But the Many-Worlders don’t need to take this lying down. They could respond, for example, by pointing to other, more specialized communication problems, in which it’s been proven that Alice and Bob can solve using exponentially fewer qubits than classical bits. Here’s one example of such a problem, drawing on a 1999 theorem of Ran Raz and a 2010 theorem of Boaz Klartag and Oded Regev: Alice knows a vector in a high-dimensional space, while Bob knows two orthogonal subspaces. Promised that the vector lies in one of the two subspaces, can you figure out which one holds the vector? Quantumly, Alice can encode the components of her vector as amplitudes—in effect, squeezing n numbers into exponentially fewer qubits. And crucially, after receiving those qubits, Bob can measure them in a way that doesn’t reveal everything about Alice’s vector, but does reveal which subspace it lies in, which is the one thing Bob wanted to know.
So, do the Many Worlds become “real” for these special problems, but retreat back to being artifacts of the math for ordinary information transmission?
To my mind, one of the wisest replies came from the mathematician and quantum information theorist Boris Tsirelson, who said: “a quantum possibility is more real than a classical possibility, but less real than a classical reality.” In other words, this is a new ontological category, one that our pre-quantum intuitions simply don’t have a good slot for. From this perspective, the contribution of quantum computing is to delineate for which tasks the giant amplitude wave acts “real and Many-Worldish,” and for which other tasks it acts “formal and Copenhagenish.” Quantum computing can give both sides plenty of fresh ammunition, without handing an obvious victory to either.
So then, is there any interpretation that flat-out doesn’t fare well under the lens of quantum computing? While some of my colleagues will strongly disagree, I’d put forward Bohmian mechanics as a candidate. Recall that David Bohm’s vision was of real particles, occupying definite positions in ordinary three-dimensional space, but which are jostled around by a giant amplitude wave in a way that perfectly reproduces the predictions of quantum mechanics. A key selling point of Bohm’s interpretation is that it restores the determinism of classical physics: all the uncertainty of measurement, we can say in his picture, arises from lack of knowledge of the initial conditions. I’d describe Bohm’s picture as striking and elegant—as long as we’re only talking about one or two particles at a time.
My conclusion is that, if you believe in the reality of Bohmian trajectories, you believe that Nature does even more computational work than a quantum computer could efficiently simulate—but then it hides the fruits of its labor where no one can ever observe it. Now, this sits uneasily with a principle that we might call “Occam’s Razor with Computational Aftershave.” Namely: In choosing a picture of physical reality, we should be loath to posit computational effort on Nature’s part that vastly exceeds what could ever in principle be observed. (Admittedly, some people would probably argue that the Many Worlds interpretation violates my “aftershave principle” even more flagrantly than Bohmian mechanics does! But that depends, in part, on what we count as “observation”: just our observations, or also the observations of any parallel-universe doppelgängers?)