This is a bit of a brain dump, not much new. I'm putting them down to see if any solutions come to mind.
The problems we have are hard. You don't get the dopamine hit of solving something or getting money from a product. You have to keep at it knowing what you are doing might be a dead end or a distraction.
Even if you identify a sub-problem to solve sometimes it is hard to know if solving that problem we really help.
Because of these two things, and all the content on biases that we read, you have to constantly doubt what you think you know and what you should be doing. Because of Goodhearts you have to worry about lost purposes when you actually try to do something uncertain.
If you can't trust yourself, you can't trust other people. Even if you trust their motivation to save the world you cannot trust their knowledge and that they are not protecting the epistemology. And they find it hard to trust you. Working with people can raise the trust level of their knowledge, but not of their sub-goals.
A whole bunch of people, you don't know very well, acting together, trying to maintain a public persona so the organisation can persist over time. That is even harder to trust.
In short it is hard to find contented feeling of a job well done or relax in the knowledge that competent people are on the job. This is different from problems that have short time frames.