Hacker News Reader

Comment by commandlinefan | Hacker News Reader

> Boiling brains down to one number

IQ does measure something, though, and it tracks pretty consistently with that more ephemeral thing that people refer to as "intelligence": if you, personally, found a dozen people you considered "smart" and another dozen you considered "stupid", you'd almost definitely find that the smart ones had high IQs and the dumb ones had low IQs. You might argue about the fine-grained distinction between the smart ones, but you wouldn't find somebody who you considered smart who somehow still had a 50 IQ.

View on HN

Just because something can be measured does not imply that metric accurately depicts reality. Moreover, all metrics should be criticized as to what part of reality they are depicting and what part of reality they are obscuring. I've yet to be convinced that IQ is anywhere near an accurate depiction of reality.

I like the way you think. Language, similarly, is not an accurate depiction of reality.

But scores that correlate well are still just averages, and are not wise to apply to individuals as you do in your closing statement. IQ tests like progressive matrices are still very particular, specialized mental activities. Over a population of people, yes it tracks well, since general intelligence coefficient tells us all intelligences are correlated. On average a person with high IQ is smart at any other given activity. But for a particular individual, it's still perfectly normal to see large differences in testing one type of intelligence compared to another.

I give an IQ test to a tribal in the Amazon jungle and a PhD in Chemistry.

The PhD in Chemistry will have the superior IQ but ask both to survive in the jungle and the tribal wins hands down.

It’s a very facetious assumption that solving a few puzzle questions captures all the facets of a human beings creativity. This is like saying that the best leet coders are the best software engineers.

You're fighting a straw man here. Nobody claimed that IQ models all facets of human creativity.

The statement "You might argue about the fine-grained distinction between the smart ones, but you wouldn't find somebody who you considered smart who somehow still had a 50 IQ." seems to claim that IQ models human creativity- or rather, human intelligence ("smartness")- comprehensively.

If it captures such a small facet of what it means to be human then its a very poor metric to optimize for. We don’t even know what trade offs we are making.

I'm not fully convinced of this, and even if I were, it might just say as much about my biases as the test's.

That's true, but that's a bar for an extremely crude instrument.

IQ is an imperfect measure of intelligence. However, it's also the only (objective) measure we have.

Yet depending on the use case, a bad map may be worse than no map at all.

Reply on news.ycombinator.com