DIY weather prediction and strategy selection

A way to run weather forecasts with orders of magnitude less computing power and picking a strategy based on what your opponent/competitor doesn’t want you to do.

Reimagining AI weather prediction

I’m not sure what the usefulness of running weather predictions on your own machine is but reducing the compute needed for accurate forecasts has got to be a good thing.

I am sure there are a set of use cases that this opens up, maybe around more specialised/localised prediction?

Choosing the right strategy

Matthew Dicks suggests an approach for picking a strategy based on:

What strategy would my opponent be rooting against? What would my opponent least want me to do? … Then do that.

His examples include how to handle discussions and arguments. I also use this sometimes when playing sport. If I was in their situation what would they not want me to do?

In our business we seem to have one main competitor and it could become very easy to focus on them and set a strategy around just beating them. How can we do what they do but better? How can we offer more features?

Maybe we should look at it through their eyes and focus on what would they not want us to do?