Seek to standardize on benchmarks to make decisions easier if they are meaningful quantitative measures.
Decide on a handful of quality candidates
Move on after a few “strike outs”, but forgive after some time.
May even be best to leave decisions to chance (50/50 coin flip) to save time if easy to undo. Example: hire quickly, fire quickly. Note externalities.
Optimize for lean feedback cycle of learning, then stop, and adapt, and iterate.
Inherantly scientific and human nature to reach for the stars (Star Trek).
Minimize long-term catatrospic loss, but be a reasonable optimist.