Like virtually everything else in biology and learning theory the optimum is not at either "edge" of the parameter space but somewhere in between. Willingness to entertain model-breaking ideas and data is a compromise between the need to avoid falsehood and error and the need to learn and update models when needed.
Like virtually everything else in biology and learning theory the optimum is not at either "edge" of the parameter space but somewhere in between. Willingness to entertain model-breaking ideas and data is a compromise between the need to avoid falsehood and error and the need to learn and update models when needed.