[ExI] Cellular automata, downward causation, and libertarian determinism

[Jason Resch]

On Sun, Oct 26, 2025 at 11:04 AM Giulio Prisco via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

> On Sun, Oct 26, 2025 at 3:56 PM Jason Resch via extropy-chat
> <extropy-chat at lists.extropy.org> wrote:
> >
> >
> >
> > On Sun, Oct 26, 2025 at 10:34 AM Giulio Prisco via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
> >>
> >> On Sun, Oct 26, 2025 at 3:23 PM Jason Resch via extropy-chat
> >> <extropy-chat at lists.extropy.org> wrote:
> >> >
> >> >
> >> >
> >> > On Sun, Oct 26, 2025, 3:05 AM Giulio Prisco via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
> >> >>
> >> >> On Sat, Oct 25, 2025 at 1:15 PM Jason Resch via extropy-chat
> >> >> <extropy-chat at lists.extropy.org> wrote:
> >> >> >
> >> >> > Can downwards causation not exist in a universe whose laws are
> fully deterministic?
> >> >> >
> >> >> > Consider that the neurons push and guide molecules
> (neurotransmitters) around as much as the molecules define the operation of
> neurons.
> >> >> >
> >> >> > The illusion that causes work only bottom-up I think stems from
> taking reductionsm too far.
> >> >> >
> >> >> > If we can do readily admit that molecules tell neurons how to
> behave, why are some so reluctant to admit that neurons can also tell
> molecules how to behave?
> >> >> >
> >> >> > I think Roger Sperry gives the best treatment of this in his
> "Mind, Brain, and Humanist Values" (
> https://www.informationphilosopher.com/solutions/scientists/sperry/Mind_Brain_and_Humanist_Values.html
> )
> >> >> >
> >> >> > Jason
> >> >> >
> >> >>
> >> >> Hi Jason. I think it's quite simple: the cosmic operating system
> >> >> chooses initial conditions, but is not limited to choose all of them
> >> >> at the same time. Rather, the cosmic operating system chooses
> >> >> "initial" conditions distributed in time, just like in my little
> game.
> >> >> To us, observers limited by time, the universe appears not
> >> >> deterministic. This is equivalent to Nicolas Gisin's intuitionist
> >> >> physics. OK, the universe is still deterministic to an observer that
> >> >> looks at the global history from outside time. But I think the cosmic
> >> >> operating system doesn't make enough choices to fix one unique
> >> >> timeline, and this is another, ontological source of nondeterminism.
> >> >
> >> >
> >> > I agree this is a viable answer to the question of no determinism in
> physics. If reality consists of a sufficiently large multiverse, then we
> can expect a vast plenitude of parallel universe states which are almost
> the same, but different only by the position of a single particle, for
> example.
> >> >
> >> > Such parallel universes contain identical copies of brain states of
> all the conscious observers, however, when any one of those observers
> attempts to deeply probe into the state of one of those particles, they
> will find it exists in an indeterminate superposition of many
> possibilities, for the reason that this same observer mind state exists
> also in all those parallel similar universes, where the particle is doing
> something a little different.
> >> >
> >> > But once the measurement is made, the observer's mind state changes
> in a way that partitions the set of similar but not quite identical
> universe she is a part of. The observer is said to have "collapsed the wave
> function" but really, she has only adjusted her knowledge in a way that
> changes the set of universes her mind state is still cool compatible with.
> >> >
> >> > There will be many such partitionings corresponding to each of the
> possible outcomes of her measurement, and from her perspective before
> making the measurement, the outcome will seem totally random and
> unpredictable, and it was, because she is never in a position to know
> exactly which universe is in (as her current mind state exists identically
> within many different universes).
> >> >
> >> > But all this is aside from my original point, which I don't believe
> downwards causation requires indeterminism.
> >> >
> >>
> >> In fact, in the OP and in my first comment (besides the last point on
> >> multiple timelines) I refer to a deterministic universe with one
> >> single timeline AND downward causation. But I still think that a
> >> universe with multiple timelines provides a stronger type of
> >> nondeterminism.
> >
> >
> > Ahh I see, I understand your point more fully now.
> >
> >>
> >>
> >> >
> >> >> More in my last book (I think I sent it to you?).
> >> >> G.
> >> >
> >> >
> >> > What is the title? I do have "Tales of the Turing Church." Or are you
> referring to a more recent book?
> >> >
> >>
> >> This one:
> >> https://www.turingchurch.com/p/irrational-mechanics
> >> You are mentioned in the acknowledgments, because you made useful
> >> observations on this list.
> >
> >
> > Oh wow, I appreciate that. I have just ordered the paperback version. ;-)
> >
>
> Thank you! See the discussion on randomness in digital computers
> (Chapter 12). I had it in a first draft, then I removed it (who knows
> why?), then I put it back after a discussion on this list where you
> also made exactly the same point.
> G.
>
>
Excellent. I somewhat remember this. RNGs and PRNGs are a bit of a
fascination of mine. Note that modern chips by Intel, etc. now include an
instruction RD_RAND which lets the computer read bits ultimately produced
by a hardware RNG (I believe it is a thermo sensor type in practice), which
ultimately could be traced to underlying statistical variations quantum
mechanical in nature. (I think someone made an argument that even a coin
flip by a human can be shown to be ultimately quantum-mechanically random,
due to how the exact force of the flip, and interaction with air molecules,
etc. influence the final result).

But I don't think there is any specialness of quantum mechanically random
events. Let's say you get a quantum mechanical random value 1 (instead of
0) and use that to drive some process. Is it any different than had you
forked the process, and provided 1 to one instance, and 0 to the other
instance? I don't see how it leads to any difference, and indeed there is
some result that shows what can be computed by a probabilistic Turing
machine is no different than what can be computed by a deterministic Turing
machine. Do the forked processes not have free will, while the quantum
determined random variable process does have free will? I am not sure I see
that.

Jason
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20251026/1011896a/attachment.htm>