Thank you for this! I really wanted to go deeper on human factors, and I think there's a lot to be said about CRM and sociotechnical systems design, especially when ML gets used for decision support. Ultimately wound up truncating that section (along with more of the economic critique) because the piece was already far too long.
There’s a paper out there, on designing IT systems from god knows when. It is incredibly dry, except for a line in it that stood out: All IT systems are political systems, because they decide how information and decisions flow.
I can only guess as to how much content you would have to explore on that axis.
You're welcome! I imagine you already know this one as well but just in case.
Learning to Learn by the late Dr Richard Hamming. See especially Chapter 2.
A point Hamming makes is that when transitions from hand to machine production occurred, usually what is built ends up changing as the old techniques don't transfer 1:1 from the old world.
So for instance, we went from nuts and bolts to rivets and welding (Dr Hamming's literal example). This required builders to produce an equivalent product to the old, built with different techniques - and crucially! - under tighter control limits.
The reason things are going all over the place with AI at the moment is that it's speed, speed, speed. They had an all hands at my company recently where the top brass talked about AI. The only thing mentioned was speed - go faster, do more, etc. Not a single soul talked about quality.
But if you know your software engineering wisdom you know that you can only pick two when it comes to speed, scope, or quality. It's going to get real dumb for a while until people realize/remember quality is how you achieve speed.