Zuse's Thesis - Zuse hypothesis - Algorithmic Theory of Everything - Digital Physics, Rechnender Raum (Computing Space, Computing Cosmos) - Computable Universe - The Universe is a Computer - Theory of Everything

Zuse's thesis: The 
Universe is a Computer

Zuse was the first to propose that physics is just computation, suggesting that the history of our universe is being computed on, say, a cellular automaton. His "Rechnender Raum" (Computing Cosmos / Calculating Space) started the field of Digital Physics in 1967. Today, more than three decades later, his paradigm-shifting ideas are becoming popular.

Konrad Zuse (1910-1995; pronounce: "Conrud Tsoosay") not only built the first programmable computers (1935-1941) and devised the first higher-level programming language (1945), but also was the first to suggest (in 1967) that the entire universe is being computed on a computer, possibly a cellular automaton (CA). He referred to this as "Rechnender Raum" or Computing Space or Computing Cosmos. Many years later similar ideas were also published / popularized / extended by Edward Fredkin (1980s), Jürgen Schmidhuber (1990s - see overview), and more recently Stephen Wolfram (2002) (see comments and Edwin Clark's review page ). Zuse's first paper on digital physics and CA-based universes was:

Konrad Zuse, Rechnender Raum, Elektronische Datenverarbeitung, vol. 8, pages 336-344, 1967. Download PDF scan.

Zuse is careful: on page 337 he writes that at the moment we do not have full digital models of physics, but that does not prevent him from asking right there: which would be the consequences of a total discretization of all natural laws? For lack of a complete automata-theoretic description of the universe he continues by studying several simplified models. He discusses neighbouring cells that update their values based on surrounding cells, implementing the spread and creation and annihilation of elementary particles. On page 341 he writes "In all these cases we are dealing with automata types known by the name "cellular automata" in the literature" and cites von Neumann's 1966 book: Theory of self-reproducing automata. On page 342 he briefly discusses the compatibility of relativity theory and CAs.

Contrary to a widely spread misunderstanding, quantum physics, quantum computation, Heisenberg's uncertainty principle and Bell's inequality do not provide any physical evidence against Zuse's thesis of a CA-computed universe! Gerard t' Hooft (Physics Nobel 1999) in principle agrees with determinism a la Zuse: proof by authority :-)

On page 343 Zuse points out that entropy cannot really grow: "If we consider the cosm as a big computer in the sense of "Rechnender Raum," not influenced by the outside [...] then the information content of this system cannot increase."

Zuse does not claim to have a complete theory of everything in form of the precise algorithm computing our universe. But his 1967 paper clearly is the first publication of the field. And in 1969 his full-fledged book came out:

Konrad Zuse, Rechnender Raum, Friedrich Vieweg & Sohn, Braunschweig, 1969.

English translation: Calculating Space, MIT Technical Translation AZT-70-164-GEMIT, MIT (Proj. MAC), Cambridge, Mass. 02139, Feb. 1970. PDF scan.

As always, Zuse was way ahead of his time - for decades his wild ideas on the universe as a computer have been ignored by many physicists. Still, some did appreciate this work. For example, "Rechnender Raum" is explicitly cited among Zuse's outstanding contributions in Peters' widely acclaimed atlas of world history, where he is listed among the 20th century's 30 most important persons. And apparently in the new millennium the time has finally come: ideas in Zuse's spirit have recently started to attract a lot of attention.

If Zuse's thesis is correct, then which is our universe's program? It may come as a surprise to some that we already know at least one algorithm that does compute our universe (and several others), taken from page 1 of Schmidhuber's 1997 paper:

Systematically create and execute all programs for a universal computer, such as a Turing machine or a CA; the first program is run for one instruction every second step on average, the next for one instruction every second of the remaining steps on average, and so on.

The method is more efficient than it may seem at first glance. A bit of thought shows that it even has the optimal order of complexity. For example, it outputs our universe history as quickly as this history's fastest program, save for a (possibly huge) constant slowdown factor that does not depend on the history size. Since some universes are fundamentally harder to compute than others, the algorithm leads to Schmidhuber's nontraditional predictions about the most likely futures of our universe.

Check out the "everything" mailing list archive for discussions of digital universes in Zuse style.


As of 2002 more than 80 streets and squares carry the name of Zuse.

Back to