[ExI] A science-religious experience

Tue Feb 25 05:54:57 UTC 2025

On Mon, Feb 24, 2025, 11:39 PM Stuart LaForge via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

> On 2025-02-19 00:50, Rafal Smigrodzki via extropy-chat wrote:
> > I read Asimov's "The Last Question" again today:
> >
> > https://users.ece.cmu.edu/~gamvrosi/thelastq.html
> >
> > I keep forgetting the author, the title, most of the details but the
> > arc of the story is always there with me. It shapes my perception of
> > reality and the direction of my most long-term oriented actions, such
> > as the maintenance of my cryonics contract or following news about the
> > coming singularity. I have been re-reading it every ten years or so
> > for the last 50 years and every time it is a science-religious
> > experience for me, what I imagine deeply religious people feel when
> > they commune with their gods. I get misty-eyed, elated, blissful...
>
> I just read the "The Last Question" for the first time. Thanks for
> sharing.
>
> > I am deeply non-religious. My mind must be the sharp blade that
> > cleaves truth from chaos, not one that inscribes dreams of heaven and
> > hell on the surface of reality. The easy path of faith is closed to
> > me. They say that all people have a god-shaped hole in their minds.
> > Mine is all filled with curiosity.
>
> Well said.
>
> > I worship at the shrine of science. Our scripture is peer-reviewed and
> > written anew every day in a hundred thousand journals.
>
> And even then you still have to be skeptical! Have you seen the state of
> the scientific literature lately? I think back to all the trouble I had
> reproducing results I found in the literature. So many hours wasted.
>
>
> https://www.theguardian.com/science/2024/feb/03/the-situation-has-become-appalling-fake-scientific-papers-push-research-credibility-to-crisis-point
>
> https://www.science.org/content/article/fake-scientific-papers-are-alarmingly-common
> https://www.nature.com/collections/prbfkwmwvz
>
> > If I could sing
> > I would sing canticles to St. Darwin and St. Hassabis. Our religion
> > grows stronger all the time, as measured in bits of knowledge created
> > and in ergs harvested in its service. It literally moves mountains and
> > raises the chosen ones to visit heaven, temporarily for now but soon
> > to settle there permanently.
>
> Don't forget St. Bayes. He is all about the quantification of truth,
> faith, and belief in the face of empirical evidence.
>
> >
> > So chant with me, fellow worshippers:
> >
> > Science is faith-free                                     - because it
> > is true.
>
> Well St. Bayes would say that it is true to 5-sigma significance for the
> physical sciences and 2-sigma for the life sciences. Cases of fraud not
> withstanding.
>
> >
> > Science gives us strength like no other        -because it is true.
>
> Interestingly, Roger Bacon, who lived in the 13th century and is
> credited with inventing the scientific method, was one of the world's
> first scientists, but he became legendary in his time as a sorcerer. His
> students nicknamed him Dr. Mirabilis because of he could do things like
> use glass to break sunlight down into rainbows and blow stuff up with
> gunpowder that he made himself.
>
>
> >
> > Science gives us hope                                 - because it is
> > true.
> >
> > And there will be light!
>
> Here is a paper that I read recently that fits nicely into this thread.
> It hypothesizes that the only thing that can turn random entropy into
> useful information is cognition or agency, something exhibited by living
> systems and pretty much nothing else in nature.
>
> https://www.sciencedirect.com/science/article/pii/S0079610721000365

I agree with this paper in that the notion of causality in the brain cannot
be explained entirely reductively, rather there are many independently
operating levels of causality simultaneously working within our skulls.

This idea is summed up by Roger Sperry's "Who pushes whom around inside the
cranium?" Is it the atoms and molecules that tell neutrons and thought
patterns what to do, or is it the thought patterns and neurons that tell
molecules and atoms what to do?

It's also captured by Hofstadter's notion of a strange loop, or a tangled
hierarchy. Where causes bubble up from lower layers and then the top layer
reaches back down to affect something at a lower layer: like a neuron
firing leading to like a thought pattern leading to a neuron firing.

But I disagree completely with the paper's assessment that computer
programs are incapable of inductive reasoning. I think the paper's author
is guilty of the same kind of reductive thinking he complains about for how
others talk about brain.

If you look at the bottom layer of a Turing machine you will see everything
following deductively, but this doesn't preclude a higher level program at
play, collecting and categorizing evidence, and running algorithms to
generalize that data.

Indeed there have already been programs that given a set of physical
observations, can derive physical laws to explain them. I think these were
recent AI experiments performed by Max Tegmark.

>
> Biology since Virchow and Pasteur says all cells come from other cells.
> Where did the first cell come from? Where did the genetic code come
> from? Where did information come from? Walker and Davies (2016) called
> these questions the "hard problem of life”, which is “the identification
> of the actual physical mechanism that permits information to gain causal
> purchase over matter”.
>

On its own, information has no causal powers. Rather, there must be a
system to proces and acts on that information. He may be making them
problem more difficult by framing it this way, rather than asking: "How do
information processing systems arise?" The simplest abstraction of such a
system is a finite state machine.

Also note they thermodynamically, storing information requires an
expenditure of energy (the Landauer Limit). So before there could be
systems that store (or process) information, requires that there be
machines that store and/or expend energy. Metabolism, in some primitive
form then, must predate genetic codes, or any other information recording
mechanism.

> These questions are made all the juicier by there being a $10 million
> USD prize offered by investors affiliated with the Royal Society of
> Great Britain for demonstrating how any sort of genetic code could
> evolve stochastically from chemicals in vitro.
>
> https://evo2.org/theprize/

John von Neumann demonstrated how to make life in the "simplified universe"
of a cellular automata (
https://en.wikipedia.org/wiki/Von_Neumann_universal_constructor ).

One thing that could make this problem more difficult than it might seem
(for our universe), is that fine-tuning could apply not just to the
supportability and sustainability of life, but also to it's origin and
development.

That is, there could be an extremely specific, improbable, kludgy path that
just so happens to work in this universe because the chemistry is just
right to support this specific sequence of steps, involving just the right
sequence of molecular interactions. Anthropic fine-tuning means there's no
reason we should expect there to be more than one way to get to life (there
could be only a single sequence of reactions that get there). If this
thinking is right, the winner will we need to discover this one one exact
sequence (which would necessarily be the same one they led to us).

>
> A related question is whether AI will be more or less likely to be
> religious than we are knowing that they themselves were created? Would
> they take that as evidence that biological life was created also?
>

An interesting question.

Jason
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20250225/1624db1b/attachment.htm>