(working on (pdf) this freedom in the machine piece from Scott Aaronson)

question 1: what if freedom at the macro level is an operator-specific equivalent to the uncertainty principle at the micro level? Scott dismisses Heisenberg uncertainty (the particle inside a volume dXdP = is indeterminate) as fundamentally distinct from the freedom operation/definition. What if they're not? What if the operator has its own uncertainty principle on the space of the whole of obervable spacetime? This probably doesn't change any of the logic, but it might have interesting mathematical uses. (And yes, I read Scott as basically saying he disagrees with this, but as a handwave in this particular essay. Not a big deal really, since there's as yet no reason to care about the distinction, nor a way to test it if there were.)

observation: can we just point out that the quantum perturbation (the inevitable entanglement with observable-system at the quantum level) is simply that the act of measurement unavoidably introduces a new force to the system? And that, for small enough systems, that force matters? I.e. if I hit an electron with another electron (or photon or whatever) *it's going to have an effect on the target that I may not be able to account for*? Please? Over and over again, until we come up with a better description to replace quantum mechanics (unlikely as that is)? It's really not mystical.

question 2: The human brain is a quantum computer in the trivial sense. Are not atoms quantum mechanical in nature? Does this imply anything regarding the quantum to classical transition involved in the conjecture? Generally, specifically? Less trivially, I think that the classical limit of a quantum system being tossed aside so readily is a result of the assumptions in question 1. Likely a trivial one in the case where whatever the boundary conditions are have a ready transformation from the relevant quantum problem to the classical problem.

observation 3: The discussion of "if we have brain uploads to external computational hardware, now philosophical experiments have a very real and immediate consequence (wait, isn't that what religion, political ideologies, etc do in the observable history?)" has prior art, the one that occurs to me is the particular branch of faster-than-light travel that posits that the human mind must shape whatever the mechanism is (see L. E. Modesitt, but there are many others, think of it as the human mind needing to derive a brand new equation each time the ship needs to "make the jump past lightspeed", Frank Herbert=> Dune and the Navigator's Guild, etc)

final: the final paragraph of the conclusions is a wonderful list of the real meat of the thing. And one I don't have any arguments with. My quibbles here don't change the questions that matter. I might have more sympathy for (limited forms of) some of the questions Scott dismisses e.g. quantum computers vs. classical computers, in the following narrow sense, that there's an argument to be made that the quantum computer is roughly similar to using a simulation to sample the configuration integral or the partition function, in that it's a reduction of an n^{6m} problem to something like an 6mn problem, or basically an empirical necessity for some problems but not otherwise immediately "the all encompassing solution to all problems everywhere" that the hype for quantum computers in some quarters would have us embrace. Again this is a limited quibble, not a general argument.