Comment by kbenson

> The interesting thing is my professor kinda knew we do this, he just acknowledged it as part of the dance of their system.

Professor's diary: It's so tiring coming up with broken experiments that still have some possible merit, but the system works, and my role is clear. If only the benefits of working under constraints weren't so clear with regard to innovation, they they are what they are and this farce continues for all of us. Maybe I'll finally feel like the private sector is the way to go next yet. Probably not, but who knows.

Also, this sounds like I've heard the military described at times, expecially in war, where the upper echelons come out with wile ideas that make no sense on the ground and mid-level officers pull wild solutions out of their asses and whatever works ends up being copied.

I could see something like this possibly developing as a natural solution when all you look at is the output and not the process, and provide a rigid framework within which different behavior can be iterated until it stabilizes on something that works. That, unsurprisingly to me, has similarities in how ML works, given given these are basically institutions that act as machines.