I wish someone would write an expert system that works like his dad. Something that had some validation of what was a worthwhile risk and what was a risk that even if you tried to mitigate the risk really would still be a risk. (For instance, even if you had enough work to presume you could employ that employee for more than 2 months, that's not a guarantee. What if the contract with a customer was for 12 months, but 2 months in, right after staffing up, they just cancelled it and said "fucking sue me"? Kinda hard for the small startup to sue... and if they did, and they won, that win would be years after they had to let the employees go.)

You can't mitigate all risk, and being able to ignore the ones that you can't' do anything about is an important skill.

This goes for technical risk as well. I try not to over-architect. I spend a lot of time trying to decide which things need to be handled now, and which things can just be added later.