Lxr

The article doesn't explain what is actually going on to produce the second chart, but here is a guess. Assume people have a true ability x, an estimated ability y which is x + some unbiased noise, and a test score z which is also x + some unbiased noise. If you take a sample and collect the lowest quartile of test scores z, then the average of the estimated abilities y in that group is higher than z because the y are produced centered at x rather than centered at z. By collecting the lowest quartile of test scores, you didn't just get the dumbest people, you also got the people who happened to perform badly on the test that day. These people may be estimating their ability accurately, in which case the estimate would be higher than their randomly bad performance that day.

This seems like too obvious a mistake to not have been noticed for this long though.

ML is a shiny object with often no discernible ROI but occasionally very large ROI, and companies are understandably nervous about missing out. Spending a small amount to hedge their bets isn't necessarily irrational.

This seems like a better explanation, because the effect is stronger for things that absorb water than for things where the water sits on top.

Similar to 厉害 in Chinese - it often translates as terrible, but mostly used to mean great.

Why use HTML in emails though?

> it creates low trust

Try looking into the camera when you speak rather than the screen.

What is the best way to profit from all this crypto hype, as a rational person who’s convinced all coins will eventually go to zero?

This. Router-level blocking will become impossible without intercepting https, and endpoint blocking also gets harder.