Ultimately, it's hilarious to posit Laplace's Demon and then try to define it at all.

The author talks about "red." People with sight have a sense of "red" that blind people don't. Laplace's Demon, though? Nah, that guy couldn't possibly know anything about time, quantum mechanics, or anything else that might blow a giant hole in every one of my arguments. Not possible.

Honestly. Even extant philosophy can point towards versions of the demon that wouldn't be surprised by anything - not even these systems that are "undecidable," because that "undecideability" relies upon McTaggart's A-series time being an objective, cosmic truth. The paper itself concedes that we should be wary of that premise because of those weird quantum mechanics experiments.

How hard is it to posit that Laplace's Demon sees the universe via B-series time instead? Nothing's unknowable then, so long as everything is determinate. By brute force, if nothing else, the demon knows every output of the algorithm.

Now, does the demon not knowing why the algorithm produces those outputs count as a surprise? Maybe? But then I repeat my point from my own top-level comment: how in the heck isn't the human "agent" equally, or more, surprised by their own actions?

If we begin to elevate "Holy shit why did I do that? I don't even know!" to a truth of the determined universe, I think we've found yet another argument against free will. Perhaps some clever chap will come along to redefine "free" (yet again, and again, and again) as "totally unmoored from literally everything," thus raising an immediate contradiction with the "will" part.