Every time an article involving personality pops up, someone inevitably mentions MBTI, and shortly thereafter a "pseudoscience" or "astrology" comment follows u

Every time an article involving personality pops up, someone inevitably mentions MBTI, and shortly thereafter a "pseudoscience" or "astrology" comment follows up to accompany it (I feel like this phenomenon should have its own name—akin to Betteridge's law of headlines). And of course I always follow up on the follow-up comment to point out that none of Big Five, MBTI, or even the latest peer-reviewed attempts at temperament classification based on various dimensionality reduction techniques (e.g., ICA) perform very well. Not one. Sure, they're all more predictive than random chance, but it's not like MBTI is a horoscope and the SOTA matches the accuracy of particle physics. It would be as though someone switched from using logistic regression to a random forest, but the nature of the data is such that there's only a very minor bump in precision/recall, and both were already low to begin with. I would ask what purpose is the model being used for in this case? It's not like quant trading where even slight improvements can lead to outsized returns. If a personality model is being used to predict human behavior for any purpose other than entertainment, then I'd argue that this practice is misguided and should be discontinued in all professional settings. Whereas if the purpose is simply to have fun, then I'd argue MBTI achieves that goal much better than the other models. So what is the point of Big Five then?