There's a compelling hypothesis in the field of machine learning called 'overfitting,' where systems become so tuned to recognizing patterns that they falter when faced with new, unexpected scenarios. Interestingly, some computer scientists are exploring ways to inject noise into computational models to keep them adaptable.

Wealth is just an indicator of how well (or not) things are going. Whatever I'm doing, there are indicators… do people like what I'm doing? Do they think I'm crazy? Am I making a lot of money? These are just indicators. Wealth is not and end in itself, it's just one of many indicator as to how you're doing.

The small act of paying attention to people is extraordinarily rewarding to every member of the team — and it’s amplified in an inverse way from their level on the hierarchy.

I give all of you permission whereby, if I change the way I behave, my beliefs, the way we are going, I give you permission to hit me in my face, you need to bring me back to reality.

Milton Friedman had a postulate that capital would move into a region hit by an adverse shock. But as soon as you interrogate this postulate, it's manifestly rubbish. When Sheffield's steel industry collapsed, investment didn't flow in saying 'oh good, a depressed region.' It flowed out to the places which were booming and accentuated the divergence.