Let’s say you want to change something in the world. Scientists will tell you to do it like this:

That assumes that figuring how what to do separately from doign it with a reasonable amount of effort in a reasonable amount of time. But the things we tend to care about are not like that (if it were that easy, it’d already be fixed). Another way to change things is adaptive management, an environmental science technique that takes place entirely in the real world:

You cannot run tests in a laboratory, because labs are hard to do properly and you don’t have time. At worst, you’re trying to change your life: you’ve only got one and you can’t spend it all theorizing.

For example, everyone trying to be more fit is doing adaptive management: you implement a diet and exercise regime, then measure weight, clothing size, appearance naked, energy levels, hunger feelings, etc.

The hardest thing about adaptive management is measurement (“what you can’t measure, you can’t manage”). A lot of things in the world are not directly measurable, so you have to find proxy measures. But proxy measures only obliquely get at your goal, for example: nobody is ever actually trying to “lose weight”, it’s a proxy measure for fitness. But using weight as a measure leads to perverse management plans like ketosis, which sheds muscle as well as fat.