I see no mention in this summary of controlling for PC and other signaling.
His entire premise should only be expected to apply where beliefs have to pay rent, e.g. in a prediction market.
[edit] I see this is an old discussion, and that Konkvistador raised my broad point already.
I see no mention in this summary of controlling for PC and other signaling.
His entire premise should only be expected to apply where beliefs have to pay rent, e.g. in a prediction market.
[edit] I see this is an old discussion, and that Konkvistador raised my broad point already.