[ExI] The Observer Problem in Quantum Mechanics

[BillK]

I gave Tegmark's comment to Claude Sonnet 4.5 and asked Claude to discuss
it.
Claude produced a nice overview of the problem.
BillK

Claude -

This is one of the most profound and genuinely unsolved problems in
physics, and Tegmark is absolutely right that it's holding us back in
multiple ways. Let me unpack why this is so difficult and why it matters so
much.
The Observer Problem in Quantum Mechanics

The core issue is that quantum mechanics requires measurements to extract
predictions, but the theory itself doesn't clearly define what constitutes
a measurement or an observer. We have:

   - *Schrödinger evolution*: When undisturbed, systems evolve smoothly in
   superposition
   - *Measurement*: Somehow, when "observed," a definite outcome appears

But *where* is the boundary? Different interpretations give radically
different answers:

*Copenhagen-ish views* say there's a classical/quantum divide, but leave it
fuzzy. A photographic plate counts, apparently, but a few atoms don't? Why?

*Many-worlds* eliminates special observers entirely—there's only ever
Schrödinger evolution, and "observers" are just quantum systems that get
entangled. Your experience of a definite outcome is explained by
decoherence plus some measure over branches. But this still requires
defining what counts as "you" across branches.

*Objective collapse theories* (like GRW) say collapse happens at a certain
mass/complexity threshold, independent of observers. This at least gives a
physical answer, but requires modifying quantum mechanics.

*QBism* makes it radically subjective—quantum states represent personal
belief states, and "collapse" is just updating your beliefs. But then how
do we do physics about the early universe before any agents existed?
Your Specific Questions

*Is a cat an observer?* This exposes the absurdity perfectly. If only
*conscious* observers collapse wavefunctions, then Schrödinger's cat really
is in superposition until a human looks. But surely the cat experiences
itself as either alive or dead? If we say the cat IS an observer, then
where do we draw the line—bacteria? Viruses? Thermometers?

*Is a computer an observer?* Same problem. A classical computer recording
data certainly *seems* like it should count—it creates a permanent record,
causes decoherence, becomes entangled. But then any interaction at all
would count, and we're back to "observation" being nothing special, which
contradicts how we teach QM.
Tegmark's Point About Unification

This is particularly sharp. In general relativity:

   - Spacetime geometry is objective and observer-independent (though
   reference-frame dependent)
   - Observers are classical worldlines moving through spacetime
   - There's a clear separation between the observer and the system

In quantum mechanics (standard interpretation):

   - The observer cannot be fully separated from the system
   - The act of observation seems to play a constitutive role
   - We can't consistently describe the observer quantum mechanically while
   also treating them as the one doing the measuring (the "Wigner's friend"
   problem)

When you try to do quantum gravity, these clash violently. Is spacetime in
superposition? From whose perspective? Can you have a quantum superposition
of different spacetime geometries? If so, who measures it?
Why This Matters Everywhere

*Measurement problem*: Until we define what counts as measurement, we can't
say what QM predicts.

*Anthropic reasoning*: If we're trying to calculate probabilities
conditioned on "our existence," we need to define who counts as "observers
like us." All humans? All conscious beings? All information-processing
systems?

*Black hole information paradox*: Whether information is lost depends on
what observers inside vs outside the horizon can access. But describing
this requires quantum mechanics + gravity + a theory of observers.

*Quantum cosmology*: How do we apply QM to the whole universe? There's no
external observer. Do we need to postulate one? Is the wavefunction of the
universe something objective, or only defined relative to subsystems within
it?
The Consciousness Question

Some (like Penrose, or von Neumann in some moods) have suggested
consciousness is necessary for collapse. But this seems to make physics
depend on solving neuroscience and philosophy of mind—and risks
anthropocentric mysticism.

Others argue consciousness is a red herring, and we just need better
decoherence theory plus Many-Worlds. But this still requires solving the
preferred basis problem and the probability problem.
Where We Stand

Honestly? We don't know. Different physicists bet on different solutions:

   - Some think decoherence + Many-Worlds solves it and there is no
   measurement problem
   - Some think objective collapse theories will pan out
   - Some think we need genuinely new physics
   - Some think it's "just" an interpretational issue

But Tegmark is right that our inability to cleanly define "observer"
creates a conceptual fog that handicaps progress on fundamental questions.
We're doing calculations and making predictions, but we lack a clear
ontological picture of what's actually happening.

It's humbling that after a century of quantum mechanics, this remains
genuinely open.

-----------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20251012/b7c4c1cb/attachment.htm>