The paper[1] claims that MIP* = RE. At first I wasn't sure whether that was true, because the paper is too long and complicated for me to verify.
But then I found a series of questions and answers written by this quantum guy Scott Aaronson[2] and by this other quantum guy Kenneth Regan[3], who both know a bunch of quantum stuff. Each one of their blog posts makes me more certain about MIP* = RE, to the point that now I'm 99.9% sure it's true, even though I could never verify it myself.
They „denote by MIP the class of languages that have multiprover interactive proof systems,“ and by „MIP∗, the ‚entangled-prover‘ analogue of the class MIP. Informally the class MIP∗, first introduced in [CHTW04], contains all languages that can be decided by a classical polynomial-time verifier interacting with multiple quantum provers sharing entanglement.“
But then I found a series of questions and answers written by this quantum guy Scott Aaronson[2] and by this other quantum guy Kenneth Regan[3], who both know a bunch of quantum stuff. Each one of their blog posts makes me more certain about MIP* = RE, to the point that now I'm 99.9% sure it's true, even though I could never verify it myself.
[1] https://arxiv.org/abs/2001.04383
[2] https://www.scottaaronson.com/blog/?p=4512
[3] https://rjlipton.wordpress.com/2020/01/15/halting-is-poly-ti...