Hacker News Reader

He addresses this: "New methods like dual n-back or nootropics are ... years later discovered to ... be tantamount to training on IQ tests themselves"

View on HN

But if IQ tests are trainable, what are they actually measuring?

If bench presses are trainable, what are they actually measuring?

Bench presses are measuring your ability to bench press. Iq tests are measuring your ability to do IQ tests (and not your IQ)

Edit: just to clarify, the fact that you can train to improve an IQ score means that it is not a very good measure of innate ability. You can't really talk about "having an IQ", all you can say is that you got 130 on your last IQ test, a bit like how you might tell someone what you got for your SATs.

How else is IQ defined?

You'd first have to define intelligence.

Why not define it as something that one can train? That certainly matches my intuitive notion of "intelligence".

Your intuitive notion of intelligence is "something one can train"? So then muscular strength is intelligence?

Uh, no. My intuitive notion of intelligence is that it is trainable. That is not its only attribute, come on.

Reply on news.ycombinator.com