I wrote the other day about the limitations of artificial intelligence (AI) in the COVID crisis. Those limitations are largely due to the fact AI is largely about gathering large sets of data and analyzing them far more efficiently and effectively than humans can, to find potential important patterns humans might miss. 

At the moment, I'd describe AI as an incredibly powerful technology seeking the best practical application to 'break through' and show it's value. It's possible that application might be management of natural disasters. There's LOTS of data already available on disasters and how humans have reacted that we could start to develop models in advance for much more efficient and effective disaster management and recovery. 

For example, what if you could predict with great accuracy both how a wildfire might spread based on geography and weather conditions, but which buildings might be most at risk, and how people in that path are likely to move to evacuate? Disaster managers could assign resources well in advance (if not already have in a practiced plan) to specific spots and take advanced action to minimize damage. And wildfires are just one example. Flooding, hurricanes, tornadoes, and more can be modeled as we tend to have quite a bit of data available for AI to analyze and build predictive models.

The promise seems evident, and there is definitely some work in this direction. But much, much more is possible. Let's keep trying to learn more and see if the promise is real or not.