Avoiding Negative Externalities - a theory with specific examples - Part 1 — LessWrong