I think apps and common is the key part of this topic. A practical approach is to define the outcome you want, test it with a small example, and then check quality, speed and reliability separately. The original point — "For AI Apps, it helps to think in terms of inputs, constraints...