Of course my question would be what’s the point of creating such a conceptual model of intelligence at all. Doesn’t model imply predictability and you might be possibly saying it’s not predictable.
So maybe ‘model’ is a bit ambitious. Analogy or metaphor is probably a better start. In either case, we should identify what properties of intelligence we’re trying to represent. A good place to start is with TNC’s observation that ‘people are scary in a way that think tanks are not.’ That is, any understanding of intelligence must account for these types of peculiar human tendencies:
- It’s possible for a young-Earth creationist to become the world’s best pediatric neurosurgeon
- Theoretical condensed matter physicists may not succeed in experimental physics. I haven’t discussed this point as much, but my physics friends and I spent hours on it. This observation for me is both mundane and profound. On one hand, it’s obvious why theorists may not do well in an experimental lab. But then again, in many ways they know physics better than the experimentalists. Yet it’s not enough.
- People might instinctively “get” a certain topic but flounder on a related one. In my case, I grasped electromagnetic waves quickly. But I didn’t do the same for electronic circuits. In the latter, I often found myself mechanically solving problems without intuition about what I was doing and why.
- Some very, very, very smart people are simply not that good at computer programming or solving differential equations. Conversely, some very, very, very smart people cannot string 5 coherent sentences together.
To perhaps crudely sum up decades of research that I haven’t read: Even if we believe there is a general intelligence (g-factor) across various cognitive tasks (and the experts appear to disagree on this), we can’t easily predict how it will manifest in different people. We can’t easily predict how it will manifest within individuals across different situations. In short, intelligence and cognitive ability are very complicated.