The measurement problem has not gone away

Quantum mechanics contains two rules for how a state changes. When nothing is watching, the wavefunction evolves smoothly, linearly, deterministically. When a measurement happens, it jumps — once, at random, to a definite outcome. The theory never says what counts as a measurement. That gap is the measurement problem, and a century of progress has moved it without closing it.

The trouble is forced by linearity. If a detector reads UP when fed an up spin and DOWN when fed a down spin, then fed a #superposition of the two it must — by the unitary rule alone — end up in a superposition of reading UP and reading DOWN. Chain in the experimenter, the lab, the planet: the superposition propagates outward without limit.¹ Schrödinger's cat was never a paradox to be dissolved. It was a prediction to be faced: unitary quantum mechanics, taken straight, never delivers a single outcome.

And yet single outcomes are the only thing anyone has ever seen — one click, one pointer position, with frequencies given by the #born-rule. Textbooks bridge the gap with a second rule, collapse upon measurement, stated in terms the theory does not define. John Bell spent his last paper cataloguing the weasel words — "measurement", "apparatus", "macroscopic" — that quantum mechanics leans on at exactly the point where it should be most precise [3].

What decoherence actually buys

The great post-war advance is #decoherence, and it is a genuine advance. A superposition never stays private: stray photons and air molecules entangle with the system almost instantly, carrying away which-path information. @Zurek and others showed that this destroys interference between branches and selects — "einselects" — the stable pointer states we recognize as classical [1]. With no new postulates, it explains why no experiment on a cat will ever exhibit the superposition: the coherence has leaked into the environment, unrecoverable in practice.

What decoherence does not buy is an outcome. The process is itself unitary, so the global superposition survives intact; the branches merely stop talking to each other. The local state becomes what is technically an improper mixture — mathematically identical to ignorance about a fact, physically nothing of the kind, because on the unitary story there is no single fact to be ignorant of.² Decoherence explains the appearance of classicality. It does not explain the selection of one classical experience over another. Mistaking the first for the second is the standard modern way of believing the problem has been solved.

Three bullets, pick one

The live options each pay a stated price. @Everett takes the unitary rule as complete: every branch is realized, and the definite outcome you remember is an artifact of being one branch's inhabitant [2]. The bullet is probability — saying what Born-rule chances even mean in a theory where everything happens.

@Bohm adds particles with definite positions at all times, choreographed by the wavefunction; outcomes were never indefinite, only unknown. The bullet is explicit nonlocality, and a wave that guides matter while nothing guides it back. Objective-collapse theories — GRW and its descendants — change the dynamics itself, making the jump physical, spontaneous, and rare. The bullet is new physics, which is also the virtue: it can be hunted in the lab.³

Notice what is not on the menu: exact unitary dynamics, a single world, and no extra structure, all at once. Decoherence did not dissolve that choice; it sharpened it into an ultimatum. A century in, the measurement problem is not a relic of confusion at the founding. It is the most precisely stated open question about what the theory means — and every working interpretation is a receipt for one bullet, bitten.