The problem I have with dividing the space of problems up into a well-defined, and poorly-defined binary and saying they each have their own distinct type of intelligence, is that it seems like the way you make progress on poorly-defined problems is by taking it or a sub-component of it and making it well-defined.
For example, maybe "Do a good job raising a family" is poorly defined, but when we think about how we evaluate if we have achieved this, it sort of caches out into a list of criteria like:
- Your children have a happy childhood
- Your family's financial needs are covered
- Your children grow up to be well-educated and have fulfilling careers.
- Your children grow up to be healthy
- Each family member develops strong personal connections to each other family member
- ...
Then these things can be pursued using the tools of well-defined problem solving. If you can't do this, then how could you even tell if someone did a good job solving a poorly-defined problem, surely at some point you have to cache it out in some sort of objective criteria.
This would be a problem for your theory because if you can convert poorly-defined problems into well-defined problems, at least in approximation, then it seems like "smart" people, should be able to leverage their smarts to achieve all of the well-defined outcomes.
My answer to this would be that the alternative virtue which you are trying to pick out isn't some sort of "smartness in solving poorly defined problems" but instead something like "directionness". That is, the bottleneck isn't that people are trying to but don't have the right smarts to solve poorly defined problems, rather actually picking good poorly defined problems to pursue and sticking to them over ones life is really challenging and a virtue independent of IQ-smarts.
This is because the space of poorly-defined questions that are relevant to a good life is actually very nebulous and itself poorly-defined. People are confronted with many different poorly-defined questions they might want to satisfy (lets call them imperatives) e.g. "have pleasurable experiences", "make my loved ones happy", "affirm my identity(s)", "be high status", "Fulfill social pressure (x,y,z)", "benefit my culture/community/tribe/nation/species".
Most people most of the time are in a state of internal strife over which of these imperatives matter and to what extent, and its doubly hard because the relative imperative importance shifts over time so your actions to satisfy imperative X one day may be viewed as totally wasted the next because obviously imperative Y is the most important one!
This directionness virtue would pick out people who are better than average able to (1) develop a coherent, non-contradictory (or less-contradictory) understanding of the importance of various imperatives to them, (2) Keep them stable over time (or at least evolving in a slower, more orderly way).
Then, the high-intelligence person may be better at satisfying any individual imperative at any given point in time, but the high-directionness person will be more satisfied that their actions over time have accumulated in a way that satisfies their current desires.
Note this is a super vague and not well-evidenced theory, and it also has the problem where it still would predict SOME correlation between IQ-smarts and happiness even if it is less important than directionness.
I like your comment, though I will say that beyond a certain level of the criteria you listed, happiness does seem to become something more intangible, and not something that can be further "well-defined". The concept of directionness seems akin to agency, which is frequently mentioned in online discussion. I don't think it's a problem that SOME correlation exists between IQ smarts and happiness - possessing IQ smarts is likely to assist people in achieving tangible, hard-to-deny goals, which would can on average lead to a good sense of achievement/ feeling that one has put in a commendable amount of effort.
Being more happy is something you can train. A little mindfullness goes a long way. Avoiding destructive patterns of thought can be taught. This isn't to say that unhappy people should be blamed for their unhappiness, but rather that most people could be a lot happier than they are if they made more of an effort. But many people are unaware of this, and expect happiness to be something completely outside of their control. How to be happy is not taught in school, but perhaps it should be.
Physics isn't at all like chess in it's so-called "well-defined problems". Dark matter, the measurement problem, baryon asymmetry, etc. There really shouldn't be a single behavioral measurement that predicts success somewhat in both physics and chess. And yet IQ does. And that's amazing. And it does this across so many domains.
But you expect it to predict moral goodness also?! You want a measurement that differentiates good and bad outcomes in every kind of human behavior?
Meanwhile you've taken happiness measurements at face value as if they represent a valid target that IQ has failed to hit. The basic self reports you've mentioned are hopelessly entangled with multiple internal validity problems like social referencing and self presentation, which make time period comparisons really really hard to interpret. They are far weaker as measurements of anything. Don't blame IQ for being as bad as happiness scales.
This should really be an essay about the metacognitive failures of smart people.
I wonder about the AI thing though. I mean, you're probably right about what it would have done with the question about going to the moon but it does seem to be able to pull together a lot of ideas to come up with something a little bit new. I've also noticed that it has made assumptions that were correct but for different reasons than I would have made the same ones. I think it's possible AI might have seen a little further than its human training would predict at that point in history. Just a thought though.
Nice Essay and a neat theory!
The problem I have with dividing the space of problems up into a well-defined, and poorly-defined binary and saying they each have their own distinct type of intelligence, is that it seems like the way you make progress on poorly-defined problems is by taking it or a sub-component of it and making it well-defined.
For example, maybe "Do a good job raising a family" is poorly defined, but when we think about how we evaluate if we have achieved this, it sort of caches out into a list of criteria like:
- Your children have a happy childhood
- Your family's financial needs are covered
- Your children grow up to be well-educated and have fulfilling careers.
- Your children grow up to be healthy
- Each family member develops strong personal connections to each other family member
- ...
Then these things can be pursued using the tools of well-defined problem solving. If you can't do this, then how could you even tell if someone did a good job solving a poorly-defined problem, surely at some point you have to cache it out in some sort of objective criteria.
This would be a problem for your theory because if you can convert poorly-defined problems into well-defined problems, at least in approximation, then it seems like "smart" people, should be able to leverage their smarts to achieve all of the well-defined outcomes.
My answer to this would be that the alternative virtue which you are trying to pick out isn't some sort of "smartness in solving poorly defined problems" but instead something like "directionness". That is, the bottleneck isn't that people are trying to but don't have the right smarts to solve poorly defined problems, rather actually picking good poorly defined problems to pursue and sticking to them over ones life is really challenging and a virtue independent of IQ-smarts.
This is because the space of poorly-defined questions that are relevant to a good life is actually very nebulous and itself poorly-defined. People are confronted with many different poorly-defined questions they might want to satisfy (lets call them imperatives) e.g. "have pleasurable experiences", "make my loved ones happy", "affirm my identity(s)", "be high status", "Fulfill social pressure (x,y,z)", "benefit my culture/community/tribe/nation/species".
Most people most of the time are in a state of internal strife over which of these imperatives matter and to what extent, and its doubly hard because the relative imperative importance shifts over time so your actions to satisfy imperative X one day may be viewed as totally wasted the next because obviously imperative Y is the most important one!
This directionness virtue would pick out people who are better than average able to (1) develop a coherent, non-contradictory (or less-contradictory) understanding of the importance of various imperatives to them, (2) Keep them stable over time (or at least evolving in a slower, more orderly way).
Then, the high-intelligence person may be better at satisfying any individual imperative at any given point in time, but the high-directionness person will be more satisfied that their actions over time have accumulated in a way that satisfies their current desires.
Note this is a super vague and not well-evidenced theory, and it also has the problem where it still would predict SOME correlation between IQ-smarts and happiness even if it is less important than directionness.
I like your comment, though I will say that beyond a certain level of the criteria you listed, happiness does seem to become something more intangible, and not something that can be further "well-defined". The concept of directionness seems akin to agency, which is frequently mentioned in online discussion. I don't think it's a problem that SOME correlation exists between IQ smarts and happiness - possessing IQ smarts is likely to assist people in achieving tangible, hard-to-deny goals, which would can on average lead to a good sense of achievement/ feeling that one has put in a commendable amount of effort.
Being more happy is something you can train. A little mindfullness goes a long way. Avoiding destructive patterns of thought can be taught. This isn't to say that unhappy people should be blamed for their unhappiness, but rather that most people could be a lot happier than they are if they made more of an effort. But many people are unaware of this, and expect happiness to be something completely outside of their control. How to be happy is not taught in school, but perhaps it should be.
Physics isn't at all like chess in it's so-called "well-defined problems". Dark matter, the measurement problem, baryon asymmetry, etc. There really shouldn't be a single behavioral measurement that predicts success somewhat in both physics and chess. And yet IQ does. And that's amazing. And it does this across so many domains.
But you expect it to predict moral goodness also?! You want a measurement that differentiates good and bad outcomes in every kind of human behavior?
Meanwhile you've taken happiness measurements at face value as if they represent a valid target that IQ has failed to hit. The basic self reports you've mentioned are hopelessly entangled with multiple internal validity problems like social referencing and self presentation, which make time period comparisons really really hard to interpret. They are far weaker as measurements of anything. Don't blame IQ for being as bad as happiness scales.
This should really be an essay about the metacognitive failures of smart people.
Really great article!
I wonder about the AI thing though. I mean, you're probably right about what it would have done with the question about going to the moon but it does seem to be able to pull together a lot of ideas to come up with something a little bit new. I've also noticed that it has made assumptions that were correct but for different reasons than I would have made the same ones. I think it's possible AI might have seen a little further than its human training would predict at that point in history. Just a thought though.
The short answer is what i want written on my tombstone. "He was too smart for his own good"
The factors that never seem to matter in "intelligence" testing are curiosity and empathy.
If you are not curious about the world and its systems, you really are not intelligent, you are learned at best.
And if you have no empathy, you don't have the engine for curiosity, which is always connected difference.