Mine is:
"Just build something people want."
I don't think that's enough anymore.
Plenty of founders build useful things.
The harder part seems to be figuring out whether the right people even recognize they have the problem.
Curious what advice you've changed your mind about.