I came across the idea on this website (https://commoncog.com/start-here/), which I feel is far more "applied" than the Farnam Street blog (https://fs.blog/). The latter is a distillation of essentially rationalist type mental models that sound good but often don't work in real life without a specific context (context really matters in real life).
The author of Commoncog actually puts stuff into practice and allows reality be the teacher. He also talks about "concept instantiation" in ill-structured domains. Most of our learning in school is geared toward structured domains where things behave in known ways. Our education system is based around abstract and principles based thinking, where a + b = c. And to be fair, this kind of thinking works very well.... in structured domains.
However, when you get into unstructured domains with lots of higher order effects like say, business or the battlefield or love, those ideas no longer work. First principles reasoning no longer gets you success -- instead you actually have to do the opposite, which is to reason analogically and "instantiate concepts" by assembling context fragments, not principles (e.g. "this has been done before under this context, which is similar to my context in these ways but different in these ways"). That's another insight that changed how I reasoned about ill-structured domains (note: only ill-structured domains! First-principles reasoning still works really well in structured domains so don't make the mistake of abandoning it... only know when to switch to analogical reasoning when you're in the realm of the ill-structured)
The specific part about rationality is here: https://commoncog.com/putting-mental-models-to-practice-part...
The author of Commoncog actually puts stuff into practice and allows reality be the teacher. He also talks about "concept instantiation" in ill-structured domains. Most of our learning in school is geared toward structured domains where things behave in known ways. Our education system is based around abstract and principles based thinking, where a + b = c. And to be fair, this kind of thinking works very well.... in structured domains.
However, when you get into unstructured domains with lots of higher order effects like say, business or the battlefield or love, those ideas no longer work. First principles reasoning no longer gets you success -- instead you actually have to do the opposite, which is to reason analogically and "instantiate concepts" by assembling context fragments, not principles (e.g. "this has been done before under this context, which is similar to my context in these ways but different in these ways"). That's another insight that changed how I reasoned about ill-structured domains (note: only ill-structured domains! First-principles reasoning still works really well in structured domains so don't make the mistake of abandoning it... only know when to switch to analogical reasoning when you're in the realm of the ill-structured)
https://commoncog.com/cognitive-flexibility-theory-the-rules...