On the surface this is a good article, but it is by someone who is not a scientist in the subject. A few errors: Those tests mentioned are out of an area about 150 years old called " cognitive testing" where the objective is to measure the speed of providing a right answer, not so much a right answer.
Fundamentally cognitive testing can measure how quickly you can come to master a science, for example, simply because you can quickly know certain facts and quickly do certain computations.
They do predict success in what they actually claim to measure. They are however, the wrong questions to ask or the wrong kind of testing if, for example, a field is highly politicized (and some are). For example a Stanford or Harvard education is probably a better predictor than an IQ test in predicting the money you will make ("success").
Modern science (after the Enlightenment) succeeds because people can replicate claimed observations. The test of science is not that you or I can have the current knowledge to observe, but that with wave after wave of new people entering a field, the truth comes out precisely because they are taught quite rightly to confirm for themselves or be sure they can trust those who have.
I have always said, read the original work first because it explains the fundamental observations. For example, James Clerk Maxwell if you want to understand the basic science of electricity and magnetism. JCM even shows you how to set up the experiments and make the measurements. (That said, it is also good to have someone explain the math until it makes sense to you what it is about, and what it is not about.)
We should trust the experts who other experts in the same field respect. Since the Enlightenment science is indeed about empirical truth. That is also exactly what our brains are about, but we must also trust others, as must everybody, when we have not personally observed the existence of something or its causation. And even then, there are numerous fallacies of reasoning we or they can get into. The only mechanism we have which works is 'enforced moderated dialogue' which is, in fact, what we see in science, in courts, and, interestingly in things like good scholarship and the United States Constitution. Ethan is certainly right precisely because the human mind will always approximate truth as best it can based on personal experience. Comprehensive truth is for practical purposes, impossible to fully learn in one lifetime. Be careful who you trust.
We are all experts at understanding lies. The reason is that our brains are classification engines that associate things by predication which place the strongest certainty of truth in what we perceive, do, and cause. Our brains are not memory and logic engines like standard computers. With deceit about, you need to see proof you are not being deceived before giving experts your trust to be your proxy with the empirical truth you could in principle have observed yourself.
Here is truth to any person. Scientifically, Descartes had it right about truth. He is considered the 'father of the Enlightenment' and is worth reading/audibling (his "Meditations") in the original:
