Controversial time reversal, any thoughts?

[vlaiv]

Internet is buzzing with new headline: "Scientists reverse time ....".

So far, I have not read the scientific paper, but I did read couple of articles, ranging from regular daily paper types that say nothing more than the bombastic headline does, ones that unsuccessfully try to explain what is going on by failed analogies and others completely refuting anything special is discovered.

From what I managed to gather, this is what has happened (I'm going to use analogy as well :D). It's not time reversal, but it is questioning randomness. Imagine you throw a dice and based on result you decide to either turn left, turn right or walk straight for 10 meters (you can envision more complicated scenarios with changing stride for example based on number on a die).

If you run above scenario for some time in the field, you will move from start position to some end position. Question is, can you "revert" laws of the game, still basing them on random events like dice throws, so when you continue the game you return to original position in finite amount of time?

It looks like that is what scientists did. They let quantum cubits evolve and then at some point when state of system is highly evolved and in the superposition of states, they applied different set of rules and were able to evolve it back to starting state (with 85% success rate for 2 cubit case and 50% for 3 cubit case).

It can be thought of time reversal, but it can also mean that we don't understand random - it's true nature or how it behaves.

Any thoughts?

[Cornelius Varley]

I let you know - 5 minutes ago 

[andrew s]

Not seen this yet but reminds me of Poincare's recurrance theorem and the quantum version.  Loosely it sates given time a system will return to its initial state or very close to it.

Seems this maybe a demonstration of it. Do you have a link to the original paper?

Regards Andrew 

[xtreemchaos]

when i hear my misses talk about which cake shes making time stops and seams to run backwards ? dont tell her.  charl.

[Ben the Ignorant]

Isn't randomness just a fluctuation around an average value?

[clarkpm4242]

If you have a good read of Richard Feynman's QED series of lectures you will note Feynman diagrams that have photons 'arriving' after they have been 'emitted' i.e. 'travelling backwards in time...

[andrew s]

8 minutes ago, clarkpm4242 said:

If you have a good read of Richard Feynman's QED series of lectures you will note Feynman diagrams that have photons 'arriving' after they have been 'emitted' i.e. 'travelling backwards in time...

But Feynman diagrams are a way of doing an approximate calculations in QED each diagram corresponds to a term in a perturbative expansion. They are not real observable entities just a very clever way of doing the maths and keeping track of the terms.

Regards Andrew 

[clarkpm4242]

8 minutes ago, andrew s said:

But Feynman diagrams are a way of doing an approximate calculations in QED each diagram corresponds to a term in a perturbative expansion. They are not real observable entities just a very clever way of doing the maths and keeping track of the terms.

Yes, they do demonstrate the mental gymnastics that are gone through to achieve a workable model e.g. speed of light being an average...

...I occasionally have fleeting moments of understanding, then seems irretrievable.

[beka]

Hi  All,

A link to the article would be good - maybe this is an example similar to the well known Quantum Erasure effect where it seems you can change past events by actions in the future - sort of ?.

Best

[Thalestris24]

I think if it's not a 100% reversal, it's not really reversal

Louise

[vlaiv]

2 hours ago, andrew s said:

Not seen this yet but reminds me of Poincare's recurrance theorem and the quantum version.  Loosely it sates given time a system will return to its initial state or very close to it.

Seems this maybe a demonstration of it. Do you have a link to the original paper?

Regards Andrew 

Don't think that is what is described as happening. From what I gathered it's more like running Schrodinger's equation backwards in time. Once system is in definite state it time evolves into superposition of states. I think paper talks about doing experiment that goes like this: quantum state is prepared and it starts evolving. At some point, environment is changed in such way that it starts "backwards" evolution of system, which then ends up in initial pure state 85% of time. But don't take that what I've just written to be so - this is just my guess based on incomplete picture of what is written in the paper (have not read it myself - just pieced things from various online articles on it).

I think this is the paper:

https://www.nature.com/articles/s41598-019-40765-6

 

[vlaiv]

2 hours ago, Ben the Ignorant said:

Isn't randomness just a fluctuation around an average value?

That just depends on distribution of random variable - some have central tendency, while some don't - uniform distribution for example does not have central tendency, every outcome has same probability of happening.

[AngryDonkey]

It's breaking the 2nd law of thermodynamics!!! We're doomed! ? 

PS: Can I build that perpetual motion machine now?

[andrew s]

11 minutes ago, AngryDonkey said:

It's breaking the 2nd law of thermodynamics!!! We're doomed! ? 

PS: Can I build that perpetual motion machine now?

Well if it was real maybe but the computer added a lot more entropy than the simulation reduced it by! I have a simulation of a perpetual motion pendulum on my PC ?

Regards Andrew 

[vlaiv]

Many comments are in relation of simulation on a computer.

I think people are missing the point - it's not regular computer and some simulation that I can run in reverse. It's quantum computer and computer is not running simulation in both direction. Quantum computer have cubits - which are bit "cells" that can be in |0>, |1> or a|0>+b|1> state

It is true quantum phenomena that drives this machines in the same way electron gets spread and its position and momentum becomes undetermined after it was measured in some exact position - by evolution of the state. I this sense one does not run regular deterministic simulation - one "plays" with quantum systems and "programming" of those systems is different than in classical computers.

[andrew s]

The paper is free here https://arxiv.org/abs/1712.10057.

I am not sure how it being a quantum computer effects the interpretation of the results rather than how the were achieved.  While the state evolution was reversed it was still done in regular time i.e. initial state - evolved state - initial state. 

There are examples of real systems being reversible see for example https://sciencedemonstrations.fas.harvard.edu/presentations/reversible-fluid-mixing. The key as in the simulation is not to have too much dissipation in the system.

I have asked on Physics Forums for views on this as they have some very knowledgeable people there and will report back if I get any insight.

Regards Andrew

[Tiki]

18 hours ago, vlaiv said:

 

It can be thought of time reversal, but it can also mean that we don't understand random - it's true nature or how it behaves.

Any thoughts?

Having not yet read any of the links, there is one thought that springs to mind.

If 'random' is defined as an absence of pattern then a random sequence (of random numbers for example) will be patternless when looked at from left to right, right to left or any other arbitrary ordering.  If there is any pattern encountered then the sequence cannot be random.

I am not sure whether it is relevant here but a random sequence can also be thought of as incompressible.

Mind bending stuff but my hunch is that the arrow of time points in one direction only.

[vlaiv]

5 hours ago, Tiki said:

Having not yet read any of the links, there is one thought that springs to mind.

If 'random' is defined as an absence of pattern then a random sequence (of random numbers for example) will be patternless when looked at from left to right, right to left or any other arbitrary ordering.  If there is any pattern encountered then the sequence cannot be random.

I am not sure whether it is relevant here but a random sequence can also be thought of as incompressible.

Mind bending stuff but my hunch is that the arrow of time points in one direction only.

You are right in saying that random sequence is incompressible, but not in part where you say that it can't contain pattern.

Pattern is something we understand "naturally" because our brain is pattern matching machine, but pattern is something "agreed" upon - i.e. you need a pattern matching machine to recognize a pattern.

When we have random source of numbers for example (1 and 0), then sequence 101001100101110101001111 is equally likely as sequence 111100001111000011110000, provided that oracle has uniform distribution.

We would say that second sequence is highly structured, but that does not tell us if it was product of random process or not. It's probability rather than pattern that defines randomness. This hints to strong relation of randomness to time - as without time we can't asses if something is truly random or not.

 

[Tiki]

10 hours ago, vlaiv said:

You are right in saying that random sequence is incompressible, but not in part where you say that it can't contain pattern.

 

Thanks for calling me up on this. 'Pattern' is too loose a word to use in this context.

 

[beka]

On 15/03/2019 at 21:39, andrew s said:

The paper is free here https://arxiv.org/abs/1712.10057.

I am not sure how it being a quantum computer effects the interpretation of the results rather than how the were achieved.  While the state evolution was reversed it was still done in regular time i.e. initial state - evolved state - initial state. 

There are examples of real systems being reversible see for example https://sciencedemonstrations.fas.harvard.edu/presentations/reversible-fluid-mixing. The key as in the simulation is not to have too much dissipation in the system.

I have asked on Physics Forums for views on this as they have some very knowledgeable people there and will report back if I get any insight.

Regards Andrew

I had a look at the paper which is very technical, I would surmise that you need a thorough understanding of QED to fully understand it. On the other hand if a reversal of state can be a measure of reversal of time then maybe we can say that time was reversed for the localized system studied. And if we physically managed to reverse state on a macro scale, say we confined a fly to a specific environment let it evolve and then reverted everything about the fly and the environment to a previous state then I guess we can say time was reversed for the fly. And if we managed to do it at galaxy scale... I think Superman did it at Earth scale once ?

All the best!