Hacker News Reader

Comment by garganzol | Hacker News Reader

What about computation completeness as in a Turing-complete vs. non-Turing-complete system as a marker of life? Morphology then can be characterized by the presence of features required for implementing computational operations.

For instance, Conway's Game of Life is a Turing-complete system.

View on HN

"Turing complete" is any theoretical construct that can do infinite recursion.

Such a thing doesn't really exist in the real world, of course, so discussing it outside of the realm of pure math is not useful.

> Such a thing doesn't really exist in the real world, of course

You are Turing-complete, including your consciousness and cognition. But your comment tries to prove otherwise. This situation is called logical contradiction, suggesting that your initial assumption may very well be false.

Turing-completeness requires the ability to recall an infinite number of bits, which is obviously impossible in physical reality. A Turing machine with that property removed is a pushdown automaton.

> Turing-completeness requires the ability to recall an infinite number of bits

It does not. The number of available bits just determines the maximum complexity of the system that can be built.

I mean, the tape being infinite is part of the definition of what a Turing machine is.

I think you are falling into the trap of thinking everything looks like a nail when you have the most familiarity with hammers

Right, this just seems like adding an alternate definition of life without moving the morphology problem.

The key point is that morphology features can be mapped to equivalent math operations. So we can build from there to formally determine whether a given object can be alive or not.

Can we unequivocally distinguish computation from non-computation in the real world? What's the difference between the physical processes inside a computer and inside a star that would let us say definitively that yes, one is computation and the other is not?

For a system to be Turing-complete, it should implement at least three operations: addition, negation, conditional jump. If there are similar processes inside a star, then yes, that star can be considered Turing-complete. "Computation" is just how we call it in math; in a star, it can be physical gas expansion, thermal convection, etc. But the outcome and its fundamental properties are essentially the same despite different forms of a particular presentation.

So are you abandoning the initial thesis that computation is a sign of life, or are you implying the sun is a computer built by some unknown form of life?

The implication is that a star may be an unknown form of life, in theory.

This implies that everything is alive since, like I said, we have no way to unequivocally distinguish computation from non-computation. Are the molecular vibrations in a rock actually computing something? How could we possibly tell? What's more, that rock might itself be a processing element of a larger computer and its own processing elements might themselves be computers, so in fact life exists fractally at all levels of magnification.

> Are the molecular vibrations in a rock actually computing something?

Not directly, but without them chemical reactions would not be possible. "The molecular vibrations", Brownian movement, or, more formally, temperature determines how fast the reactions could occur. It's a physical equivalent of a clock signal in a computer.

Without temperature (t=0), "molecular vibrations" do not occur, so all chemical reactions stop, making active computations based on those reactions impossible. This is a physical equivalent of a halt state in a digital computer.

> What's more, that rock might itself be a processing element of a larger computer and its own processing elements might themselves be computers, so in fact life exists fractally at all levels of magnification

This is logically wrong. Water is a chemical substance. It makes life possible, but it's not alive by itself because it does not meet the formal criteria. Just like a single transistor is not a CPU.

>Not directly

How can you possibly tell? Yes, Brownian motion looks random and meaningless to us, but how could you tell it does not look meaningful to someone?

>It makes life possible, but it's not alive by itself because it does not meet the formal criteria.

If you accept that the sun might be alive you must accept that water might be alive, since you can't definitively say that computation is not happening in either of them.

>Just like a single transistor is not a CPU.

I can't say for certain that a transistor is not a CPU to someone. A single atom could hypothetically be powerful enough to compute something.

> Brownian motion looks random and meaningless to us, but how could you tell it does not look meaningful to someone?

Random generator is not a CPU either.

>A single atom could hypothetically be powerful enough to compute something.

I agree. But if we cannot extract the useful effect, that potential power is diminished to zero.

Interesting idea, but what reasons do we have to think this would this be true?

We can look at a cell as a Turing-complete bio-chemical machine. Given the right chemical substances, it produces the desired outcomes. For example, an intra-cell protein building machinery is very much an interpreter executing code (from RNA). All other stuff like replication, cognition, etc is built upon these basic cellular blocks.

Right, but how can we tell this is not an accidental property?

If we inhibit that property then a cell dies. This is not a proof, of course, but let's call it a lemma.

Reply on news.ycombinator.com