Embed Notice

I think of it as sort of like calling one's shots in billiards. You decide what change you would like to see in the system you are designing for, you come up with an "intervention" (based on studying the problem, reading into other people's approaches to trying to solve the problem, and maybe doing a bunch of rounds of iterative experimentation), you decide, up front, how you will determine whether or not the problem has been solved, and then you implement the intervention, and then you check those criteria you previously identified as the ones that will determine whether or not the problem has been solved.

This is what I mean by rigor. This is pretty sciencey, no? It's not necessarily a controlled experiment, but it does have the form of an experiment. But it's not a *mere* experiment, either. It's not just a trial to see whether or not something will work. It's an attempt to actually do something that will work. With some slightly more rigorous testing as to whether it did.

🧵