What on earth does the following footnote mean? Are we back to consciousness
where the "quantumbuck" stops?
Understanding Deutsch's Probability in a Deterministic Multiverse by Hilary
The following objection is sometimes raised against the decision-theoretic
approach: in an Everettian context, all outcomes of a decision are realized,
and therefore it simply does not make sense to make choices, or to reason
about how one should act. If that is correct, then while we may agree that
probability can in principle be derived from rationality, this is of no use
to the Everettian, since (it is claimed) the Everettian cannot make sense of
If this was correct, it would be a pressing 'incoherence problem' for the
decision-theoretic approach. The objection, however, is simply mistaken. The
mistake arises from an assumption that decisions must be modelled as
Everettian branching, with each possible outcome of the decision realized on
some branch. This is not true, and it is not at all what is going on in the
decision scenarios Deutsch and Wallace consider.
Rather, the agent is making a genuine choice between quantum games, only one
of which will be realized (namely, the chosen game). To be sure, each game
consists of an array of branches, all of which will, if that game is chosen,
be realized. But this does not mean that all games will be realized. It is
no less coherent for an Everettian to have a preference ordering over
quantum games than it is for an agent in a state of classical uncertainty to
have a preference ordering over classical lotteries.