[harekaze]

(1) The details of popular IQ tests are considered commercial secrets by testing companies. The reliability data and validity data are kept secret to people other than selected experts connected to those companies. This opaqueness does nothing to enhance the credibility of IQ tests.

(2) The claims that IQ tests are supported by consensus of psychologists ignore that, scientific knowledge does not rely on consensus to justify its credibility, but the explanatory power and predictive power stemmed from such knowledge. Plus, most psychologists don't do research on IQ tests, the so-called consensus can be regarded to an extent as textbook claims in disguise.

(3) There have always been disagreements. Multiple intelligence theory is one that made its way to textbooks. But there are others. For example, the mutualism model by van der Maas, Kan, and Borsboom (2014) claims that IQ tests do not constitute as a reliable measure that can reflect the underlying mental capacity (if there is one, not none or multiple), the value of IQ tests can only be (weakly, in my opinion) justified by pragmatic considerations [0]. The mutualism model is somewhat tangential to the claims made by Taleb, but my point is that IQ tests are not something that should be glorified.

(4) Psychometrics in general, has deep issues. Operationalism plagues the discipline. The practice of many (I shall reserve properly that it's not all) psychometricians does not meet the idea of validity as instigated by current recommended guidelines, and what should be constitutive of validity is still somewhat an open question that remains to be explored. These are not direct evidence for or against the credibility of IQ tests, but nonetheless should make us cautious how we reach our judgments.

[0] https://www.mdpi.com/2079-3200/2/1/12

EDIT: grammar