A thought on this, and apologies if it repeats something already said here. Basically: question the structure that leads to someone saying this to you, and question how easy it is to talk about 3^^^^3 people as opposed to, say, 100. If suddenly said person manifests Magic Matrix God Powers (R) then the evidence gained by observing this or anything that contains it (they’re telling the truth about all this, you have gone insane, aliens/God/Cthulhu is/are causing you to see this, this person really did just paint mile-high letters in the sky and there is no Matrix) should be more than enough to tip the balance of evidence in favor of “yeah, holy crap, this person deserves my $5!” In short—don’t even take seriously your model of an agent as being truthful or untruthful; it could be outside-context-problem wrong, especially if it behaves pathologically, being overly sensitive to small amounts of evidence for bad reasons. Similar idea to your whole model possibly just being wrong if it reports 1-1/3^^^3 certainty.
A thought on this, and apologies if it repeats something already said here. Basically: question the structure that leads to someone saying this to you, and question how easy it is to talk about 3^^^^3 people as opposed to, say, 100. If suddenly said person manifests Magic Matrix God Powers (R) then the evidence gained by observing this or anything that contains it (they’re telling the truth about all this, you have gone insane, aliens/God/Cthulhu is/are causing you to see this, this person really did just paint mile-high letters in the sky and there is no Matrix) should be more than enough to tip the balance of evidence in favor of “yeah, holy crap, this person deserves my $5!” In short—don’t even take seriously your model of an agent as being truthful or untruthful; it could be outside-context-problem wrong, especially if it behaves pathologically, being overly sensitive to small amounts of evidence for bad reasons. Similar idea to your whole model possibly just being wrong if it reports 1-1/3^^^3 certainty.