Usually, forms of life are associated with organisms like humans, plants and animals. The probability that such can evolve in an environment is very low due to the many constraints that have to be fulfilled, e.g.

  • Planets have to be in a habitable zone.- planets require an atmosphere to shield cosmic radiation etc.
  • Certain elements are required to create chemical bonds which then form cells.

Cells can grow into larger structures like organisms. According to Darwin's evolutionary theory those cell structures will evolve to become better adapted to their environment, possibly becoming more complex and "advanced."

But can we also expect extraterrestrial forms of life like a machine - similar to a computer or robot?

Under the right circumstances it could be possible that elements bond themselves to simple electronic components like AND and OR gates, very much like cells. If evolution takes place those gates could form more complex electronic circuits and represent some type of logic. This would be some kind of digital evolution.

Future generations could end up as "robotic" kind of intelligence that can transform and optimize itself.

Are there any constraints that prevent the existence of a "robotic" forms of life with intelligence?

  • 39
    $\begingroup$ What is life? What is machine? What is intelligence and how you distinguish robotic one from our kind? $\endgroup$
    – Mołot
    Commented Aug 18, 2016 at 22:47
  • 3
    $\begingroup$ @Molot has placed his finger precisely on the problem. In my opinion the concepts of "machine" and "life" are mutually exclusive by definition. $\endgroup$
    – Lumberjack
    Commented Aug 19, 2016 at 0:06
  • 1
    $\begingroup$ Carl Sagan seems on topic. $\endgroup$ Commented Aug 19, 2016 at 9:45
  • 5
    $\begingroup$ @Lumberjack Most biologists — especially those that adhere strictly to modern evolutionary synthesis, aka. neo-Darwinism — tend to disagree: informally, and in at least a few instances also more formally, cells (as well as compartments and molecules inside the cell; and aggregations of cells, organisms) are seen as a particular class of machines. There’s really no contradiction unless you define “machine” as “not alive”. In fact, I’d venture that the people who answer “no” below aren’t biologists because their arguments aren’t supported by modern biology. $\endgroup$ Commented Aug 19, 2016 at 15:39
  • 2
    $\begingroup$ Actually materials aren't that great difference. Carbon fiber is increasingly popular, for example. Also, simplest definition of machine is "sth that turns energy into intended action". Is wedge a simple machine? Sure it is. Are there shells that can be wedges? Sure there are. And these are not constructed but grown. See? It's all about definitions. $\endgroup$
    – Mołot
    Commented Aug 20, 2016 at 0:22

23 Answers 23



The literal answer can be yes, as long as machines can satisfy the definition of life.

I would say that such mechanical life would have evolved through a biological phase first, where some sort of biological sentience designed the self-replicating machines that then took over.

I agree with other answers that direct "digital evolution" sounds implausible.

  • 3
    $\begingroup$ G'day, cobber. Welcome to Worldbuilding SE. Just a suggestion to improve your answer. Can you give a few more reasons how machine life would have developed and why "digital evolution" is implausible. Meanwhile have fun here. $\endgroup$
    – a4android
    Commented Aug 19, 2016 at 2:19
  • 3
    $\begingroup$ by technical definition humans are factories of factories of machines as well (mixing programming and industrial anaologies), when analysed from component level up, so I am with @a4android :) Any sufficiently complex mechanism starts to cross over into something that appears organic, whilst any sufficiently simple organism can appear almost mechanical. For example what's the difference between a single celled organism and a sufficiently simple but micronised machine (for example nanowerk.com/news/newsid=23703.php ) $\endgroup$
    – GMasucci
    Commented Aug 19, 2016 at 14:12
  • $\begingroup$ @GMasucci Agreed. And I think Sci-Fi writers agree too, as they often make alien technology appear organic, even though it's artificial just like our machines. $\endgroup$ Commented Aug 19, 2016 at 14:17
  • 1
    $\begingroup$ There's been more than one story about how sentient machinery took over the world after their masters died. $\endgroup$
    – Skye
    Commented Aug 20, 2016 at 7:20
  • 1
    $\begingroup$ The machines don't even have to take over -- they just have to be capable of interstellar travel (and colonization, if you want humans to discover machine-only worlds) when their biological creators are not. $\endgroup$ Commented Aug 21, 2016 at 2:05


To understand why not, first consider how biological life formed.

The building blocks of biotic life, particularly amino acids, have been show to form spontaneously when you add energy to a methane, hydrogen, amonia, water mixture. The exact process for getting RNA is less clear, but it is pretty clear that RNA is capable of replication and passing on traits. It is also capable of constructing both proteins and DNA, under the right circumstances. Blobs of lipids form cells automatically when suspended in water, just because they're polar, which could lead to RNA molecules being encased in cells.

Basically, biological life was able to form and evolve into what we have today because there exist simple precursor components which are 1) simple enough to spontaneously form from inorganic material, but 2) can replicate and pass on changes in their own design to their replicated copies

So why not electric/electromechanical life based on simple logic gates?


A lot of commenters have pointed out that there are many, many ways of representing simple logic gates, and they are absolutely correct. Simple boolean logic can be encoded in any number of ways, from simple chemical compounds as might be found in both biology and nature, to interlocking mechanical levers, to the BJTs and MOSFETs we use in modern circuitry.

This answer is about why we wouldn't expect to see evolved electronic life-forms which are similar to man-made electronics and robotics.

End Edit

Electronics fail on basically every facet of the requirements for life to form and evolve.

The building blocks our computer's logic gates are transistors. To function, a bipolar junction transistor (BJT) is made of three layers of P and N type materials, either NPN or PNP. P and N type materials are semiconductors, usually based on silicon which have been "doped" my mixing in a particular impurity to make the material either generally positive or generally negative. When sandwiched into layers NPN or PNP, controlling the voltage on the middle layer allows you to control the flow of current and the relative voltage between the outer two layers. Differences in the relative doping of the layers control which way current is allowed to flow at all, and the voltages involved.

So let's say we got really lucky, and happened to form a BJT at random from some lucky bits of silicon. Now what?

...well, nothing. In order for a transistor to do anything at all, it needs to be connected to a power supply, and have something else control the voltage on the base. Without that all you've got is a particularly odd rock. There's no mechanism available for a lone transistor to do anything at all, let alone replicate itself and produce another transistor.

Systems other than Transistors

(Edited to explain better/more concisely)

What about sustems other than transistors? After all, there's a lot of ways to implement a logic gate.

The problem here is that logic gates are devices that implement boolean logic, and boolean logic is a concept. There a lots of ways simple chemical processes in early replicators (of any kind; silicon based, carbon based, or otherwise) could be seen as implementing logic gates, but the system probably wouldn't replicate by virtue of being a logic gate, especially since boolean logic does not itself encode a concept of replication.

Lets say you find a simple compound or mixture of compounds which releases electrons at one point iff it is absorbing electrons at two other points. Sounds like an and gate.

However the system itself is not implementing a logic gate. You are modeling the system as a gate because you can see how the behavior is similar to the boolean and function. The system might or might not be able to be usefully incorporated into a replicator, but it won't replicate by virtue of being an and gate.

  • 2
    $\begingroup$ OK, so according to what you wrote, we're not likely to see self-assembling clocked logic gates doing well-defined computations. ... so what? It's not the only way to compute, but even more importantly, it's not at all clear what that matters to life in the abstract. $\endgroup$
    – hmijail
    Commented Aug 19, 2016 at 12:57
  • 1
    $\begingroup$ @KonradRudolph Molecular machinery and systems built thereof are machines as far as I am concerned, but I used the metalic vs carbon-based distinction for biological vs mechanical life because it fit the OP's question. Molecular interaction in cells are Turing complete but are not really what the OP seemed to be asking about. $\endgroup$
    – zstewart
    Commented Aug 19, 2016 at 18:41
  • 2
    $\begingroup$ The last paragraph in the section entitled So why not electric/electromechanical life based on simple logic gates? is precisely why creationists don't believe the section preceeding that title. There's just too much that all has to be functional at once (circuit and power supply in your example) to think that it could come about by chance. The term for that is irreducible complexity, and is often applied to biology in the same way that you applied it to electronics. $\endgroup$
    – AaronD
    Commented Aug 20, 2016 at 3:40
  • 2
    $\begingroup$ It seems incredibly conceited and narrow to me to assume that if electron-centric life evolved on another planet, that it must necessarily be constructed in the same way that the electronics components we invented are also constructed. $\endgroup$
    – Jason C
    Commented Aug 20, 2016 at 10:10
  • 1
    $\begingroup$ I’m with @AaronD: at the risk of sounding harsh, the answers on this question look like an exercise in creationism. The arguments sound superficially plausible but have nothing whatsoever to do with modern biology. There may be some property that fundamentally prevents the evolution of electronic circuits but none of the answers so far has given a compelling reason. Worse, nature disproves them trivially. Logic gates exist at various levels of the organism (e.g. gene regulatory networks, neuronal networks), as do electronic circuits (see action potential). $\endgroup$ Commented Aug 20, 2016 at 12:40

Once I read a kind of mindblowing short story partially concerning this question. It gave me a big idea about what after all could happen. Of course from what we know it is unlikely. But hey, the universe is big and who knows what is going on at all the spots we won't ever be able to see due to the universe's expansion.

So my answer here is:


It is verry unlikely this could happen but it could. How unlikely it is depends on your definition of life and machine, which you didn't disclose.

i.e. I could imagine that it somehow would be possible that some kind of RNA could be able to adapt to logical elements (as others have here welldescribed so far) resulting in biological entitys that morph through diferent stages of life of biological parts interacting with robotics.

Or if we consider anything non bioligical as robots, what about a galaxy where gravity designs some kind of source for pushing to specific optimized designs? Here comes also in the question about are you limiting the scale of size for what is considered living?

For anyone beeing interested to make an own view of what I tryed to explain here, this is the story I was refering to:


by Terry Bisson

"They're made out of meat."


"Meat. They're made out of meat."


"There's no doubt about it. We picked up several from different parts of the planet, took them aboard our recon vessels, and probed them all the way through. They're completely meat."

"That's impossible. What about the radio signals? The messages to the stars?"

"They use the radio waves to talk, but the signals don't come from them. The signals come from machines."

"So who made the machines? That's who we want to contact."

"They made the machines. That's what I'm trying to tell you. Meat made the machines."

"That's ridiculous. How can meat make a machine? You're asking me to believe in sentient meat."

"I'm not asking you, I'm telling you. These creatures are the only sentient race in that sector and they're made out of meat."

"Maybe they're like the orfolei. You know, a carbon-based intelligence that goes through a meat stage."

"Nope. They're born meat and they die meat. We studied them for several of their life spans, which didn't take long. Do you have any idea what's the life span of meat?"

"Spare me. Okay, maybe they're only part meat. You know, like the weddilei. A meat head with an electron plasma brain inside."

"Nope. We thought of that, since they do have meat heads, like the weddilei. But I told you, we probed them. They're meat all the way through."

"No brain?"

"Oh, there's a brain all right. It's just that the brain is made out of meat! That's what I've been trying to tell you."

"So ... what does the thinking?"

"You're not understanding, are you? You're refusing to deal with what I'm telling you. The brain does the thinking. The meat."

"Thinking meat! You're asking me to believe in thinking meat!"

"Yes, thinking meat! Conscious meat! Loving meat. Dreaming meat. The meat is the whole deal! Are you beginning to get the picture or do I have to start all over?"

"Omigod. You're serious then. They're made out of meat."

"Thank you. Finally. Yes. They are indeed made out of meat. And they've been trying to get in touch with us for almost a hundred of their years."

"Omigod. So what does this meat have in mind?"

"First it wants to talk to us. Then I imagine it wants to explore the Universe, contact other sentiences, swap ideas and information. The usual."

"We're supposed to talk to meat."

"That's the idea. That's the message they're sending out by radio. 'Hello. Anyone out there. Anybody home.' That sort of thing."

"They actually do talk, then. They use words, ideas, concepts?" "Oh, yes. Except they do it with meat."

"I thought you just told me they used radio."

"They do, but what do you think is on the radio? Meat sounds. You know how when you slap or flap meat, it makes a noise? They talk by flapping their meat at each other. They can even sing by squirting air through their meat."

"Omigod. Singing meat. This is altogether too much. So what do you advise?"

"Officially or unofficially?"


"Officially, we are required to contact, welcome and log in any and all sentient races or multibeings in this quadrant of the Universe, without prejudice, fear or favor. Unofficially, I advise that we erase the records and forget the whole thing."

"I was hoping you would say that."

"It seems harsh, but there is a limit. Do we really want to make contact with meat?"

"I agree one hundred percent. What's there to say? 'Hello, meat. How's it going?' But will this work? How many planets are we dealing with here?"

"Just one. They can travel to other planets in special meat containers, but they can't live on them. And being meat, they can only travel through C space. Which limits them to the speed of light and makes the possibility of their ever making contact pretty slim. Infinitesimal, in fact."

"So we just pretend there's no one home in the Universe."

"That's it."

"Cruel. But you said it yourself, who wants to meet meat? And the ones who have been aboard our vessels, the ones you probed? You're sure they won't remember?"

"They'll be considered crackpots if they do. We went into their heads and smoothed out their meat so that we're just a dream to them."

"A dream to meat! How strangely appropriate, that we should be meat's dream."

"And we marked the entire sector unoccupied."

"Good. Agreed, officially and unofficially. Case closed. Any others? Anyone interesting on that side of the galaxy?"

"Yes, a rather shy but sweet hydrogen core cluster intelligence in a class nine star in G445 zone. Was in contact two galactic rotations ago, wants to be friendly again."

"They always come around."

"And why not? Imagine how unbearably, how unutterably cold the Universe would be if one were all alone ..."

  • 3
    $\begingroup$ Pretty good short novel, and by the way it answers my first question which was then closed, on world building. That's very interesting. I know it doesn't answer that one with a lot of scientific argument but at least it's very good base to write other stories. $\endgroup$
    – Kaël
    Commented Aug 19, 2016 at 11:29

Arthur C Clarke had a progression of life in the novel 2001. Biological life moves into machine form (starting with cyborgs and ultimately moving into full uploads. Like becomes the spaceships needed to explore the universe), before finally becoming beings of energy who "live" in the fabric of spacetime.

The people who said "no" are correct with current and foreseen technology, but at a vert fundamental level biological life is very mechanical in nature, relying on chemical synthesis, electrochemical energy and even quantum effects (photosynthesis). Ions move through "gates" in the membranes of cells, flagellum are powered by "molecular motors" and so on. Futurists like K Eric Drexler used this as inspiration for the original conception of nanotechnology, and there is no conceptual reasons that such principles cannot be created by us in the future, including self assembly and reproduction.

In terms of would we meet machines as aliens, it seems likely. Once again Arthur C Clarke lays out the reasons. The universe is 13 billion years old, and the Solar System is only 4.5 billion years old. Life arose perhaps 3.5 billion years ago, and complex life @ 500 million years ago. So life has had an immense amount of time to have arisen in the Universe, and to evolve into complex forms that we might not even recognize. Clarke stated the argument as "Apes or Angels". Given the very tiny slice of time that we have been around, it is highly unlikely that we would meet any civilization near our own level of development. We could run into a planet where life is developing or perhaps at a photo "human" stage, or we would discover beings who evolved billions of years ahead of us.

And of course, over billions of years, they could have easily moved from cyborgs to full uploads, and artificial machine bodies might actually be the most common form of life in the Universe today.



Biological life has to replicate itself. Even tiny and relatively simple microorganisms can replicate in a wide variety of common environments. Robots are designed to do a job, not to make more robots. Before you can have a life form with a silicon chip brain you need to make the chips. Look at how we do that. Super high purity silicon is made using a complex chemical process. Under ultra clean conditions silicon is melted (melting point just above copper) gradually cooled to form a crystal, sliced and polished. Then its coated in all sorts of dangerous chemicals and blasted with UV light (not just scattered UV, the pattern is projected onto it). I am not an expert on this but you get the picture. Many processes, chemicals and temperatures that anything remotely resembling biology can't stand. It's also a very specific chemical process, one tiny mistake and the whole things useless. Even if a chip was produced, that wouldn't be what you want. A whole robot needs to be produced by natural forces. Complete with the complex tools it needs to reproduce and the programing to do so. Brains however can be made under common conditions with abundant materials and are forgiving to small mistakes. But they don't pop into existence on their own either. The minimum self reproducing system is what you want to look at. We have not yet made a self reproducing robot. The larger and more complex something needs to be to reproduce (and more precise) the less likely.

  • 4
    $\begingroup$ "Many processes, chemicals and temperatures that anything remotely resembling biology can't stand" - bah, that's exactly how scientists described deep ocean vents a few decades ago. $\endgroup$
    – Kevin
    Commented Aug 19, 2016 at 5:21
  • 10
    $\begingroup$ "We have not yet made a self reproducing robot." So just because we ape-descended life forms — who are so amazingly primitive that we still think digital watches(x) are a pretty neat idea — have not yet invented self-replicating machines, then no-one else in the universe will?! I am sorry but you just earned yourself a big fat -1 for that unimaginative premise. (x) xkcd.com/1420 $\endgroup$
    – MichaelK
    Commented Aug 19, 2016 at 8:14
  • $\begingroup$ Good answer. I just want to point that some robots are already built for self-replication. This makes your point no less valid: Machines are designed to serve the purpose of a conscious third party and optimised within the parameter set by this third party. Living organisms not. $\endgroup$
    – runlevel0
    Commented Aug 19, 2016 at 9:25
  • $\begingroup$ @runlevel0 if a conscious third party would create an organism optimised within the parameters set by them, would that disqualify them from being called "living" ? E.g. if we'd create something from scratch, currently there are teams working on synthetic bacteria but there's no reason why in a hundred or thousand years we couldn't create mammal-comparable animals, wouldn't they be living? $\endgroup$
    – Peteris
    Commented Aug 19, 2016 at 10:42
  • $\begingroup$ @MichaelKarnerfors: Your point is that some one might someday invent self-replicating robots. But the OP and Donald seem to be talking about undirected evolution, rather than invention. Donald's "We have not yet made" is an a fortiori argument: If we haven't been able to invent such things yet, then all the more does it seem unreasonable for such a thing to develop without an inventor. This argument might be mistaken, but the point that someone somewhere might invent such a thing doesn't invalidate the argument. $\endgroup$
    – LarsH
    Commented Aug 19, 2016 at 14:34


On video, no less: https://www.youtube.com/watch?v=2_9ohFWR0Vs

Penrose, Von Neumann and others already thought long and hard about this, even before DNA was discovered - and managed to devise ways in which simple blocks could "self-replicate".

In fanciful terms, we visualized the process of mechanical self-replication proceeding somewhat as follows: Suppose we have a sack or some other container full of units jostling one another as the sack is shaken and distorted in all manner of ways. In spite of this, the units remain detached from one another. Then we put into the sack a prearranged connected structure made from units exactly similar to those already within the sack... Now we agitate the sack again in the same random and vigorous manner, with the seed structure jostling about among the neutral units. This time we find that replicas of the seed structure have been assembled from the formerly neutral or ``lifeless'' material.


  • $\begingroup$ It's not clear that we can make the leap from "in fanciful terms, we visualized" to "we can expect" it to happen. Why should we expect the units in the bag to assemble into replicas of the seed structure? And how far does the process have to advance before it's capable of shaking the bag, inserting "prearranged" connected seed structures, and any other external inputs required? $\endgroup$
    – LarsH
    Commented Aug 19, 2016 at 14:41
  • $\begingroup$ @LarsH, "we can expect" because they DID it and it happened. See the video, or take a look at the link. It's actually an absurdly easy example. Regarding the "prearranged": the beauty of the process is that it's simple enough that once two units connect at random (which is easy enough), it will cause other units to connect in the same way. $\endgroup$
    – hmijail
    Commented Aug 19, 2016 at 17:45
  • $\begingroup$ Sorry, I was unclear as to what I meant by "it." They did something, and it happened, but not life. Special units were carefully engineered to connect, disconnect, and spread this configuration to others through physical contact. The construction is clever, no doubt about it. But that's just it ... no interesting behavior arose without complex, cleverly designed units introduced from outside. Merely spreading tilt angle from one unit to another is hardly "self-replication," any more than a toppling line of dominoes. Calling the connected pieces "life" doesn't make it so either. $\endgroup$
    – LarsH
    Commented Aug 19, 2016 at 21:32
  • $\begingroup$ So, some guy put some simple chemicals in a bottle, applied some sparks, and some hydrocarbons formed. So what? That's not life! Right? You dismiss the wooden pieces as "complex, cleverly designed units". But carbon atoms, and oxygen, and hydrogen, and etc etc, are much more interactive than any wooden piece you can engineer. Pump up the complexity of the wooden piece (add magnets? electrical charges?), add eons of time (how many atomic collisions happen per second in any aqueous solution?), and... who knows what would happen. $\endgroup$
    – hmijail
    Commented Aug 20, 2016 at 10:03
  • $\begingroup$ Indeed, who knows. But notice your scenario starts with pumping up the complexity of the units ... something that is external to the "shaking the bag" process, never accomplished by that process. Penrose et al. got more interesting behavior by introducing more complex, cleverly engineered units from outside, not by continuing to agitate simpler units for long periods of time. $\endgroup$
    – LarsH
    Commented Aug 20, 2016 at 15:00

Yes but with a big ol' asterisk

You are correct in saying that a computer and a brain are both complex, in fact some could argue a brain is much more complex than a computer. So following simple probability rules we can assume that computer aliens are way more common, right?

Well, yes, if you squint really, really hard.

It really depends on whether you mean natural computers or man made ones. The brain is really nothing more than the world’s most complex computer and following that logic, then yes, all intelligent alien will be computers. But if you mean computers like the ones man makes the answer is not likely, in fact if it wasn't for the infinite space and life possibilities, the answer would be a hard no.

  • $\begingroup$ Computer like we humans make are very unlikely. Our computer designs spend a lot of effort pretending to do things in order, just so our programmers can pretend to understand what is going on. Really efficient designs would be unclocked and purely event driven, and programmers would have no idea what happens when or in what order. $\endgroup$
    – Zan Lynx
    Commented Aug 19, 2016 at 15:38

A big YES.

Actually, I think it's more probable that we will find alien intelligent machines than alien intelligent organic species. Look at us, humans: we're extremely fragile and unsuited for extraterrestial travel. In order to move to another planet, or even our own Moon, we need to encapsulate ourselves in very complex and cumbersome spaceships which would carry with us a suitable atmosphere, prepared food, facilities which we need to function properly, like beds (or at least chairs) to sleep on, toilets, showers, etc. The spaceship needs to shield us from cosmic rays and for longer journeys it has to provide some sort of artificial gravity and a recycling system. In the same time, in order to send a probe to another planet we just have to build it and throw it in a given direction. (Of course I oversimplify but the difference in the amount of work needed to be done is immense).

With more probes and other types of machines in the outer space, the human presence on not terraformed planets will not be necessary. Machines can dig out minerals for us from asteroids and send them back to Earth. Machines can build a system of solar panels around the Sun, providing us with energy. They can explore the Solar system and send us the results. Heck, they can even wage wars against machines from another country on Earth. The only problem is distance which makes difficult for human beings on Earth to control the machines due to the light speed limit and to fix machines broken in some accidents or assaults far from mother stations. So, we will probably want the machines to be more intelligent and able to make decisions on their own, at least in a limited way, and we will want them to be able to fix themselves and other of their kind, as well as to build new machines from materials dug out from asteroids. Scroll a thousand years into the future and you will have swarms of intelligent robots running around the Solar system, serving (or not) their organic masters who live somewhere there on a small planet (or maybe they're already extinct), but otherwise autonomous.

And why would it be different with any other civilization in the universe?

So, it's a no if we're talking about robots built from scratch by evolutionary processes, but YES if it means that there was a civilization of organic predecessors who built the first intelligent, replicating robots, and they took it from there.


This question stems from very narrow-minded idea of what life actually is. Let me point out what is it that we usually see as machines: systems created to perform tasks. Life, on the other side, is system (possibly randomly) created that infinitely reproduces. There is no purpose to it.

Your idea of robotic life includes AND and OR cells. This is already flawed because not all machines use AND and OR cells. And I wouldn't be so sure that life on Earth is devoid of such contraptions.

The answer is therefore yes, but it's more complicated. Idea of machine life with WiFi and USB ports is childlish. But idea of life that exists due to incredibly complex mix of metal crystals and acid/base liquids that drive the being by creating electric current through redox reactions is much more plausible.

You could expect such beings to live underground and feed on minerals used to extract acid+metal mix required for the red-ox reactions. You could expect them to stop moving once acid is neutralized. They might die, just like us, if electric charge is used to keep some internal reactions going.

Simply put, spontaneous robotic life wouldn't be much different from ours.


TL;DR: yes

The definition of a lifeform is that it must replicate itself. Usually, biologists also insist in that a living being must have a metabolism.

A machine is a thing that performs any work.

So, the two things are not mutually exclusive, and hence a living machine is theoretically possible.

The more complicated question to answer is the Turing dispute. Who says that animals and plants are not machines? In a wide sense, all these lifeforms also satisfy the definition of a machine. This would make the answer to your question trivially yes.


Probably yes

If you are martian, the the only extraterrestrial "life form" is the Rover sent by Nasa on Mars.

While I find not plausible natural evolution of machines, if a biological race born, then create evolving machines and later die, we will see only such machines so we can met the machines as an extraterrestrial life form.

From e evolution point of view a machine will be immune to all common diseases so it is a great advancemente (however it will be weak to computer viruses?)

However I find that current circuits are not suitable for artificial intelligence, while software can do amazing things (see Machine Learning) it miss something, I would bet any machine lifeform would have a analogic-neuronal signal instead of a digital one. Of course we can simulate analogic signals but at great expense of computing power, wouldn't it be better if using directly analogic signals?


You seem to have missed a key factor in evolution.

It all starts when random physical processes happen to produce a basic form of life. At this point, it's true that those random processes could produce something that behaves like a logic gate just as easily as they could produce something that behaves kinda like DNA or RNA.

The next step is that this basic form resists processes that would destroy it, and somehow produces copies of itself so that it becomes more common. DNA and RNA have chemical mechanisms by which they produce more of themself. As they do this, some of the copies come out slightly different, because of errors or imperfections in the copying process, or mutation caused by environmental factors. These differences allow for a general diversifying - variations on a theme. Those variations that are poorly suited to the environment will die, while better-suited to the environment will die less, and therefore you'll end up with more of the "fitter" ones about. There's no notion of "superior" or "more advanced" here - it's simply a matter of which ones are most able to avoid dying.

A logic gate does not have a way to produce more of itself, and therefore cannot "evolve". For this to work, your initial basic form needs to be able to self-replicate. A self-replicating system built from NAND gates (for example) would be very complex, which makes the chances of it forming spontaeously small enough that we would never expect to actually see it happen.

The distinction - and the thing that many people forget or don't realise - is that DNA is more than just a component of an animal like us humans. DNA is a complete self-replicating system in its own right, which happens to live inside of another self-replicating thing (the cell) which forms a part of an even larger self-replicating system (the organism). DNA can survive and multiply just fine by itself, given the right environment. When you have a DNA test, they take a sample and multiply it using a process known as PCR so that they have more to test. Basically, they're taking a part of you, and getting it to make more of itself. On the other hand, you can't take a chip out of your computer and get it to build a few dozen copies of itself.


Other answers are missing a major concept. They're focussed on whether silicon-based life could evolve naturally. An alternative possibility is that silicon-based life is designed, but then its designers/creators disappear (for whatever reason).

There are various SF novels based around this. A good one is Saturn's Children which not only considers the society of robots that might be established after the end of the human race, but also the implications which the absence of humans might have on that society.

  • $\begingroup$ Scientist actually think that naturally evolved silicon based life is a real possibility. It's one of the few possibilities other than carbon. Of course, being silicon based hardly makes it a machine or computer. $\endgroup$
    – RubberDuck
    Commented Aug 21, 2016 at 16:47
  • $\begingroup$ Yep, it's considered possible. (Or at least there's no hard evidence of it being impossible right now. :) But if we consider a society of designed robots, then that throws evolution out of the window and opens up a whole bunch of interesting options. $\endgroup$
    – Graham
    Commented Aug 22, 2016 at 9:08

Not with how you describe it.

The other answers have explained why we will encounter constructed machine life, and a few have covered some conceptual reasons mechanical life couldn't evolve. They haven't covered any practical, technical reasons, though.

Let's look at this very, VERY literally first

First, consider Humans. We exist very happily in the mid-to-lower range of temperatures in which liquid water forms under 1 atmosphere of pressure. The water in our bodies provides a nice medium for all our fancy chemistry to happen. If you freeze us solid, all of the chemicals and compounds we rely upon to live get locked in place, ceasing all that fancy chemistry. If you heat us up beyond boiling, again our liquid medium becomes useless - not to mention all the other nasty problems of being too hot. The same goes for most life on earth.

So now think about a machine. Something simple, like an electric motor. You need magnets, conductive metals, and some kind of casing. In order to form those conductive metals, you need to heat them to their melting point - which is going to vastly exceed the curie point of our magnets, stripping them of their magnetism and rendering them useless. This is assuming we have a medium to move all of this around in.

Now, a little less literally

Instead of thinking of machines as we know them, let's define machine life as 'Life based on metal.'

This makes things easier... But not much. In order for iron-based life to keep its medium-fluid flowing, it would of course have to live at the melting point of iron. There's not many chemical reactions that occur at these temperatures, and if it dips below the melting point of iron we have all the same problems of a frozen human. You also now have the problem of coming up with a planet hot enough to melt iron, and having this species develop technology when they melt everything they touch.

Even if you choose lower melting points, most or all of the problems remain - There's simply not enough complex chemical reactions avaliable to sustain a 'lifeform.'



If there is even a tiny possibility of a self-replicating robot getting formed by chance, then given a large enough universe and enough time, you can basically guarantee that it will happen.

(I'm going to avoid the "alive" debate...and just treat a self-replicating robot as alive).


  1. By arranging metals, sand, etc., self-replicating robots can be formed.

  2. Nature arranges metals, sand, etc., in a chaotic random fashion.

  3. After enough random trials, Nature will eventually arrange them into a self-replicating robot.

Granted, it would probably take a lot of time to put together a jigsaw puzzle by shaking it...but if the universe is really big and has a lot of time...what's stopping it?

Is it easier to shake elements together and make a single-celled organism or a robot?

Just my two cents, but it seems like both would be pretty hard.

P.S. Robot don't need a perfect habitable temperature, and atmosphere, so nature might have many more opportunities (a couple orders of magnitude) which could help offset the improbability.


No - sort of:

As other answers stated quite detailed, a purely mechanical machine species is very unlikely to evolve, because it lack an "evolutionary" mechanism of procreation. We can likely assume the robots would reproduce asexually, therefore to individuals are incapable of sharing positive traits or offseting negative ones with positive ones. (Positive and negative with respect to their current environment and 'life style').

On the other hand hot-swaping parts of their programming might be a very beneficial trait in situations, where the robot is sufficiently advanced to do this at will. (Less at will, but similar to bacteria exchanging plasmids).

An other issue would be the power source. Robots as we understand them use electricity, would a robot driving its energy from say photosynthesis still be considered a robot in terms of this question? Chemically driven robots sound suspiciously like animals, but have a much easier way of sharing/distributing energy throughout the ecosystem.

On the other hand: consider natural reactors, most people would agree that a nuclear reactor is not something that can just happen but .... https://en.wikipedia.org/wiki/Natural_nuclear_fission_reactor

An other point in favour of natural computers: Alastair Renolds wrote a short story featuring a glacier acting similar to a brain. His point was that the glacier could compute thoughts and hold memories by different chanel conducting melt water. These channel were constantly rewired by some sort of worm. While this example is obviously not about life (no procreation, metabolism debatable) it still show cases an alien computer, which developed without outside influence, so maybe its what the OP was looking for?


Like many others have asked and assumed that machines produced by a non-machine civilization would become sentient and create a civilization of their own then certainly we would likely find many of them through the universe. If, however, you are referring to naturally occurring (i.e. naturally evolved) then I say it is extremely unlikely. We have already seen that the physical laws of this universe are such that organic life can exist. We also know that evolution of computer life occurs much faster than organic life. So, if computer life/intelligence were to have evolved naturally you would expect it to be the dominant life form in the universe and we'd likely have already been encountered it long ago (especially given that machines would be much more suited to space travel). In the universe, if some thing can happen you can assume it has happened (a lot) and the fact that we don't see computer life would point to the fact that there isn't any (and likely won't be any).


Is a machine "evolving" possible?
Sure. It's possible. Silicon based life could produce something that we might think of more as a "machine", but it would be far from, but if you want to stretch to the realm of the possible, silicon based life that looks more mechanical is where you're going to get such a thing.

Is the above probable?
No. Silicon is the 2nd most abundant element on Earth. The stuff is every where and it didn't happen here. Why? Carbon is just that much better than the only known alternative.

Technically... we're evolving into "machines" so, if you want to mean that it's 100% probable and possible.

  • $\begingroup$ Your question “Is the above probable?” is a red herring: Biological life is also extremely improbable. Yet it happened, and most (all?) modern biologists think that it has almost certainly happened more than once in the Universe. As for the argument that “carbon is just that much better than the only known alternative”, this is a typical parochial argument. Carbon is better suited here on Earth. It may be worse under other conditions. Granted, people have a hard time coming up with a suitable set of conditions but that’s an argumentum ad ignorantiam. $\endgroup$ Commented Aug 20, 2016 at 12:45
  • $\begingroup$ No. Silicon life would rely on the same processes, but less robust than carbon. If you have Carbon sufficient to create life it will always develop faster that silicon. We have a massive amount of silicon on earth and Carbon still won out and there is more carbon in the universe than Silicon, so the probablility that Silicon life happens is just not likely to happen. And life isn't considered "improbable" at all. All indications is that if it can happen it will happen, and pretty quickly too so we're dealing with a completely different thing with silicon. $\endgroup$
    – Durakken
    Commented Aug 20, 2016 at 12:52
  • $\begingroup$ You say “And life isn't considered "improbable" at all” — This is very wrong. Most contemporary researchers in the field argue that abiogenesis is extremely improbable, even under ideal conditions. The literature is unanimous on this. And there’s a lot of evidence for this; to name just two pieces: we’d expect to see a lot more life in the Universe otherwise, and we’d expect to see the formation of complex patterns in laboratories pretty much all the time. Regarding your silicon argument: this argument is known as carbon chauvinism. $\endgroup$ Commented Aug 20, 2016 at 13:17
  • $\begingroup$ "We'd expect to see a lot more life" <- that's just a dumb statement. "We'd expect to see formation of complex patterns in labs all the time" We do, we just don't call it life, because life is thousands of steps down the line with complex interactions that if you got, you're probably not setting your experiments up properly. As far as Carbon Chauvinism. No, that's called "the facts" Carbon chauvinism is an unwaranted preference or belief that life can't form or be as worthwhile as that of carbon. Obvious silicon life can form, just unlikely to do so "naturally" or thrive due to carbon. $\endgroup$
    – Durakken
    Commented Aug 20, 2016 at 14:01
  • $\begingroup$ You can call it dumb but it’s not made by me, it’s made by some of the most respected researchers in the field of abiogenesis, and it’s their assessment based on decades of research, rather than the opinion of some bloke on the internet who thinks it sounds dumb. — In addition there’s just no reason whatsoever to assume that abiogenesis should be frequent. There’s no data to support this assessment, contrary to what you claimed earlier. $\endgroup$ Commented Aug 20, 2016 at 14:03

First off, we are biological machines. All life forms are. The only reason you might bristle at this suggestion is that we've laden the term "machine" with a definitional proviso: that machines are made by people. Setting that quip aside, we are data and matter processors just the same.

But consider this. We are sending out our machines out into the universe much faster than we are sending ourselves. Shouldn't we expect other sentient species to do the same?

Although this is speculative, I'd argue it's likely we will eventually trade our meat brains (and bodies) for machines that can host our minds as information. If we do that, then it will be possible to transmit our minds (as information) across vast spaces at the speed of light to be hosted by other machines of the same kind. We probably should expect other sentient species to come up with the same idea.

In that vein, the first sentient life forms we come across may may appear in the guise of artificially crafted machines completely devoid of biological life. But those machines may be ferrying biologically evolved minds as information across the vast expanses of space between stars over millions or even billions of years.


Perhaps. I'm not sure if we already have an universal accepted definition for life. But I think there are some characteristics:

  1. Living things need to consume
  2. Living things excrete
  3. Living things reproduce
  4. Living things evolve
  5. Living things die (?)

For now the only reference we have is the life on earth, say life that is made out of carbon. But I think there is no reason to belief that only carbon-based life is possible or even that life is just possible under condition that are satisfying for us. Maybe there is a planet somewhere in the universe that is inhabited by completely different forms of life that exists under completely different conditions.

Then there is always the possibility of artificial life. We could program very simplistic robots that are designed to do specific task. To fulfill the task it may be necessary that the robots are adapting and learning; while the robots are going to accomplish the task, new problems might come up and the robots seek to solve that new problems. To do so they could ask other robots for help and finally change themselves to solve the problem. It is also possible that they will make replicates with appropriate changes in order to fulfill the new tasks. That would be some kind of evolution. But there are some differences. For example, the evolution would be less open ended and much more conducted and controlled.


Hard one.

I think we can´t answer that Question to your full satisfaction.

There are missing a lot of informational dependancys related to your question. Which have yet to be discovered by humanity.

  1. What is Life? (Is a Virus alive or dead?)
  2. What is intelligence? (Is there a measurable boundry?)
  3. What is consciousness? (Holy c... we know nothing....)
  4. Is it possible that meaningfull valid information is created by random physical processes? (What is the origin of genetic information?)

I can´t imagine that genetic information/biological programming like "reproduce, evolve, repeat" is created by chance... Was there perhaps a Sender/Author of that information?


Thank you for your input!

I gave it a second thought and come to the following conclusion:

Literally: Can we expect machines as extraterestrial form of life?

Yes, we can. If you are familiar with the Grey Goo Theory, Nanomachines capable of reproduction would fit into this quite well.

By Context: Can we expect machines as extraterestrial forms of life? (By natural procreation and evolution)

No. As the term machine says it is something that was "artificialy" created.

If there is no deeper secret to the past of Life as we know it, machines are made. Natural Life is not. It popped miraciously out of a pool of protein slime.

By the way... If a Life form based on a different element than carbon visits us with a space ship i can imagine i would be considered "alive".

Life can develop from different elements than carbon, hydrogen, oxigen etc... The chances are minimal thoug. A common star can fuse elements up to carbon within its lifespan. We know heavier elements are results supernovae which are not that common.

In our universe living things tend to use the most eenrgy efficient ways. Why dissolve things in sulfuric acid when you have liquid water at hand?(a pretty powerful solvent due to smaller molecules)

Why use sulfur as basic element when you have carbon (which binds better)? It is all a question of effective use of energy.

And then you build basic AND and OR Gates from them? Naaah... As we already see deterministic turing machines have limits that inhibit them to solve problems where lots and lots of variables are involved. (Traveling Salesman Problem, Knapsack Problem... just see Karps 21 NP-complete Problems)

If Life develops "circuits" they wouldn´t be deterministic logic gates... it just would not make sense regarding to effectiveness and mental fitness of the resulting Life form.

yeah.... so i assume NO

Thank you Gentleman!

  • $\begingroup$ Correct me in case I'm wrong but AFAIK we have answers to all your 4 questions. for 1, 2 and 4 its even pretty clear, we are just not fully aware of the reasons behind. And at point 3 there are different ideas trying to answer it but... While your idea is correct, this is a pretty thin argument for answering this OP. $\endgroup$
    – Zaibis
    Commented Aug 19, 2016 at 11:37

Apparently, your question does not imply or require abiogenesis, or, it does not matter whether the machines were designed or not.

Self replication of machines is very common, in fact, our factories are doing it all the time, albeit on a very large scale: what makes our cars, phones or computers? Machines makes them. What makes the machines? Other machines make them.

I am not going to detail, but what you can find within our industry is that, you have different machines that can make parts for each other, but not usually itself; and you also have machines that assembles other machines from such parts, but not usually itself.

Then if you have a pair of such machines, one Machine assembles the other, while both assembles the other machines that makes the parts of all the machines, including the part-making machines, then you get a improvable, self replicating system that operates on a very large scale.

Which is exactly what we see in cells: rna polymerase assembles the ribosome, the ribosomes assembles the rna polymerase, and the ribosomes makes all the other enzymes that makes the amino acid and nucleus acid parts of the different enzymes and trnas, which includes a DNA polymerase that copies the instructions for the machines (dna or rna). Exactly what we think a macroscale self replicating system would likely work like.


You may want to check Stanisław Lem's The Invincible novel. It presents an idea of mechanical 'life' forms, evolving out of self-replicating robots that were left alone after the civilization that created them had fallen.


  • 1
    $\begingroup$ I think this is actually the opposite of an answer to the question as asked, the OP has asked about the evolution of machines from raw chemistry not the creation of machines by organic beings. $\endgroup$
    – Ash
    Commented Aug 8, 2017 at 12:59

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .