I want to welcome this tech, albeit with a bad taste of AI marketing in my mouth. But all of this feels surface level and none of it talks about the previous models / tech, at least without skimping on some details.

What's improved? What COULD be improved? Who are you aiming to serve?

Since machine learning is inherently based on previous historical data, will this model be able to accurately predict the unpredictable? This is assuming that the next forecasts or, worse, weather disasters will be similar to the ones we've seen before.

Clearly, that's not the case.

I don't want AI grifters or techno-optimists, I want to hear from those with boots on the ground & a career dedicated to this kind of work. Google touts the WeatherNext 2 to "surpass previous (the) previous model on 99.9% of variables," and those are just words to me. What does that actually mean?

Obligatory bogo sort mention

Legitimately a classic, shed a tear

HAPPY BIRTHDAY!!! 🎉🎉🎉🎉🎉🎉