Great point. This is going to get OT but I don't care.
I was thinking alot about patterns, mostly when it comes to the AI crap.
Obviously, understanding weather patterns is a very important part of weather forecasting, whether it be short-term, medium-range, or long-range. However, I think there is a big misconception within the field that there is a strong correlation between weather patterns and what results.
For example, major severe weather outbreaks in the Great Plains typically have a deep trough digging into the West with ridging building across the East. However, not every trough digging into the West is going to produce a major severe weather outbreak. During the winter, not every ridge in the West/trough in the East is going to produce above-average snowfall here.
But AI focuses on this...all about "identifying patterns" and then saving time because instead of calculating complex equations, a solution is just based on a slew of "historical patterns". BADDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD
Understanding the pattern and it's relevance is one thing, however, IMO what's more important is how the pieces are moving and evolving within the pattern which may not be directly related to the pattern at all.