(Un)predictability in physics
Before the RANPHYS (Randomness and irreversibility in physics) project, research devoted to the development of generalised approaches for unpredictability was limited. The belief that the outcome of quantum measurements is intrinsically unpredictable was based on theoretical results, such as the Kochen-Specker theorem. To fully understand quantum unpredictability, it is necessary to have general models. The RANPHYS model relies on the ability of a predictive agent to accurately and reproducibly predict the outcome of an experiment using information that the agent extracts from the 'environment' as input. Importantly, it takes into account the epistemic limits of the observer. Scientists used the computability theory that originated in the 1930s with the study of computable functions and the so-called Turing degrees to provide an information-theoretical approach to unpredictability. With this model, relativised with respect to the means of the predictive agent, different physical systems and theories can be uniformly analysed. The RANPHYS model is non-probabilistic, a feature that may be overly restrictive given that highly probable outcomes seem predictable. Nonetheless, the uncertainty still present in the high probabilities represents an important unpredictable aspect of the physical processes. On the other hand, certainty is needed if model predictions are to cover definite properties of physical systems. Scientists used this new model to study the unpredictability of different quantum phenomena. They found that quantum complementarity leads to a form of relativised unpredictability that is weaker that the Kochen-Specker indefiniteness. The results have been published in the peer-reviewed Information journal.
Keywords
Unpredictability, non-probabilistic model, RANPHYS, computability theory, physical systems, quantum phenomena
