I think I’ve made some progress in adopting Norvig’s thinking. I need to:

  • Understand back propogation mathematically.
  • Understand what a ‘wicked problem’ is.

He has a talk where he’s focusing more on the problems and the metrics we’re looking for. Emphasising that it’s a misunderstanding of the problem rather than the program that normally causes issues.

  • He talks about trade offs at one point. Where the solution is to decide what we want to optimise for and to engage and present data to all stakeholders.