One should always strive to consider side effects as far as possible, but there are obviously limitations. The number of effects to consider increases exponenti

One should always strive to consider side effects as far as possible, but there are obviously limitations. The number of effects to consider increases exponentially with the order of thinking. Weighing them against each other becomes harder as well, as the scope increases and effects may no longer be tightly tied within the same domain. Probabilities necessarily become part of the equation, and they only become fuzzier the higher the level. A sufficiently high order of thinking becomes essentially to predict butterfly effects. Not that this makes the task impossible, within some scope. Games like Chess have a similar kind of chaotic unpredictability, yet people get really good at thinking many moves ahead and ignoring dead-end paths. Higher order is better, as far as you can be confident. But the real world is huge, and full of unknowns, so higher order conclusions will be as highly up for debate.