Andreas Albrecht
These are topics of interest which I would like to hear from others about, and also which I have thought a lot about and would be happy to present a talk and or lead a discussion (I am especially keen to discuss my new work listed in item 7c, either informally with others or in a talk):
1. Relationship between arrow of time and "tuning puzzles" in cosmology
2. Relationship between cosmic inflation and the arrow of time in cosmology.
3. The crucial role played by the Bunch Davies vacuum in cosmic inflation
4. Entropy/information theoretic analysis of all of the above
o e.g. Bunch Davies vacuum as "infinite blank tape" and cosmic inflation as "daemon"
5. Can the requirement of the emergence of classicality bring insights into the "initial conditions" of the universe?
o Especially the emergence of semiclassical spacetime?
6. The role of the arrow of time in the physics of einselection
o Can classicality behavior/einselection exist in equilibrium (without an arrow of time)?
7. A unitary version of the "Caldeira Leggett model":
o The Caldeira Leggett model (and its offshoots) are a useful tool for studying decoherence and einselection
o However, the standard treatments always take a special order of limits to arrive at irreversible equations which approximate the unitary evolution in a special regime (of course when an arrow of time is present)
o I have developed (and am studying numerically) a CL-like model which I can evolve in a fully unitary manner. This makes it possible to extend the analysis outside of domains where the arrow of time is present, and address a variety of interesting topics including topic 6 above.
Charles Bennett
My alternate title for this workshop: Toward a Mathematical Home for our Classical Phenomenome, highlights the responsibility of cosmology to come up with a mathematical model of the physical universe within which the body of phenomena we observe, our phenomenome, is typical of those the theory predicts we should observe. This task, of course, is ill-defined insofar as it depends of the definition of "we", i.e. the anthropic selection criterion. But for reasonable criteria, a wide range of otherwise reasonable cosmological models, including the standard LCDM model---roughly those that equilibrate under short-range interactions to a positive temperature---appear to suffer from a Boltzmann brain problem, or what Sean Carroll, following David Albert, calls "cognitive instability."
To me, part of the problem seems to come from the disparity between the kind of distributions that arise from equilibrium statistical mechanics, eg. canonical or microcanonical, and the kind considered in algorithmic information theory, e.g. the universal semimeasure, defined as the output distribution of a universal Turing machine on random input. Distributions of this latter kind are universally fertile, producing self-organization on a set of positive measure under any computable definition thereof. Moreover they are stably embeddable, through the theory of fault-tolerant computation, in simple dissipative (irreversible) dynamics, but not in simple equilibrium (reversible) dynamics. On the other hand, universes with perpetual dissipative dynamics are unattractive because they seem to need some implausible deus ex machina to perpetually hold them away from equilibrium, which is no better than the tacit assumption of a special initial condition to explain the observed universe's atypically low entropy.
The question of whether a classical or quantum system with short-range interactions can self-organize under equilibrium dynamics is of course related to the question of topologically stabilized passive memory, which is known to be exist in quantum systems with 3 or more dimensions (for classical memory) and 4 or more dimension for quantum memory. Recent work by Matthews, Meklo and Poulin, which I'd like to hear more about at the workshop, suggests that topologically stabilized passive memory may even be possible in classical 3d systems. But the mere existence of stable memory, including even arbitrarily complex "logically deep" topologically stabilized states, is not sufficient if they are so infrequent, among the states visited by the system's dynamics, as to be outnumbered by local partial replicas---Boltzmann brains.
So perhaps we need to give up locality of interaction, undermining familiar notions of space. Could a quantum theory of emergent spacetime give us other kinds of spacetime rich enough to generically produce structures of unbounded logical depth even at equilibrium?
Adam Brown
Statement of Interest: holographic quantum complexity; classical and quantum chaos; the power of systems in thermal but not complexity equilibrium; virtual observers.
Susannah E. Glickman
As a young historian studying the development of quantum information and computing, I hope to get a sense of this cosmological aspect of the field. I am curious to see the kinds of questions and lines of inquiry researchers pursue as well as the means through which they pursue them. I also find this type of theorizing really exciting and interesting. I look forward to participating to the extent I am able remotely.
David Poulin: A three-dimensional self-correcting classical memory?
A self-correcting memory is a passive device that stores information robustly despite fluctuation of its external parameters like temperature, magnetic field, pressure, etc. It is well established that a memory can be stabilized locally using fault-tolerant cellular automaton error correction, but such a dynamical process consumes power. In contrast, a self-correcting memory is stabilized thermodynamically and does not require external power. In particular, we are interested in self-correcting memories that arise from systems composed of localized, bounded degrees of freedom interacting locally.
The prospect of quantum technologies has generated a growing interest for robust quantum memories, and the possibility of conceiving a self-correcting quantum memory in three spatial dimensions or less is being debated. Here, we take a step back and consider the simpler task of conceiving a classical self-correcting memory in three dimensions. We present a simple and well known lattice model and argue that it could serve as a self-correcting memory.
Renato Renner
Do we need to modify quantum theory in order to avoid contradictions in multi-agent scenarios (where each agent uses the theory from their perspective)? If yes, how could such a modification look like?
