Imagine you're told that among all possible ways to tune a guitar, the "open strings" tuning — no fingers on the fretboard — always produces the lowest possible hum from the body. That's intuitive: do nothing, get the minimum vibration. Now someone shows you a very specific, slightly weird fingering that produces an even quieter hum. Not by much — a whisper quieter — but provably, analytically quieter. That's what this paper does to the vacuum state in quantum phase space. The committed claim: pure Gaussian states (the quantum vacuum being the simplest) do NOT minimize Wigner entropy among all Wigner-positive quantum states. The Wigner entropy conjecture, proposed by Van Herstraeten and Cerf themselves in earlier work, said they do. Now the same authors kill their own conjecture with explicit analytical counterexamples. The mechanism is precise and satisfying. Take a Wigner-negative state — something with negative quasi-probability regions in phase space — and mix it with another state chosen so the mixture restores Wigner positivity. The resulting blend is Wigner-positive but has been sculpted by the negativity of its ingredient. The entropy of this mixed state can be strictly less than the vacuum's Wigner entropy by up to ~10⁻³. The states achieving this can be found arbitrarily close to the vacuum in state space, which is physically striking: the conjecture doesn't just fail at exotic corners, it fails in the neighborhood of the state it was supposed to protect. The physical engine behind this is what the authors call a "subtle interplay between the uncertainty principle and non-Gaussianity." Gaussian states saturate the uncertainty principle and are the natural minimum-entropy candidates — but non-Gaussian states can redistribute their Wigner function in ways that slightly compress entropy without violating positivity. The paper also disproves the stronger Wigner majorization conjecture, which would have implied the entropy conjecture as a special case. Validation is purely analytical — this is a mathematical disproof, not a simulation or experiment. The counterexamples are constructed explicitly, with closed-form expressions for the states and their entropies. There is no numerical optimization to question. The violation also extends to Wigner-Rényi α-entropies for 0 < α < 2, broadening the result beyond Shannon entropy. Notably, the authors credit AI tools for helping develop the key idea, building on prior characterizations of extreme non-negative Wigner functions. This is an honest and increasingly common acknowledgment in mathematical physics — the AI didn't prove the theorem, but it apparently helped identify the structural opening. The paper is compact (5+12 pages, 1 figure, 3 tables) and decisive. It closes one question cleanly but opens a harder one: if Gaussian states don't minimize Wigner entropy, what does? The true minimum remains unknown, and characterizing it will require understanding the full geometry of Wigner-positive states — a problem that may be substantially harder than the conjecture it replaces.
