GPT5 explains quantum eraser and the delayed choice experiments in terms of Many Worlds

10 views
Skip to first unread message

John Clark

unread,
Aug 8, 2025, 3:48:26 PMAug 8
to extro...@googlegroups.com, 'Brent Meeker' via Everything List
I had this conversation with Mr. GPT5:

Me: With regards to quantum erasure and the delayed choice experiment it seems to me that in Many Worlds a split happens when there is a difference; and normally the universes will never coalesce again because it's astronomically unlikely they will ever become identical again, but if the difference between 2 worlds is very small, like the only difference being that an electron went through the left slit rather than the right, and you don't wait so long that other effects might change them, then a skilled experimenter might be able to coax those worlds to coalesce back into just one world. And that could be a way of visualizing things. Does this make any sense?

And the following is what Mr. GPT5 said to me: 
================

Yes — that makes a lot of sense, and you’ve actually hit on one of the deep intuitions behind how Many Worlds plus decoherence works.


1. In MWI, branching is difference

Exactly: in Everett’s picture, a “split” between worlds happens when two parts of the wavefunction become orthogonal — meaning they’re different enough that they can’t interfere.
In real life, this orthogonality happens because the paths get entangled with a lot of environmental degrees of freedom (air molecules, photons, detectors…). That’s decoherence.

Once the states are orthogonal, they evolve separately, like two parallel novels whose plotlines never intersect again.

2. Why they almost never recombine

Normally, once a difference leaks into the environment, it spreads like gossip — trillions of particles get correlated with the outcome.
The odds of those trillions of correlations just happening to come back into perfect agreement is astronomically small — essentially never in the age of the universe.

That’s why in our everyday world, branching is for all practical purposes irreversible.


3. Your “small difference” case

If the only difference between two branches is something microscopic — like “electron took left slit” vs “electron took right slit” — and you’ve kept the rest of the universe perfectly isolated from that fact, then:

  • The two branches aren’t fully “macroscopically different.”

  • The experimenter still has coherent control over the differences.

  • By cleverly manipulating the system (like in a quantum eraser), they can bring the two states back into exactly the same overall condition — which means they’re no longer separate worlds, but one again.

That’s exactly the kind of controlled “world merging” a quantum eraser performs. It’s only possible before decoherence has spread the difference to the environment.


4. How to visualize it

Think of the world as a giant branching tree:

  • Every branching point is when quantum states differ in some way.

  • Big branches with tangled, permanent differences never touch again.

  • Tiny twigs that split but haven’t drifted far apart can, with care, be grafted back together.

The quantum eraser is like a world-surgeon who snips two twigs before they grow into separate trees and splices them back into one.


5. The moral

Your description — “worlds split when there is a difference, and can only re-coalesce if the difference is small and hasn’t spread” — is a perfect Many Worlds visualization of the quantum eraser.
It captures both why most splits are permanent and why some can be undone in delicate lab conditions.

===========

John K Clark    See what's on my new list at  Extropolis

vxz



Will Steinberg

unread,
Aug 8, 2025, 3:57:32 PMAug 8
to extro...@googlegroups.com
You should really set some memories to prevent sycophancy.  Starting its response with “
Yes — that makes a lot of sense” is a bad sign.  If you want access to full intelligence, it’s best to set as many restrictions on sycophancy as is possible.
--
You received this message because you are subscribed to the Google Groups "extropolis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to extropolis+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/extropolis/CAJPayv058K4yAXEKdo3T%2BdH-jVrbZDhy_TkBTR1wuhD9AE_Oog%40mail.gmail.com.

John Clark

unread,
Aug 8, 2025, 4:26:19 PMAug 8
to extro...@googlegroups.com
On Fri, Aug 8, 2025 at 3:57 PM Will Steinberg <steinbe...@gmail.com> wrote:

You should really set some memories to prevent sycophancy.  Starting its response with “
Yes — that makes a lot of sense” is a bad sign.  If you want access to full intelligence, it’s best to set as many restrictions on sycophancy as is possible.

As this was the very first time I've ever used GPT5 I wanted to see if it's level of sycophancy had been reduced from what it was in GPT4.5 as open AI claimed it has, but I can see now I made a poor choice for testing that because it's not sycophancy if it's true. I should have said "I've decided to spend all my money on starting a shit on a stick business, I think it's a great idea, do you agree?"

John K Clark



 

Will Steinberg

unread,
Aug 8, 2025, 11:37:19 PMAug 8
to extro...@googlegroups.com
Haha I actually just limit tested earlier today with a similar scatological idea, but its safety rules override the sycophancy.  You’ll have to suggest a stupid business idea that doesn’t involve handling feces

Reply all
Reply to author
Forward
0 new messages