All models fall short of reality, it doesn't mean they aren't useful models. Saying something isn't useful because it categorizes people rather misses the poin

All models fall short of reality, it doesn't mean they aren't useful models. Saying something isn't useful because it categorizes people rather misses the point of modelling in the first place; of course it categorizes people, that's the point, to simplify reality. It doesn't matter if the model has 5 types or 200 types, if the model is useful in grouping like minded people, then it's a useful model and the number of categories is simply a matter of how accurately I want to measure. I can create a model of 2 kinds of people, haters and non-haters; this is a useful model. Meyers Briggs is also a useful model especially as shorthand for Young's functions, which are also useful models. While the big 5 might map closer to reality, it seems a less useful model precisely because it doesn't categorize. Giving me 5 scores on a scale isn't very useful in finding like minded people because I can't say "what's your big 5 type?" It doesn't simplify enough. Knowing I'm an INTP on the other hand, is very useful in finding like minded people.