Although on the flipside, let's pretend it's 2017's and you are Nvidia selling GPU's for Bitcoin - maybe demand will dry up at some point? Do you stop scaling production as this might be the max of the market, or do you follow the market and increase production?

It's always easier to see the right move in hindsight!