Pretty much the only thing I think AI could be useful for - forecasting the weather based off tracking massive amounts of data. I look forward to seeing how this particular field of study is improved.
Bonus points, AI weather modeling, for once, saves energy relative to physics models. Pair it with some sort of light weight physical model to keep the hallucinations at bay, and you’ve got a good combo.
You’ll also need more accurate remote sensor data (precipitation happens in a very narrow range of temperature, pressure and humidity), better observation data, better terrain models (microclimate is influenced by the interaction of terrain and the atmosphere). The forecast grid sizes used now are based on choosing the smallest grid size we can afford to compute that yields meaningful forecast data. Computational cost more than quadruples each time grid size halves. “More than” because altitude levels matter too, though for terrestrial forecasts, the ones near the ground matter a whole lot more than what’s happening at 6km.