I knew Kolmogorov complexity was used in Solomonoff induction, and I was under the impression that using Universal Turing Machines was an arbitrary choice.
Solomonoff induction is only optimal up to a constant, and the constant will change depending on the language.
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
If continuing the discussion becomes impractical, that means you win at open threads; a celebratory top-level post on the topic is traditional.