then in expectation code2 will outperform code1
Deciding whether a decision theory outperforms another is already decision-theory complete, and thus circular/trivial/not a valuable argument, even if it at first doesn’t seem like it. See this
Deciding whether a decision theory outperforms another is already decision-theory complete, and thus circular/trivial/not a valuable argument, even if it at first doesn’t seem like it. See this