Many scientists and philosophers seek to provide an overall account of the world, often described as a God's eye view, that aims to picture how the universe is as a whole. In this piece, theoretical physicist Eric Cavalcanti argues that quantum mechanics throws a wrench into this idea. Using a simple thought experiment, he shows that at the heart of quantum mechanics, there is a paradox that prevents a God's eye view. The upshot is: reality is different for everyone.

Imagine being able to tell the complete story of the universe. Not merely what you have witnessed, or could discover, but everything that actually happens. The story might be unimaginably complicated, and it might even involve countless parallel universes, with different versions of you living alternative lives. But there would still be a complete story, whether or not anyone could read it.

Must reality admit such a story?

In his influential 1814 philosophical essay on probability, Pierre-Simon de Laplace imagined a “vast intelligence” who, armed with complete knowledge of the present state of all objects and the forces acting between them, could predict the entire history of the universe. For such an intelligence “nothing would be uncertain”, he said, “and the future, like the past, would be present before its eyes”. Probabilities, Laplace continues, are needed only due to our ignorance of what is known by the absolute cosmic narrator.

Quantum mechanics complicates Laplace’s story. The quantum state of an isolated system still evolves deterministically and reversibly in time, according to an equation developed by Erwin Schrödinger more than 100 years ago. However, in the standard view, probabilities in quantum theory are not mere lack of information: complete knowledge of the quantum state does not allow us to predict the outcomes of every possible observation we can make. What happens to a system between observations, and how to describe the process of observation itself --- the measurement problem --- remains a hotly contested issue in the foundations of quantum mechanics.

But even a genuinely random universe could have one complete account of what actually happened. To understand the possible narratives available for a cosmic narrator to tell, let’s populate our story with some characters. The story follows the scenario of a paper with collaborators Howard Wiseman and Eleanor Rieffel. It is an extended version of the paradox of “Wigner’s friend”, proposed by Eugene Wigner in 1961 to illustrate the difficulties associated with the measurement problem.

An observer inside the computer

Our first character is Alice, a quantum computer programmer. She is running an artificial intelligence algorithm, much like today’s large language models, called Quall-E. Turning this story into reality would require a quantum computer vastly larger and more reliable than anything that exists today, but as far as the laws of quantum mechanics are concerned, it is a physically possible story.

Imagine Quall-E at a desk in a simulated room. Alice sends Quall-E an image of a live cat, and instructs it to write down what it sees. She reads the reply: “I see an image of a live cat”. She tries a dead-cat image, and Quall-E replies “I see an image of a dead cat”. In these cases, we could say that Quall-E has observed the image --- live cat or dead cat, as the case may be --- and accurately reported his observations to Alice.

Quantum mechanics gives Alice other options. Instead of either a living or dead cat, she can send Quall-E a quantum superposition of the two images, much like in Schrödinger’s famous thought experiment.