Agency cannot be a purely quantum phenomenon

TL;DR

This paper argues that a purely quantum agent evolving unitarily cannot satisfy three minimal agency conditions: world-model construction, deliberation over consequences via that model, and reliable maximization of expected utility. Central obstacles arise from the no-cloning theorem, which forbids perfect copying of unknown quantum states required for world-model replication and testing multiple actions, and from the linearity of quantum dynamics, which prevents a universal mechanism from selecting the best action across arbitrary action sets. The authors show that even with unlimited perfect copies or with approximate cloning, the fidelity and generality needed for robust agency fail, except in special cases where a known classical basis (decoherence-induced) makes copying feasible and a single best action deterministically implementable. They also provide circuit-level analyses demonstrating that realistic quantum agency implementations fall short of true agency unless classical resources and basis knowledge are incorporated. The work has broad implications for the emergence of classicality, the feasibility of quantum simulations of agential behavior, and the assessment of quantum theories of agency, free will, and consciousness.

Abstract

What are the physical requirements for agency? We investigate whether a purely quantum system (one evolving unitarily in a coherent regime without decoherence or collapse) can satisfy three minimal conditions for agency: an agent must be able to create a world-model, use it to evaluate the likely consequences of alternative actions, and reliably perform the action that maximizes expected utility. We show that the first two conditions conflict with the no-cloning theorem, which forbids copying unknown quantum states: world-model construction requires copying information from the environment, and deliberation requires copying the world-model to assess multiple actions. Approximate cloning strategies do not permit sufficient fidelity or generality for agency to be viable in purely quantum systems. The third agency condition also fails due to the linearity of quantum dynamics. These results imply four key consequences. First, agency requires significant classical resources, placing clear constraints on its physical basis. Second, they provide insight into how classical agents emerge within a quantum universe. Third, they show that quantum computers cannot straightforwardly simulate agential behavior without significant classical components. Finally, they challenge quantum theories of agency, free will, and consciousness.

Figures (2)

• Figure 1: This table attempts to characterize the performance of four different quantum agency circuits, by evaluating our 26 test inputs. These circuits use between $N=2$ and $N=5$ environment states, as shown. The left block shows the case where the agent is given $N$ perfect copies of the pure state $|\psi\rangle$, and the right block shows the case where the agent must make $N$ symmetric clones of a single $|\psi\rangle$, which really means creating a symmetric entangled state of $N$ qubits and acting the deliberation unitaries on the first $N-1$ of this state. $Q_{IX}$ deliberates using $I$ and Pauli $X$ for its two control qubits. $Q_{IHX}$ deliberates using $I$, Hadamard $H$, and $X$ for its three control qubits. $Q_{IX'Y'Z'}$ deliberates using $I$, and rotated Paulis $X'$, $Y'$ and $Z'$ for its four control qubits. Lastly, $Q_{(I)X}$ deliberates using only $X$ for its single control, and blindly applies $I$ for control $|1\rangle$. Because $Q_{(I)X}$ does not deliberate on at least two choices, we think of it as a proto-agent, but it is noteworthy that it performs just as well as $Q_{IX}$ with perfect copies, and very similarly for clones.
• Figure 2: For each example quantum agency circuit, the complete unitary is given, along with the details of the controlled unitary $C_U$. The target state is $|0\rangle$, so where a deliberation unitary produces a $|0\rangle$, $C_U$ applies that unitary, or a superposition of all deliberation unitaries that gave $|0\rangle$. $X'$, $Y'$, and $Z'$ are the Pauli matrices rotated to have equal expectation values $1/\sqrt{3}$ for the target state $|0\rangle$.