general sense are terrifying graphs of graphs with meta edges everywhe

So much to unpack here. 1) Intelligence isn't linear. How could it be. Problems in the general sense are terrifying graphs of graphs with meta edges everywhere that are conditionally relevant depending on what "value" is at any given vertex. Being good at one type of problem doesn't necessarily translate into being good at another type of problem. Being gifted doesn't mean you're necessarily smarter than your peers it just means that you're better than they are at a particular type of problem that you aren't generally expected to be good at. 2) Problems get harder at an exponential rate while utility from intelligence improves at a linear rate. Example: Brain surgery is pretty hard. Fighting a gorilla is pretty hard. Now imagine trying to perform brain surgery while simultaneously fighting a gorilla. If you are gifted as a child then you're solving problems that are harder than what your peers can solve. But as the set of problems you want to solve gets more difficult the value of your giftedness is going to be lessened. 3) Utility = effectiveness * alignment * time (more or less). Having your head in the game (alignment) can make a real difference in whether or not you're useful. Being able to commit more time can make a real difference in whether or not you're useful. And intelligence can affect effectiveness, but experience or having good mentors can also raise your effectiveness (sometimes more than intelligence can). 4) As far as I can tell, some people like to unify all of their experiences into a singular framework of understanding the world and some people like to compartmentalize their experiences into different spheres. To some extent we all do both. Unifying as a child can make you look gifted because you'll start to understand complicated topics where your peers only mimic competency (think having an algorithm that calculates outputs versus a lookup table that just gives out memorized results). However, often in the real world (tm) mimicking is more than effective enough for what people want to get done. So the compartmentalizers are completely adequate. However, the compartmentalizers also can deal with incompatible scenarios much better than the unifiers. If you try to unify the cold logic of a computer with the fuzzy interpersonal relationships of office politics, then you aren't going to be making any progress without years of studying both. However, if you can compartmentalize the two, then you can start being productive and advancing on day one. There's going to be some cap to what you can do as a compartmentalizer, but that cap is probably far above what you need for nearly all domains that most people are likely to encounter.