Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Caledonian2 comments on That Alien Message - Less Wrong

111 Post author: Eliezer_Yudkowsky 22 May 2008 05:55AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (164)

Sort By: Old

You are viewing a single comment's thread.

Comment author: Caledonian2 22 May 2008 02:14:35PM 0 points [-]

I will further add that it wasn't Einstein's IQ that made him one of the greatest physicists ever. It's been estimated to be between 160 and 168. Feynman's is even less impressive - 124. *I* can beat that without even trying.

There are other aspects of intellectual functioning besides those that contribute to IQ test performance. The 140-average world will not have one out of every thousand people becoming new Einsteins.

Comment author: bigjeff5 11 October 2011 06:27:08PM 11 points [-]

125 according to a high school IQ test in the 1930's.

He was doing integral and differential calculus at age 15 (much better than anybody in my high school managed as a high school freshman). He was formulating mathematical concepts he hadn't yet been taught before he entered college.

I think it more likely that the IQ test he took was wrong than that he was only a little above average intelligence. His body of work is further evidence of this.

Comment author: Jack 11 October 2011 07:10:42PM *  3 points [-]

If intelligence is just defined as the g factor then it isn't clear what grounds we have to dispute test results. Instead, what cases like Feynman's suggest is that while a single variable account of intelligence is powerful predictor academic and job performance, income, etc society-wide, the cognitive abilities of individuals can be considerably disparate. This is why modern tests have multiple indexes. I took WAIS III when I was 18 and they test runners refused to actually give me my full IQ because the verbal IQ and the performance IQ were 30 points apart. (not that it wasn't trivial to just average them if I wanted to).

Comment author: TitaniumDragon 16 April 2013 05:11:00PM 8 points [-]

I know I am several years late to this party, but I felt it appropriate to hop in.

Einstein, with an IQ of 160+, really has an unmeasurably high IQ - at that IQ level, you are outside of the bounds of statistical probability used to construct the test. 160 correlates to such a rare IQ that you cannot use IQ tests to measure differences between them in a very predictive fashion because there aren't enough people with IQs that high.

Now, what about the case of Feynman?

Well, there are tons of possibilities here:

1) It was a bad IQ test.

2) He got extraordinarily unlucky (Remember, your variability on a modern test is +-5 points; if the older tests had higher margins of error, it could be he was significantly more intelligent than this)

3) IQ is not a perfect indicator of g. This is actually known. It strongly correlates with g, but it is not identical to g - it is entirely possible that he was smarter than the IQ test indicates because of this discrepency.

4) He did badly on the test for some external reason (he was tired, the test didn't get graded properly, he got the wrong person's score back... any number of possibilities that could theoretically lower his IQ).

5) He really DID have an IQ of 125 in high school, but via concerted effort increased his intelligence greatly over time. In other words, he may have had significant untapped potential. Did he take the IQ test before or after he went on his crazy math-learning spree? This is especially true given he was still in adolescence.

6) He may really have only had an IQ of about 125, maybe as high as the mid 130s, and simply made better use of it than most people.

Comment author: gwern 16 April 2013 05:36:06PM 2 points [-]

Now, what about the case of Feynman? Well, there are tons of possibilities here:

Some of which can be verified if you trace this widely repeated anecdote back to the source; see my old comment on the topic: http://news.ycombinator.com/item?id=1251164