No, if we’re rationalist, we should figure out if the cost of doing the tests is worth the expected gain from getting the test results. If we’re fairly certain that it’s a scam already (so test results won’t change the situation much) and if the tests are expensive, it might be a better idea not to test.
Shouldn’t we first determine whether the amount of effort needed to figure out the costs of the tests is less than the expected value of ((cost of doing tests—expected gain)|(cost of doing tests > expected gain))?
No, because you don’t want ot get into an infinite regress about figuring out the cost of figuring out the cost to, etc. So you need to have a general policy about when to stop the regress, and ensure that your policy is good in most of the cases you’ll actually run into in practice, while acknowledging that it may not be optimal 100% of the time.
No, if we’re rationalist, we should figure out if the cost of doing the tests is worth the expected gain from getting the test results. If we’re fairly certain that it’s a scam already (so test results won’t change the situation much) and if the tests are expensive, it might be a better idea not to test.
If we’re rationalist, we don’t start our reasoning with “If we’re rationalist, then...”.
Shouldn’t we first determine whether the amount of effort needed to figure out the costs of the tests is less than the expected value of ((cost of doing tests—expected gain)|(cost of doing tests > expected gain))?
No, because you don’t want ot get into an infinite regress about figuring out the cost of figuring out the cost to, etc. So you need to have a general policy about when to stop the regress, and ensure that your policy is good in most of the cases you’ll actually run into in practice, while acknowledging that it may not be optimal 100% of the time.