I genuinely don't understand the point in contention here. I get the positive manifold, which is a thing out there in the world. I get that it is a positive manifold across tests that it might be argued have an ambiguous relationship with what we subjectively call intelligence.
-
-
-
Causality is the point of contention. g has no referent which could be studied via causal modelign. Even weaker in fact, there is no physical phenomenology. There is no part of the brain you can point to and say, "this is the g part of the brain" afaict. Phlogiston comp is apt.
- 2 more replies
New conversation -
-
-
Oh I like that. Kolmogorov-Chaitin is my go to mental model for intelligence. See also Schmidhuber. Wisner-Gross almost had a good related idea but it appears to be bs upon deeper look.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
And I think that should be taken as strong evidence that it is an ill-posed construct that is primarily of social and ideological utility. Trust it no further than mathified superstition. "Data without generalization is just gossip" etc.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Yes, it's a good read, but be sure to also read this very good and thorough critique of Shalizi's arguments https://humanvarieties.org/2013/04/03/is-psychometric-g-a-myth/ …. Having been through all these pieces, I would not say IQ stats can really (currently) be taken down in the way detractors might hope.
-
Nor can they be propped up as strongly as proponents hope :D
- 1 more reply
New conversation -
-
-
I think the weakness is less in the heritability claim (which I totally accept) and more in the coherence of the thing being inherited. It just doesn't seem that meaningful to me.
-
As Shalizi says, intelligence is the measure of how good are you at being a clerk in industrial societies.
- 1 more reply
New conversation -
-
-
I'm a little bit obsessed by that Shalizi post. I've read it every year for years and my understanding is that it's an attack on the idea that it makes sense to reduce correlated cognitive abilities to a 'general' component and a 'specific' component.
-
It doesn't directly undermine the idea that cognitive abilities exist and are - for the most part - correlated. There is a suggestion that 'new' cognitive abilities are barred because they *aren't* 'g-loaded', which is plausible.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
I basically don't trust a pop-stats idea until Shalizi has taken a run at it