Turing's thesis and physics Matthew Szudzik Abstract: Alan Turing claimed that every mathematical function which can be effectively computed by humans is a function that can be computed by a Turing machine. This claim, known as Turing's thesis, serves as the theoretical foundation for computer science. It guarantees, for example, that if a human can imagine an explicitly-described procedure, then that procedure can be automated by a computer. In the decades since Turing proposed his thesis, various generalizations have been proposed. For example, the physical generalization of Turing's thesis states that any function which can be effectively computed by a physical process is a function that can be computed by a Turing machine. But this simple generalization has been historically difficult to formalize. What is the correct formal definition of the physical generalization of Turing's thesis? We will discuss some of the historically significant attempts to answer this question, and describe our own recently-proposed solution.