>>12498126Disagree, to a degree.
I purposefully chose a rando problem that involves pattern matching to illustrate a point: IQ more or less recognizes pattern matching, which tracks with intelligence/problem solving. Someone who approaches the pareto-frontier optimization problem will face a pattern-matching problem, and if you are good at pattern matching, you could come up with some solutions (generally, I'm thinking stochastic embedding like t-SNE or more like UMAP).
In case you're wondering, this is legit an "unsolved" problem, and something I'm working on. You wont find this in a text book or paper other than formulating the problem and some "solutions" that are okay (pair-based optimization).
That idea is that pattern-matching is unversal, and therefore the IQ test will test that. IF you take IQ test after IQ test, you stop testing the
universal pattern matching and start asking "can I take an IQ test well" by learning the specific IQ test.
