I think the reason behind it is fundamentally open-ended problems without clear end points. Where it’s possible (due to non-transitive preferences / opinions on when something is done / what to focus on) to infinitely loop over the same parts without making a lot of headway.
The way that I approached solving this meta problem is by trying to come up with a consistent way to tackle anything that can be executed with maximum speed. (i.e. where I don’t have to kind of passively “feel” that I am done and have clear completion criteria).
Reminds me a lot of Cynefin but a more explicit/structured approach to solving the Complex quadrant.
What I do when I have an open-ended problem is this:
- Put it into an outlining tool like Workflowy - Decide what outcome I want (e.g. this plan is done, this task is done, I have decided between N things, I have identified the things to decide between) - Brain dump various thoughts and just kind of go where the gut pushes me - Then stop and inductively organize / structure the outline
From there I continue iterating on it and investigating things (this often include carrying out & completing specific tasks in the outline), and I specifically check in on these 3 rules to determine when to stop:
- The outcome is hit, and I can clearly summarize in 1-2 sentences - I am looping over the same things repeatedly (reminds me of a cycle detection algo of a recursive fn) - The next thing I would look at will not change the outcome
I’ve thought about this problem a lot.
I think the reason behind it is fundamentally open-ended problems without clear end points. Where it’s possible (due to non-transitive preferences / opinions on when something is done / what to focus on) to infinitely loop over the same parts without making a lot of headway.
The way that I approached solving this meta problem is by trying to come up with a consistent way to tackle anything that can be executed with maximum speed. (i.e. where I don’t have to kind of passively “feel” that I am done and have clear completion criteria).
Reminds me a lot of Cynefin but a more explicit/structured approach to solving the Complex quadrant.
What I do when I have an open-ended problem is this:
- Put it into an outlining tool like Workflowy
- Decide what outcome I want (e.g. this plan is done, this task is done, I have decided between N things, I have identified the things to decide between)
- Brain dump various thoughts and just kind of go where the gut pushes me
- Then stop and inductively organize / structure the outline
From there I continue iterating on it and investigating things (this often include carrying out & completing specific tasks in the outline), and I specifically check in on these 3 rules to determine when to stop:
- The outcome is hit, and I can clearly summarize in 1-2 sentences
- I am looping over the same things repeatedly (reminds me of a cycle detection algo of a recursive fn)
- The next thing I would look at will not change the outcome