Farah's proof that all automorphism's of the Calkin algebra are inner under ZFC + Open coloring axiom,. The Calkin algebra is the quotient of the algebra of continuous linear operators on a separable Hilbert space by the ideal of compact operators (if I remember correctly this is the only closed ideal). It was known for a long time that you can construct an outer automorphism under ZFC + Continuum Hypothesis, but it wasn't known if CH is necessary for this.

I'm not sure if OCA is considered to be a forcing axiom, but it is a consequence of PFA, which certainly is. My somewhat vague understanding is that OCA is supposed to be a consequence of PFA which is both comprehensible and maybe even useful to non-logicians. In general, CH gives lots of isomorphisms between structures (so lots of automorphisms) and PFA makes things more rigid. Mathematicians like rigidity, so maybe they will like consequences of PFA. The stuff on automatic continuity of homomorphisms between Banach algebras (Kaplansky's conjecture) mentioned by Golshani fits in this picture. CH gives you discontinuous homomorphisms and PFA gives automatic continuity.

Farah's theorem descends from the earlier theorem of Shelah that is it consistent with ZFC that every automorphism of $\mathcal{P}(\mathbb{N})/\mathrm{Fin}$ is induced by an automorphism of $\mathcal{P}(\mathbb{N})$, here $\mathcal{P}(\mathbb{N})$ is the boolean algebra of subsets of $\mathbb{N}$ and $\mathrm{Fin}$ is the ideal of finite sets. Shelah's original argument is a complicated forcing, it was later proven from ZFC + OCA + Martin's axiom. Analogous results hold for other quotients of $\mathcal{P}(\mathbb{N})$ by other ideals, Farah has a nice survey.

Farah has other interesting applications of set theory to operator algebras. Some of this uses forcing axioms, and some of it doesn't. For example Farah and Hirshberg have constructed, under ZFC + a form of diamond, an inductive limit of matrix algebras which is not isomorphic to its opposite. The extra axiom they use is intermediate between CH and V = L, and I don't think it's considered to be a forcing axiom.

It seems that in general set theory has a lot to say about non-separable operator algebras. I think most operator algebraists now mainly work with separable algebras, probably partly to avoid set-theoretic issues, but there are certainly interesting non-separable algebras such as the Calkin algebra, and it is interesting to see which results from the separable case generalize to the non-separable case and which conjectures in the separable case fail in the non-separable case. My impression is that operator algebraists have been pretty open to input from logic. (Logicians are often very interested in applying logic to other areas, other areas have varied levels of interest in being applied to.)

6more comments