It from Qubit: David Deutsch on Quantum Information and Reality

David Deutsch’s “It from Qubit” (2002) argues that quantum information provides a powerful way to understand physical reality—but cannot, by itself, explain why the laws of physics exist or take their particular form.

The paper is primarily a conceptual argument about the foundations of physics, rather than a presentation of new experimental results.

1. Why “qubit” rather than “bit”?

Deutsch begins with John Wheeler’s question, “It from Bit?”: could the physical world arise from information?

His first move is to emphasize that classical bits are approximations to a fundamentally quantum reality. A classical bit has two possible values, whereas a qubit has a continuum of possible states, despite a binary measurement having only two possible outcomes.

For example,

    \[ |\psi\rangle=\alpha|0\rangle+\beta|1\rangle, \qquad |\alpha|^2+|\beta|^2=1. \]

The amplitudes vary continuously, while measurement yields a discrete result. Quantum theory therefore combines continuous evolution with discrete measurement outcomes.

2. The mathematical building blocks are quantum observables

Deutsch favors the Heisenberg picture, in which the state remains fixed and observables evolve.

He describes a qubit using three operators satisfying the Pauli algebra:

    \[ \hat q_x\hat q_y=i\hat q_z, \qquad \hat q_x^2=\hat q_y^2=\hat q_z^2=\hat I, \]

with cyclic permutations of the first relation.

Each operator has eigenvalues +1 and -1. These operators encode more than a collection of ordinary yes-or-no answers: their noncommuting relationships capture the qubit’s quantum structure.

Classical bits and classical computation emerge approximately when decoherence suppresses observable interference.

3. Universality connects simple qubits to complicated physics

The central argument concerns universal quantum computation. Simple quantum gates can be composed into networks that perform arbitrary quantum computations to a desired accuracy.

Deutsch distinguishes several aspects of universality:

  • Simple gates can generate general quantum computations.
  • One quantum computer can emulate other quantum computers.
  • Quantum computers can simulate physical systems.
  • Such computers must ultimately be physically constructible for this picture to describe nature fully.

The third point connects directly to QFT: a field can be approximated on a lattice, with qubits encoding the degrees of freedom near each lattice site and gates reproducing local interactions.

The important qualification is that representing or simulating a field with qubits does not establish that spacetime is literally a lattice of qubits.

4. Why this does not mean the universe is “made of information”

Deutsch rejects the inference that computational universality explains reality’s ultimate origin.

Information processing requires physical systems obeying physical laws. Consequently, appealing to computation alone does not explain those laws.

He also criticizes the idea that our universe is software running on an external “Great Simulator.” Such a proposal, he argues, relocates the unexplained physics to the simulator’s hardware.

His distinction is:

Physics explains what computations are possible; computation alone does not explain which physics exists.

5. His stronger claims require interpretation

Deutsch describes qubits as “multiversal objects,” reflecting his commitment to the Everett, or many-worlds, interpretation. This is his interpretation of the formalism, rather than a conclusion uniquely established by quantum computation.

Similarly, the paper’s broad claims about efficient physical simulation should be read with care: simulation resources depend on the system, state preparation, required accuracy, and requested output. Universality does not make every physical calculation easy.

6. Why he ends with constructor theory

A computer can simulate a process without making that process physically achievable. Simulating an engine, for example, does not construct one or supply its fuel.

Deutsch therefore proposes a broader framework—constructor theory—concerned with which physical transformations are possible, which are impossible, and what resources they require. Computation would become one special case within that framework.

The paper’s central takeaway is that qubits may provide a common language for describing physical systems, but a complete account of nature must also explain the physical laws and constraints governing them.