I had to research intelligence as part of my master's thesis. You can only measure intelligence if you have a group of very similar people. So let's say me, my sister, and three of my four brothers. One of my brothers is autistic, which makes him (neurologically) a different "type" of person, with a different way of organizing thought. The rest of us would be somewhat similar and you could measure us against each other mentally. My one brother would be what is called an outlier and would mess up the whole thing for everyone. If we included him, because his thinking system is so different, he would be likely to test as way higher or way lower than what would be accurate. It would be like testing five circles and then one square. How much of a circle is the square? It's not a circle, so you can't include.
When you have a very large population, it doesn't matter that you have a few outliers. They average out. So it doesn't mess up your measurements of everyone else. It might matter to that one individual that they are being tested as retarded or genius, but to the country or world it's not a big deal- not worth changing the system. Which is why we still use intelligence tests.
If this is something that truly bothers you, I would encourage you to read up on how intelligence is measured and tested. It is eye-opening. Here's the wikipedia to get you started if you're interested:
Intelligence quotient - Wikipedia
I could be wrong and I didn't look this back up all the way, but this part here:
Each successive edition has re-normed the test to compensate for the
Flynn effect, ensuring not only that the norms do not become outdated which is suggested to result in inflated scores on intelligence measures, but that they are representative of the current population (Flynn, 1984, 1987, 1999; Matarazzo, 1972).
Wechsler Intelligence Scale for Children - Wikipedia
I think that's referring to this thing that happened in early tests where they gave the tests they developed to groups of people and everyone scored "retarded" (the official term then). So they were like, oh, we can't say that regular white males are retarded, so they adjusted it until regular white males were getting 100 on the tests. They did not do that with women, non-whites, or any other minorities. So minorities always got lower scores on IQ tests, which played into the idea that white men were genetically superior. That's why when kids apply to college any kind of minority gets padding to their test scores before they are considered. Seems like it would be easier to correct the tests, don't you think? It's really fascinating when you start reading about it.
Anyhow, there's all that if you want it.