Can a robot be called as Living thing?

[scienceisbest]

Can a self replicating, or self growing robot (programmed to make logical decisions) can be called as Living thing?

If not, what is the definition of living thing?

 

[Oldfart]

It depends upon one's definition of living thing. I think that most folks would insist that the creature be biological rather than mechanical in nature in order to be "living."

But in the far flung future, it may be possible for science to create biologically-based creatures to serve us. Then we may be asking if they could be considered as "robots."

OF

 

[scienceisbest]

In biological life, energy and informations are stored in matter made up of mainly C, N, H etc (Evolution has chosen those atoms) where as in robot, there are stored in inorganic materials. So both does the same job in different platform. It is possible to restart a robot which has stopped working due to the lack of energy by giving energy as input.
Is it possible, at least theoretically, to make a dead man alive using the same principle..

 

[DaveC426913]

The basis of the definition of life is:
- it eats and excretes
- respirates
- grows
- reproduces
- reacts to stimuli

There's all sorts of nuancing but it starts with those.

 

[Oldfart]

The basis of the definition of life is:
- it eats and excretes
- respirates
- grows
- reproduces
- reacts to stimuli

There's all sorts of nuancing but it starts with those.

It seems easy enough to imagine the building of a machine that can do all those things. I have a feeling that the term "living" is too vague, as it would permit endless quibbling. Perhaps we can even speculate that "living" is a fundamentally meaningless term, or at least should find a home under the philosophy place.

 

[DaveC426913]

It seems easy enough to imagine the building of a machine that can do all those things.

A machine that could reproduce itself would not be easy.


What makes you think life is so complex?

 

[Archosaur]

Here's a sort of thought experiment:
If a person has a prosthetic leg, is he still human? Of course!
What if he has two prosthetic legs?
Two prosthetic legs and an artificial heart?
What if every organ is replaced with a mechanical substitute, even the brain, the contents of which are "downloaded" into a network with transistors instead of neurons?

After which gradual step is he suddenly no longer human?

 

[Oldfart]

A machine that could reproduce itself would not be easy.


What makes you think life is so complex?

Maybe not easy, or even very practical, but it seems obvious that it could be done.

I didn't say it was complex. My point was that the definition of "living" is complex.

OF

 

[Danger]

I disagree that it must eat and excrete, only because that the human concepts of those functions are personalised. They might not be recognised in a normal manner, but are unmissable by those who grew up with the crap.

 

[Born2bwire]

I disagree that it must eat and excrete, only because that the human concepts of those functions are personalised. They might not be recognised in a normal manner, but are unmissable by those who grew up with the crap.

Maybe, but I think that some form of consumption is required in order to reproduce oneself. Otherwise, you continually strip yourself of resources. In that manner, I would think of consumption as being able to take in readily available resources and manipulate them to suit your needs. Whether it is our eating and digestion of foodstuffs to provide us with energy and matter or a robot taking in scrap metal or ore to smelt into materials to build the frame of a new robot.

In this manner, sure we can construct robots that are self-assembling, but they are unable to build the parts themselves.

 

[Danger]

I agree, Born. I just felt like playing Devil's Advocate in order to point out that neither intake nor exhaust will necessarily be recognizable by our species. Really, I've taken a crap or two that would never be recognized as having come from a human...

 

[DaveC426913]

Here's a sort of thought experiment:
If a person has a prosthetic leg, is he still human? Of course!
What if he has two prosthetic legs?
Two prosthetic legs and an artificial heart?
What if every organ is replaced with a mechanical substitute, even the brain, the contents of which are "downloaded" into a network with transistors instead of neurons?

After which gradual step is he suddenly no longer human?

We were discussing life versus non-life. Human versus non-human is a very different question.

The central paradox here is often known as the http://en.wikipedia.org/wiki/Ship_of_Theseus" [Broken].

If Theseus replaces every wooden plank on his ship one by one, is it still the same ship?

 

[Archosaur]

Human versus non-human is a very different question.

One of the unstated assumptions of my argument was that the human is also alive after each gradual step. You could redo the questions, this time asking "After which gradual step is he suddenly not alive?"

Of course, that would be a very complicated robot. A more pertinent question would be "What is the simplest robot that could be described as alive?"

But before I go there, do we all agree that a sufficiently complicated machine could be called "alive"?

 

[K^2]

I would only disagree with DaveC's criteria on eating/excreting and respiration. By that definition, anaerobic bacteria are not alive, since they require no respiration. Response to stimuli is also questionable, but only because of vagueness.

I would start with reproduction as most important criterion. It must self-reproduce however. I would not classify something like a virus as alive. Otherwise, we must include all existing technology in the definition.

I would then replace all statements about feeding/breathing or whatever into the single most broad definition. It must consume energy to maintain it's own Gibbs' free energy at above equilibrium level. This is all we really need to exclude "simple reproduction" of things like growing crystal.

Anything that follows these two criteria will evolve with changing environment. If it doesn't end up evolving to respond to stimuli or to do any of the other things we associate with living things, then it obviously doesn't need them.

Basically, as long as these two criteria are fulfilled, I'd call it alive regardless of origin.

 

[DaveC426913]

I would only disagree with DaveC's criteria on eating/excreting and respiration. By that definition, anaerobic bacteria are not alive, since they require no respiration.
...
I would then replace all statements about feeding/breathing or whatever into the single most broad definition. It must consume energy to maintain it's own Gibbs' free energy at above equilibrium level. This is all we really need to exclude "simple reproduction" of things like growing crystal.

Yes. Eating and excreting are terms we put a lot of meaning on. Ultimately, it must consume raw materials and emit waste materials. Eating/excreting and respiration are both subsumed under that.

Response to stimuli is also questionable, but only because of vagueness.

If it responds to no external stimuli, is it alive at all?

I would start with reproduction as most important criterion. It must self-reproduce however. I would not classify something like a virus as alive. Otherwise, we must include all existing technology in the definition.

Even that has its vagueness. There are organisms that can't reproduce without the help of other organisms (some parasites and symbiotes). Is that kind of like saying a machine needs the symbiotic help of another machine to produce its raw materials?

Anything that follows these two criteria will evolve with changing environment. If it doesn't end up evolving to respond to stimuli or to do any of the other things we associate with living things, then it obviously doesn't need them.

Basically, as long as these two criteria are fulfilled, I'd call it alive regardless of origin.

And it must grow.

 

[Danger]

I find myself in the peculiar position of both agreeing and disagreeing with someone.
I consider a virus to be alive. My best argument in favour of that is the fact that inoculations are specified to contain a killed virus.

 

[scienceisbest]

I think "having the free energy above equilibrium level" is the most basic condition. Rest of all the conditions, like growth, reproduction, consciousness etc are the properties gained in the process of evolution, depending upon at what stage of evolution it has reached. May be a growing crystal can evolve to a stage where it can do all the function what now we consider as the characteristics of living thing.

 

[K^2]

If it responds to no external stimuli, is it alive at all?

If I remove support from under a rock, it drops. And many living organisms can't respond to more than a pressure change. So it's way too vague.

Even that has its vagueness. There are organisms that can't reproduce without the help of other organisms (some parasites and symbiotes). Is that kind of like saying a machine needs the symbiotic help of another machine to produce its raw materials?

I agree. This needs work. What I'm trying to get to is difference between requiring a host for environment, in case of a parasite, and needing host to do the actual reconstruction, in case of a virus. Former should be classified as alive, while later should not. Any thoughts on how to formulate it better?

And it must grow.

Reproduction sort of forces growth at some stage. If it never grows past that point, I see no reason to disqualify it from being alive.

I think "having the free energy above equilibrium level" is the most basic condition. Rest of all the conditions, like growth, reproduction, consciousness etc are the properties gained in the process of evolution, depending upon at what stage of evolution it has reached. May be a growing crystal can evolve to a stage where it can do all the function what now we consider as the characteristics of living thing.

But if it doesn't reproduce, how can it evolve?

 

[Danger]

You make a good point, Scienceisbest. There has been a lot of speculation by reputable scientists as to whether or not a silicon-based life form is possible, since its valence shell is equal to that of carbon. I like the idea of the possibility, but have no evidence to support it.

 

[DaveC426913]

May be a growing crystal can evolve to a stage where it can do all the function what now we consider as the characteristics of living thing.

Except that a growing crystal is not alive. So we know that definition is too loose.

Let's also be wary of this word 'evolve'. You are using it simply in the transformative sense (white can evolve into black), but that is not the same as replication with adaptation.

 

[DaveC426913]

If I remove support from under a rock, it drops. And many living organisms can't respond to more than a pressure change. So it's way too vague.

Yes it is. But vagueness doesn't disqualify it; it simply measn we need to refine it.

Reproduction sort of forces growth at some stage. If it never grows past that point, I see no reason to disqualify it from being alive.

Well, if it simply made an identical copy of itself whole and unchanging, like a photocopier, would that count as life?

 

[Archosaur]

Well, if it simply made an identical copy of itself whole and unchanging, like a photocopier, would that count as life?

I can't see why not. If a bacterium assembled the next generation outside of itself using ambient materials, would we call it less alive than one that undergoes binary fission (which necessitates growth)?

Also (devils advocate here) if reproduction is a defining characteristic, then are sterile hybrids alive?

 

[SW VandeCarr]

The common freatures of living systems are metabolism and replication. Metabolism maintains a bounded volume in a lower state of entropy than it's environment, but through releasing heat, excretion and other activities, metabolism raises the overall entropy of the environment if we take the environment + life to be a closed system (the usual assumption).

The futurist literature (Kurzweill and others) has discussed the idea of self-replicating "nanobots"; minute robots that can eat matter and covert it to energy for their own metabolism and for replication. I would think that such "creatures" could be thought of as alive.

EDIT: Re: Sterile hybrids; an individual need be not capable of replication to be considered alive if it's a product of replication, but I suppose you could think of a number of odd exceptions. Also nanobots would be more lifelike if they could accumulate new information that would allow them to adapt.

 

[DaveC426913]

I can't see why not. If a bacterium assembled the next generation outside of itself using ambient materials, would we call it less alive than one that undergoes binary fission (which necessitates growth)?

Also (devils advocate here) if reproduction is a defining characteristic, then are sterile hybrids alive?

I think that may be too literal.

It is a technicality that sterile hybrids can't actually reproduce. They have all the necessary blueprints and machinery for replication of themselves, it's just that there's a "bug" in the system.

 

[Pythagorean]

Well, if it simply made an identical copy of itself whole and unchanging, like a photocopier, would that count as life?

Doesn't sound like it would be conducive to adaptation and diversity, which should be important to a classification of life.

 

[Newai]

I find myself in the peculiar position of both agreeing and disagreeing with someone.
I consider a virus to be alive. My best argument in favour of that is the fact that inoculations are specified to contain a killed virus.

Virii are ambiguous. They show symptoms of both life and death. Robots, not so much. All they know is on/off and move.

 

[Archosaur]

All they know is on/off and move.

This is a pretty shortsighted view. First off, there's a multi-billion dollar industry that says they can do a lot more, including recognize facial expressions, and even eat. Secondly, they're able to do more and more complex things every year. The solid line you're trying to draw between robots and living things simply isn't there, and that grey area is going to start being challenged in the future.

 

[Dr Lots-o'watts]

It's arguable that the concept of life can apply to abstractions, such as pixels and algorithms: http://en.wikipedia.org/wiki/Conway's_Game_of_Life

 

[Noesis]

We were discussing life versus non-life. Human versus non-human is a very different question.

The central paradox here is often known as the http://en.wikipedia.org/wiki/Ship_of_Theseus" [Broken].

If Theseus replaces every wooden plank on his ship one by one, is it still the same ship?

What an interesting paradox. I have thought about essentially the same question multiple times. From a particular viewpoint, the crux of answering the question lies in establishing some threshold where once crossed the object in question no longer retains its identity. I believe this can be given quite a measurable, i.e., mathematical, representation.

Already the framing of the question establishes some threshold since it is assumed that the ship, or the human vessel, remain static throughout time. Small changes are also likely to be dismissed: If a splinter is removed from the ship or a hair removed from my body, we are still likely to not question the persistence of the object's identity. I think interesting conclusions can be drawn from these thoughts, although some are, granted, fairly academic.

Of more practical concern: I think this provides a simple algorithm to determine what is responsible for something's function. Basically: remove things until it stops doing what it was doing. Removing planks one by one might provide considerable insight as to what keeps the ship afloat in the first place, in case Theseus was wondering. If technically feasible, I think this would be a fantastic way to attempt to identify the basis of consciousness. Ideally one could start by removing one molecule at a time and analyzing throughout, initially it'd be far more practical to start at a coarser grain by, say, removing entire neural compartments or neural assemblies.

I think the algorithm can also be abstracted away from a material setting, although will probably be more difficult to carry out in general. For example, in regards to an argument of what constitutes life, one could identify a living object and then start to subtract parameters until it was clear that the object could not possibly be alive without these remaining parameters. Then this fundamental set must be the basis of life. This might sound very oversimplified since, after all, it has been traditionally very difficult to establish a working definition of what life is. However, I think the main impediment has been what Dave mentions in another post: human versus non-human is a very different question than alive, or animate, versus inanimate.

 

[Noesis]

This is a pretty shortsighted view. First off, there's a multi-billion dollar industry that says they can do a lot more, including recognize facial expressions, and even eat. Secondly, they're able to do more and more complex things every year. The solid line you're trying to draw between robots and living things simply isn't there, and that grey area is going to start being challenged in the future.

I agree. I think one of the terms that will one day have to be redefined is artificial intelligence. Because what, after all, does natural intelligence really mean? Intelligence is a general concept and should be independent of its substrate.

 

[DaveC426913]

What an interesting paradox. I have thought about essentially the same question multiple times. From a particular viewpoint, the crux of answering the question lies in establishing some threshold where once crossed the object in question no longer retains its identity. I believe this can be given quite a measurable, i.e., mathematical, representation.

Already the framing of the question establishes some threshold since it is assumed that the ship, or the human vessel, remain static throughout time. Small changes are also likely to be dismissed: If a splinter is removed from the ship or a hair removed from my body, we are still likely to not question the persistence of the object's identity. I think interesting conclusions can be drawn from these thoughts, although some are, granted, fairly academic.

Your viewpoint ignores the idea that new additions can be imbued with identity once added.

i.e. if half the planks on Theseus' ship are replaced with new planks, it does not follow that only half the ship is Theseus' ship; it is more reasonable that the new planks are inaugerated into the "Theseus' Ship Club".

Of more practical concern: I think this provides a simple algorithm to determine what is responsible for something's function. Basically: remove things until it stops doing what it was doing. Removing planks one by one might provide considerable insight as to what keeps the ship afloat in the first place, in case Theseus was wondering. If technically feasible, I think this would be a fantastic way to attempt to identify the basis of consciousness. Ideally one could start by removing one molecule at a time and analyzing throughout, initially it'd be far more practical to start at a coarser grain by, say, removing entire neural compartments or neural assemblies.

I think the algorithm can also be abstracted away from a material setting, although will probably be more difficult to carry out in general. For example, in regards to an argument of what constitutes life, one could identify a living object and then start to subtract parameters until it was clear that the object could not possibly be alive without these remaining parameters. Then this fundamental set must be the basis of life. This might sound very oversimplified since, after all, it has been traditionally very difficult to establish a working definition of what life is. However, I think the main impediment has been what Dave mentions in another post: human versus non-human is a very different question than alive, or animate, versus inanimate.

I don't think this will work. I would not have to remove many pieces from a computer program for it to stop working. In fact, the removal of a single character - virtually any single character I might care to choose - is quite likely to be fatal. I would then erroneously conclude that that single character is the most important component in the program.

Likewise, I would not have to remove many components of a human for it to stop working too. I cannot thus conclude that the particular component I last removed is the difference between life and death.

Same would likely apply to consciousness.

Thus is the nature of complex, interdependent systems. I think these are excellent examples of 'the whole is greater than the sum of its parts'.

 

[darkfrog]

Certain definitions become useless at certain points. Do we call Archeopteryx a bird or a reptile? When a new class of something can be identified, it is reasonable to create a new definition or new system of taxonomy.
There may come a time when we have replicating nanobots that will need to be classified as somewhere between biological life and non-replicating inorganic material.

 

[Newai]

This is a pretty shortsighted view. First off, there's a multi-billion dollar industry that says they can do a lot more, including recognize facial expressions, and even eat. Secondly, they're able to do more and more complex things every year. The solid line you're trying to draw between robots and living things simply isn't there, and that grey area is going to start being challenged in the future.

Of course the industry says that. Would you expect them to put it so succinctly as I? Especially with all the different brands of A.I. they tout, none of which are even a true A.I. Anyway, all those nifty things they can do still boil down to their transistors: on/off. They don't reproduce, they don't evolve, they don't grow. They are no where near living organisms yet. My central nervous system, by comparison, knows on/off in a similar way, but it is not by itself. Without the other components that make me alive, that system would not function.

 

[bobze]

Most biologists would classify life based loosely on the 5 principles of "biology";

1. Ability to reproduce with fidelity
2. Cell is the smallest unit
3. Convert energy from one for to another
4. Regulate internal environment
5. Evolve

I don't also think I've met a biologist that doesn't understand the our definition of life is not a natural "law" of life. Ergo, by our definition such a robot would not be "living". Similarly, we don't consider viruses living (because they don't meet some of the above criteria for life).

That isn't to say that our definition of life is the definition of life, it is merely an agreement to allow discourse on applicable subjects. Certainly things which aren't alive by our definition could be alive under another definition of life, or posses life like properties.

 

[bobze]

argh, double post.