One exercise that I find helpful is... Finding what the components of $Thing will do, where else they may be useful, and being critical on if they're needed.

One exercise that I find helpful is... Finding what the components of $Thing will do, where else they may be useful, and being critical on if they're needed. Sometimes you find that it's not as big as anticipated, there's existing work, or the idea just doesn't hold water. In general, it helps to break the problem down. Trying to bite the whole thing at once is overwhelming Sometimes I find that if I do 20% of this, it fixes a much bigger problem than what I was even looking at