A Computational Anthropic Principle: Where is the Hardest Problem in the Multiverse?

内容摘要

The anthropic principle is an inevitable constraint on the space of possible theories. As such it is central to determining the limits of physics. In particular, we contend that what is ultimately possible in physics is determined by restrictions on the computational capacity of the universe, and that observers are more likely to be found where more complicated calculations are possible. Our discussion covers the inevitability of theoretical bias and how anthropics and computation can be an aid to imposing these biases on the theory landscape in a systematic way. Further, we argue for (as far as possible) top-down rather than bottom-up anthropic measures, contending that that the latter can often be misleading. We begin the construction of an explicit computational measure by examining the effect of the cosmological constant on computational bounds in a given universe, drawing from previous work on using entropy production as a proxy for observers by Bousso, Harnik, Kribs and Perez. In addition, we highlight a few of the additional computational considerations that may be used to extend such a measure.