Systems must take performance hits to get out of local optima. Actions that move you against the gradient usually look stupid over a short time horizon.
This is a specialized version of the fact that Decision quality is timescale dependent.
Web URL for this note
Comment on this note
SpaceX and Blue Origin are two examples of modern definite optimistic thinking
Roadmaps can enable faster progress than blind exploration
The tech tree model suggests that explicitly drawing out tech trees is important
The structure of the tech tree can inform actions that are short term suboptimal but long term useful