But IQ performance likely predicts success in data science, and not only that, it does predict success in a variety of subjects. I'm also with the post on the bottom -- I don't think Lior believes in his post. I interpret this as humble-bragging. There's enough silliness in this post that permits the author to wiggle away; I'm sure the author already agrees that it's unfair to pick two highly prejudicial examples, or reduce causal modelling down to Occam's razor.
Or perhaps it's a jab at subtle misconceptions on causal learning and empiricism. Either way it's not meant to be taken seriously at face value.
But do get started on IQ.