Maybe the best way to illustrate our point is to use weather models, as it's pretty tangible for regular people.

Weather models are ridiculously advanced running on super computers crunching a massive global network of real time data. But they still aren't anywhere close to perfect.

That being said, just because it rains on your birthday when it was said to be sunny, doesn't mean you delete your weather app, call meteorology pseudoscience, and start a substack of "the forecast was wrong again" blog posts.

People intuitively grasp this foolishness because they constantly interact with weather models. But for things they have almost no contact with, it's easy to write it off on a single "bad forecast"

(I'll also admit the caveat that not every model is as good (or bad) as weather models, another dimension at play to throw a wrench in peoples gears)