From Emil, Some people have IQs of 160:
Hoel is right about that. People who claim very high IQs are usually narcissistic or mislead. There is no normal way to obtain reliability estimates in that range for normal people. But that’s something we can address by designing better tests! It won’t be entirely possible to prevent people from cheating and taking the test many times, or secretly training beforehand, but if we really wanted to, we could make a company that tests people in person using modern computer adaptive testing and give people proper high-range results. Is there a market for this? Probably not, but maybe. Certainly, online tests of this kind should be made. In fact, I am coincidentally working on a project like this right now.
Creating such a test likely requires an exceedingly large number of questions to disentangle luck from skill at the highest end of IQ, and it would be impractical to proctor it in a timed and controlled setting. And also, validating such a test would be close to impossible too. Or the test being compromised due to the answers being circulated/shared online, which was a problem with earlier tests. It’s hard enough creating a single test that validates well, let alone many tests. There is likely no market for it either, because top companies don’t need IQ tests to find top talent: they recruit, such as from MIT or other top-20 schools.
The SATs have undergone many revisions over the past century, but it’s generally assumed that tests taken prior to 1995 have a very high potential IQ ceiling, high enough that the Triple Nine Society, which has a 99.9-percentile IQ cutoff, uses it. But administering the SATs is quite an involved process, and is not something that can be done online easily, nor something that people would want to do voluntarily (part of the reason the SATs are so accurate for screening is because so many people partook in it). Due to smart phones, smart watches, etc., cheating would be a bigger problem compared to 1995.
I disagree with it being narcissistic. Sure, IQ is not as easily quantifiable as height, but how is it any more narcissistic for a smart person to claim to have a certain IQ, such as 160, especially with supporting evidence, than it is for a tall person to proclaim he or she is tall? (Although it is possible for the individual to be mislead, as he says.) An IQ of 160 is 4 standard deviations, or one in 30,000 people. If one’s skills or abilities are commensurate with a 1/30,000 rarity for a sufficiently g-loaded task, then at least we have something in the ballpark of being supporting evidence. If anything, tall people are at a greater advantage in society even compared to smart people, such as higher wages, better dating market, and more social status, yet why it it socially acceptable to have pride in tallness but not smartness, especially given that neither of them are earned?
One way to ascertain very high IQ is to compare how someone does relative to his or her peers. If Wall St. is a battle of wits between traders to find the best methods and strategies, to be the best of an already self-selected group for high IQ, is another tier above that. You cannot be as accurate at forecasting and finding pattern in price action or charts unless your IQ is high–and or unless you are a prodigy/savant or something like that. Stock chart patterns are like advanced versions of IQ tests in which money is at stake.
It’s like knowing where the puck will be. For example, I predicted that BTC would lag ‘big tech’ as the stock market recovers, and sure enough it has, and now the media is acknowledging such. The lagging continues as Bitcoin is still below 28k and the 3x funds keep going up, meta at $266. A few days ago Barron’s, a leading financial publication, wrote about the major underperformance of crypto relative to big tech stocks, well after I predicted it. And now today from Bloomberg: Crypto Rues Bitcoin’s Decoupling From AI-Fueled Tech Stock Gains.
It’s one thing to get a few things right or to predict a general trend, but to get everything right including the individual constituents, especially when it’s assumed they should be correlated, is way more indicative of skill than luck. The assumption was that Bitcoin, Ark Investments, and QQQ would go up together due to being correlated, but I was the only one, correctly, to predict that only QQQ would go up and the rest would lag significantly. The assumed high correlation between the assets means getting the all the orderings right is not halved with each correct trial by more like cut by 1/4, so getting all right is around 1/32, depending on the conditionals.