A Universe made by a Turing-like Machine

[martillo]

Inspired in the thread "Nothing can come from nothing so something always existed!" at the Physics' Forum I'm exploring the idea that what could have ever existed is a Turing-like Machine running over a limited discrete 3D Space. 
And thinking in an entire Universe evolving from something very simple I think in an evolving Turing-like Machine from may be a very simple binary automata to a more complex one with time over an also evolving 3D Space.
The elementary elements of the 3D Space would have some attributes which would let define elementary particles in some places. The elementary particles would obey some rules which would constitute the physics' laws of the Universe. The Turing-like Machine applies then the Physics' Laws over the particles. The elements' attributes and the rules (the physics' laws) could have also evolved becoming more complex with time.

Of course the machine would operate at may be an unbelievable huge speed.

I'm just exploring the idea and any comment is wellcome.

I must mention that my knowledge on the Turing Machine is very limited. I know just the basic principles of operation.

Edited January 11, 2021 by martillo

[Halc]

A Turing machine runs over a 1D space, not a 3D one.

It isn't a binary machine (except for the tape it reads). It is an n-state machine which can in principle run without numbering the states.

A Turing machine is also painfully inefficient. It takes trillions of steps (clock cycles) to complete what a modern processor can do in one clock cycle.

You seem to have a model of an emulation of a universe, which does not require any particular speed at all, but it does require a finite data set, and our universe does not seem describable with a finite data set.

[martillo]

4 hours ago, Halc said:

A Turing machine runs over a 1D space, not a 3D one.

A finte discrete 3D space can be represented by a 1D list of all of its elementary elements.

4 hours ago, Halc said:

It isn't a binary machine (except for the tape it reads). It is an n-state machine which can in principle run without numbering the states.

I'm thinking in an evolving machine that evolved with the Universe it describes and the initial Universe could have been even a very simple point-like one with a binary state. Both the machine and the Universe would have evolved to the current much more complex ones.

4 hours ago, Halc said:

A Turing machine is also painfully inefficient. It takes trillions of steps (clock cycles) to complete what a modern processor can do in one clock cycle.

I'm thinking in an initial not so efficient but simple Turing-like machine with the capacity of evolve modifying itself to a much more complex and efficient one of course.

4 hours ago, Halc said:

You seem to have a model of an emulation of a universe, which does not require any particular speed at all, but it does require a finite data set, and our universe does not seem describable with a finite data set.

I'm thinking in a limited but increasing Universe with a variable finite number of elementary elements with finite number of attributes like location, velocity, acceleration, energy, etc with a finite number of physics laws over them representing their possible interactions (the forces) and their behavior (F=mA, Momentum and Energy conservation, etc).

Even Super String Theory of the Universe state it could have up to eleven dimensions, still a finite number of them.

Why do you think the Universe would be not describable by a finite data set?

Edited January 11, 2021 by martillo

[Halc]

7 hours ago, martillo said:

Even Super String Theory of the Universe state it could have up to eleven dimensions, still a finite number of them.

Why do you think the Universe would be not describable by a finite data set?

It only takes one unbounded dimension to have an infinite data set. Our universe appears to have at least four. The location of a single particle is not expressible as a finite size number. Try it. What is your absolute location (not relative to something that doesn't have its own known location)?

Quote

A finte discrete 3D space can be represented by a 1D list of all of its elementary elements.

True, but there is no evidence that space is either finite or discreet.

Quote

I'm thinking in an evolving machine that evolved with the Universe it describes

…

I'm thinking in an initial not so efficient but simple Turing-like machine with the capacity of evolve modifying itself to a much more complex and efficient one of course.

 

Then it isn’t a Turing machine, which by definition cannot change itself. I’m just wondering why you’ve chosen a Turing machine. It’s only benefit is that it has no limit to its data set.

Quote

and the initial Universe could have been even a very simple point-like one with a binary state.

No viable model of the universe expresses an initial state of a point. That’s kind of a pop-science naive view. You can double or square the size of a point all you want and it will remain a point. Likewise, you can double or square the size of a finite size thing all you want and it is going to remain finite in size.

Yes, the initial state is expressed as a singularity, but a singularity is just a place where the normal laws of physics don’t apply. It doesn’t mean a point.

Edited January 11, 2021 by Halc

[martillo]

58 minutes ago, Halc said:

It only takes one unbounded dimension to have an infinite data set. Our universe appears to have at least four. The location of a single particle is not expressible as a finite size number

A finite discrete Space could have a grid with locations that could be represented by finite numbers or symbols.

58 minutes ago, Halc said:

Then it isn’t a Turing machine, which by definition cannot change itself. I’m just wondering why you’ve chosen a Turing machine. It’s only benefit is that it has no limit to its data set.

I said "Turing-like" machine. It could be not exactly a Turing one.

The main characteristic of a Turing machine that calls my attention is its simplicity. It would be well suitable for an initially very simple Universe running by a very simple finite automata.

58 minutes ago, Halc said:

No viable model of the universe expresses an initial state of a point. That’s kind of a pop-science naive view. You can double or square the size of a point all you want and it will remain a point. Likewise, you can double or square the size of a finite size thing all you want and it is going to remain finite in size.

I said point-like not exactly a point. The elementary element in a discrete Space is an elementary volume with no zero size. But yes, the initial Universe could be not a single element one. I don't know at this time.

58 minutes ago, Halc said:

Yes, the initial state is expressed as a singularity, but a singularity is just a place where the normal laws of physics don’t apply. It doesn’t mean a point.

We don't know which and how were the laws of Physics at the beginning of the Universe. Surelly not the current ones.

I'm not talking about points anywhere. I talk about not zero size elementary elements of a discrete Space.

Edited January 11, 2021 by martillo

[martillo]

On 1/11/2021 at 10:43 AM, Halc said:

I’m just wondering why you’ve chosen a Turing machine.

 

On 1/11/2021 at 10:43 AM, Halc said:

No viable model of the universe expresses an initial state of a point. That’s kind of a pop-science naive view.

 

I was thinking about and seems to me now you could be very right on these two comments. 

It was my intuition that tends to think something beginning from nothing. As if a first nothing existing before everything and forever is what has sense. But the Parmenides premise of "Nothing can come from nothing" tell us that this is wrong. So something must always have existed and it doesn't need to be so simple like an elementary automata or an unstable point in a "fields foam". This is a turn again on the same error of thinking in something coming from nothing. We can think then in something always existing and with some inherent complexity, complex enough to generate the Universe we are living. So we can well think in some eternal and powerfull enough computing machine executing the curent Laws of Physics over this Universe. Of course I know there are different theories about which Universe is this but this is another story. I think they all could run in that powerful machine.

This way I would fall into the  "Simulation Theories" as they are being currently called, I know. It's only that with the Parmenides premise they are having more sense to me now except that I don't think precisely in a simulation, I think this Universe and life is for real...

Edited January 14, 2021 by martillo

[pzkpfw]

There's an XKCD for this.

https://xkcd.com/505/

[martillo]

32 minutes ago, pzkpfw said:

There's an XKCD for this.

https://xkcd.com/505/

Very, very appropriated... 😄

Edited January 15, 2021 by martillo

[FragmentedCurve]

I get the feeling you'd be interested in the recent work of Stephen Wolfram and his crew if you haven't already been looking at it. He builds up geometry from just nodes and edges (graph theory) with the goal of deriving what we know about physics already.

https://www.wolframphysics.org/technical-introduction/basic-form-of-models/

From what I understand, when he published his big book NKS approximately 20 years ago, he knew cellular automata (and graphs) was too limiting for his purposes because it imposed structure. In recent years, a colleague of his Jonathan Gorard put together the piece Wolfram was originally missing which was building his models with hypergraphs. After that collaboration they feel they've made a great deal of progress.

I haven't been following their work closely, so I might be off with the details. Anyway, you might be into it.

[martillo]

2 hours ago, FragmentedCurve said:

I get the feeling you'd be interested in the recent work of Stephen Wolfram and his crew if you haven't already been looking at it. He builds up geometry from just nodes and edges (graph theory) with the goal of deriving what we know about physics already.

https://www.wolframphysics.org/technical-introduction/basic-form-of-models/

From what I understand, when he published his big book NKS approximately 20 years ago, he knew cellular automata (and graphs) was too limiting for his purposes because it imposed structure. In recent years, a colleague of his Jonathan Gorard put together the piece Wolfram was originally missing which was building his models with hypergraphs. After that collaboration they feel they've made a great deal of progress.

I haven't been following their work closely, so I might be off with the details. Anyway, you might be into it.

Thanks for your suggestion. I'll take a look. You know, I'm interested to know if a Turing machine could surge from an underlying basic "mathematical structure". I was thinking in something like a "fields foam" which could be "continuos" in its nature but this is not essentially necessary and may be graphs or hypergraphs of nodes an edges could be a great one as a source of a Turing machine. A Turing machine has the same computational capability of any computer and so it can run any program a computer does and has the advantage of being a very simple machine. This way, anything from video games to artificial intelligence could run in it and this turn it possible to produce a Universe with intelligent life in it. Its inefficiency of having to run in a much higher speed to achieve the same results is not a problem for me. I can set any unveliebable huge number for its speed as it would be needed. I'm just interested in a possible explanation of the origins of the Universe, not to build a practical one. I think you could feel that this is actually possible like I feel.

Thanks a lot.

Edited January 15, 2021 by martillo