CPU, memory, and etc sre just a collection of these gates, connected by these gates.
For sure, you need to go further in this direction to build a general purpose computer, but it is in this direction that you have to go.
No, it can't be. You could unroll a constant length loop into a network of these domino calculators (which is how you get Turing-completeness in homomorphic encryption from a few logical primitives), but you cannot implement arbitrary looping that real computers can do.
>CPU, memory, and etc sre just a collection of these gates, connected by these gates.
No, they aren't. Those implement the additional functionality necessary to do that for arbitrary cycles of calculations, not just one static loop unrolled into a fixed-size circuit.
>For sure, you need to go further in this direction to build a general purpose computer, but it is in this direction that you have to go.
No, the functionality implemented here cannot generalize to arbitrary computation until you implement other, orthogonal primitives.
These dominoes are distinct in that they cannot flip the state back and therefore cannot rerun the same sections, and need need more and more dominoes to do more computations, but fundamentally I dont see how that is any different than providing a longer and longer input of energy into a modern computer.
Can you clarify what "orthogonal primitives" are missing?
At each step it creates an output, which can be repetitive, or not repetitive.
The generated sequence is still finite unless were talking about an infinite amount of power. If we allow an infinitely long domino then they are equivalent.
Sure. From my earlier comment[1] distinguishing the core primitives -- labeled a) through d) -- necessary to call something a computer:
>>if you want to get the core, impressive functionality of the things-we-call-computers, you have to include a bunch of other, nontrivial, orthogonal functionality, like a) the ability read and execute a lot of such instructions, and b) to read/write from some persistent state (memory), and c) have that state reliably interact with external systems. Logic gates (d) are just one piece of that!
>These dominoes are distinct in that they cannot flip the state back and therefore cannot rerun the same sections, and need need more and more dominoes to do more computations, but fundamentally I dont see how that is any different than providing a longer and longer input of energy into a modern computer.
In addition to SonOfLilit's comment, a fundamental difference is that, for a modern computer, the provision of electricity is specified as part of how the computer works, while the domino "computer" did not specify the apparatus necessary to reset and execute the subsequent instructions (nor the mechanism whereby such subsequent instructions would be provided) -- the a) and possibly b) in my model above.
So yes, if they had presented a "domino computer" that included such functionality (and rounded out the a-d), I would indeed agree with calling that a computer!
Why does (the absence of) such apparatus matter? Well, the whole point of the project was to provide intuition behind how/why some physical mechanisms give rise to a "computer". Since people already understand how dominoes work, they will therefore understand how the computer built from them works! Except they don't, because, for it to be a computer, you'd have to add on something they don't have an intuition for (the resetters etc).
So yes, by all means, use this as an intuition pump for yet-another-way to physically implement a calculator. Just, not a computer, which is something else and a tad more mysterious.
I don't see why it would need to specify this. Anyone who has played with dominoes knows you can set them up again. Why would anyone waste time stating something so obvious?
We already have a term for the machine where you have to reset it yourself after every calculation: a calculator. Hence the original comment.
>electricity
pick one.
This is generally where the pedant would derail the conversation into the usual Turing* Completeness* infinite tape semantics.