>PRESENTATION:
Chris Fuchs is a research staff member at Bell Labs, Lucent Technologies, an Adjunct Professor of Physics at the University of New Mexico, and an associate editor for the journal Quantum Information and Computation. His interests range from quantum channel theory to quantum cryptography to quantum foundations. Perhaps the most outlandish thing he has done is publish a collection of quantum emails; see
http://www.arxiv.org/abs/quant-ph/0105039

In the neo-Bayesian view of quantum mechanics that Appleby, Caves, Pitowsky, Schack, the author and others (maybe D'Ariano?) are developing, quantum states are taken to be compendia of partial beliefs about potential measurement outcomes, rather than objective properties of quantum systems.
Different observers may validly have different quantum states for a single system, and the ultimate origin of each individual state assignment is taken to be unanalyzable within physical theory---its origin, instead, ultimately comes from probability assignments made at stages of physical investigation or laboratory practice previous to quantum theory. The objective content of quantum mechanics (i.e., the part making no reference to observers) thus resides somewhere else than in the quantum state, and various ideas for where that "somewhere else" is are presently under debate---there are adherents to the idea that it is purely in the "measurement clicks," there are adherents to the idea that it is in intrinsic, observer-independent Hamiltonians, there are adherents to the idea that it is in the normative rules quantum theory supplies for updating quantum states, and so on. This part of the program is an active area of investigation; what is overwhelmingly agreed upon is only the opening statement of this abstract---that quantum states are compendia of beliefs. Still, quantum states are not simply Bayesian probability assignments themselves, and different representations of the theory (in terms of state vectors or Wigner functions or C*-algebras and the like) can take one further from or closer to a Bayesian point of view. It is thus worthwhile spending some time thinking about which representation might be the most propitious for the point of view and might, in turn, carry us the most quickly toward solutions of some of the open problems. In this talk, I will explore various issues to do with the above and explain why I prefer a representation of quantum mechanics that makes crucial use of a single probability simplex.