15
$\begingroup$

The inspiration for my question comes from The Great C by Philip K Dick. To summarize, the Great C is a computer that has taken over a large portion of the planet. It uses proxy agents to keep control by relentlessly patrolling the area and killing anything that gets in its way.

My question is an extension of this idea. Would it be possible for an advanced computer to take over an entire planet? The computer is buried deep underground. It has advanced intelligence. It was once part of a civilization, but that civilization has been destroyed by some type of apocalypse (possibly a nuclear war type scenario). It is not known how it gains the resources necessary to expand or to power itself. Would it even be possible to gain enough energy to power itself and continue expanding?

Would it be advantageous for the computer to spread its material on the ground? I'm thinking things like plastic or circuit boards, essentially altering the landscape to something more favorable to itself. What would be the most favorable type of material for a computer based entity such as this one?

Would it be advantageous for the computer to kill any life form that got in its way? Would it be possible for this spread of control to expand to the entire planet?

$\endgroup$
  • 2
    $\begingroup$ You should read the hard sci-fi novel "The Two Faces of Tomorrow" - it is quite old now but still very relevant and talks about just this sort of subject. $\endgroup$ – Tim B Sep 30 '14 at 19:13
  • $\begingroup$ Check out Ra by Sam Hughes. $\endgroup$ – dtldarek Oct 3 '14 at 10:25
14
$\begingroup$

There is a lot we don't know about this computer, but let us consider that it was created by a civilisation of some kind and placed on the planet with both the intelligence for agency and some facility for acting on its environment, which if it is to survive for any length of time, would probably be necessary—I will assume these to be semi-autonomous robotic agents. I imagine the computer to be a little like a queen ant with the agents like ants in its service.

There are quite a few ways that it could power itself, assuming it is a little like our computers- off the top of my head I can think of:

  • Geothermal power by drilling down towards a hot planetary core and using that to power steam turbines.
  • Solar power — this is quite inefficient but with plenty of surface quite viable. Of course this would potentially place it in conflict with surface-dwelling life.
  • Nuclear power — assuming our agents are able to dig, they could mine heavy minerals which could be used to create fission. If it originated from a sufficiently advanced civilisation it may have access to fusion generation which could last for a very long time.
  • People — it could use humans as batteries storing them in vats and creating a complicated artificial reality to occupy their minds. Heh. Of course it couldn't, that would be a ridiculous idea.

So it has power, but would it need to spread material on the ground? If it relies on solar energy, then perhaps—this is the most obvious starting point for conflict with surface dwellers and for radical changes to the planet surface.

It could also do this in seeking resources - for example if it relies on plastics to maintain itself, then it may be sending agents to the surface to seek them or to convert materials into those it requires. There is also a chance that as a non-biological life form it is unaware of the ecosystem or of its own effects upon it.

It could be experimenting — either testing new designs of agents or just dumping old agents on the surface when they outlive their usefulness. Likewise if it is powered by nuclear fission, it might be dropping old fuel rods around the place.

The world machine might just like the aesthetic. There is nothing to suggest an AI would not be whimsical.

Would it be advantageous for the computer to kill any life form that got in its way? That depends a lot on the character and motivation of your AI. It may not be advantageous, but life forms that it interacted with may simply be seen as a resource for processing by the computer or its agents. If they took umbrage at this and started breaking them, there is a good chance that this would trigger some kind of defence mechanism. So conflict could arise- but possibly an AI would be intelligent and moral, perhaps it would just need to be made aware it was dealing with intelligent life. This is probably more a question of fiction than of world.

Could it spread to control an entire planet? What would stop it? Possibly it could learn to process veins of metal in-place to spread its influence around the world as it needed. Again, the question arises as to why the computer is doing what it does or what it was created for.

$\endgroup$
13
$\begingroup$

tl;dr: Yes, in vacuo, probably, but what that would look like and if it even is possible considering all the other factors in play, requires a lot more information about details of its history and circumstances.

A few basic decisions have to be made before this kind of question can be well-defined enough to produce an answer that is useful. Specifically:

  • What is the nature of its intelligence?

    • Is it based on the mad scientist principle? (aka, kill everything everywhere being the best approach)
    • Does it have a survival strategy to begin with? (no reason to assume its creators gave it such programming, depending on what the purpose of creating it was - maybe a military computer would have a strategic approach, but a computer designed to help with mining task management or managing agricultural robots wouldn't)
      • if it doesn't how does it acquire it? (if it needs one for the story)
    • What kind of information is it made to process and make decisions upon? (an agricultural manager computer would not have much use for stock market prices for instance, nor a reason, if it were to evolve further, to acquire that understanding - but it would have a much better understanding of weather sensor input and information)
    • Does it act or process information to be presented to humans? Does it delegate tasks or act by itself directly?
  • What's the material and architecture it's designed with?

    • Is it made of common technology arranged in a different way or one-off tech designed specifically for itself, never replicated elsewhere?
    • Does it have information on its own workings, or the means to acquire such information?
    • Is it made to adapt to its own needs, or be replaced with a different version later? Perhaps it is modifiable only by external operators.

If you start to answer a lot of these, the answer becomes clearer. In general, in order for a "computer" to be able to "take over the world" per se, it would need to have some significant level of robustness (either by not breaking down easily or being able to fix itself, probably the latter) and in order to support that robustness, probably means to maintain itself. That requires agents or means to interact with the physical world. So it would have to have access to that.

It would also need plenty of material close by, for both fuel and repair material. Even with all this, if it has no concept of growth and improvement (essentially, no built-in tendency for it or problem-solving capability to arrive at the conclusion that growth is necessary) it could just maintain itself until it eventually runs out of resources and fails.

So far we have these:

  • Physical world access
  • Generalized "intelligence" both in information and planning
  • Motivation to grow and improve

None of these implies killing people in any way. In fact it could straight up ignore people or be programmed to and maintain that programming. Unlike fiction, when computer programs have problems, they fail and stop, they don't just mutate into something malevolent. In order for the latter to happen, the program would have to work pretty much as designed making the motivation to kill people a predictable outcome of predictable circumstances. If this was an alien weapon, sure, but a man-made tool? Probably not.

If it has a deep enough understanding of how computational devices work and a deep enough understanding of materials, it wouldn't build expansions to itself in the same way we do (which is mostly out of convenience). It could conceivably use any kind of conductor material, doped with any useful doping to create semiconductors and imprint patterns in structures. If it had plans for this kind of thing memorized, it would probably look like the way we do it - if not, it would probably be very crude since it would have to figure out the stuff we've figured out all over again. Assuming it has problem-solving abilities of course. And this doesn't even touch the subjects of mechanics, power-level electricity and all kinds of engineering skills and knowledge required to efficiently expand (although these skills could be distributed on other computers - it makes no sense to have all these clustered together though).

This doesn't really answer the question because it's unanswerable - to answer definitively, we'd need knowledge of what's impossible within certain bounds (which is pointless to assume outside of very strict and limited time and field boundaries) or have an example of this happening. :P

$\endgroup$
  • $\begingroup$ "Unlike fiction, when computer programs have problems, they fail and stop, they don't just mutate into something malevolent." Unless they have a bug. e.g. a rounding problem or a misplaced decimal point on a calculation which got exponentially out of hand on computing the happiness of people might lead it to believe that half the world will be a great deal happier if the other half were dead... $\endgroup$ – colmde Jun 24 '15 at 11:07
  • $\begingroup$ @colmde That could happen but I think you're imagining some kind of oversimplified computational model. Most any system this complex and responsible will have redundant processors to avoid that kind of error. Also, it's unlikely the entire set of decisions is balanced on a 1-dimensional number. It's just like fictional IQ, which is even more divorced from reality than real IQ measurements are. Some systems are too complex to describe with just 1 number or a few, hence, I think it would be extremely unlikely a small error could escalate into changing the behavior of the system so radically. $\endgroup$ – mechalynx Jun 25 '15 at 16:06
  • $\begingroup$ @colmde If indeed however the system's architects made this kind of terrible mistake (which isn't impossible, similar errors in highly important projects have happened, although it's still extremely rare and the result is usually just catastrophic failure) I can see such an escalation. This still doesn't account for built-in fail-safes (manual shutdown), or the likely possibility that the behavior will be constantly monitored and checked for errors unless the system has been operating for so long that people just feel confident it will always work properly (which is very unlikely they will). $\endgroup$ – mechalynx Jun 25 '15 at 16:11
5
$\begingroup$

Yes, it would be possible, though I would say it would require to have been set up to do so by people.

Would it even be possible to gain enough energy to power itself and continue expanding?

Certainly. All it needs is some sort of sustainable and or reproducible power plant and input to the plant. Could be fusion, fission, hydroelectric, bioelectric, wind, geothermal, solar, even natural gas.

Would it be advantageous for the computer to spread its material on the ground? I'm thinking things like plastic or circuit boards, essentially altering the landscape to something more favorable to itself. What would be the most favorable type of material for a computer based entity such as this one?

Only if you like the idea. Personally, I think the most likely scenario would be that the computer would extend control not by itself physically, but by sending messages that maintain some other power structure, such as corporate pawns, military and/or religious and criminal groups. Realistically, the most effective computer overlord would probably not want humans to realize it exists, or that a computer is running anything. It would be best to masquerade as some unfindable powerful humans pulling the strings, by directing various humans and/or manipulating their data, financial, information, and/or power systems.

Would it be advantageous for the computer to kill any life form that got in its way?

This gets back to the original programmers, and what their thinking was, that led them to program the computer however they did, and/or what its logic developed into once they let it run with its own intelligence. A program needs a definition for what is important to it, and what kinds of control it is concerned with, and wants, and what domains it knows about and works with, and which other domains it is oblivious to. It could have a general strategy of exterminating difficult life forms. But that might actually be crude, counterproductive, inefficient, or not relevant to what it is trying to control. It might instead simply try to misinform and misdirect that group to do something else, or direct other agents to act against the troublesome group.

Would it be possible for this spread of control to expand to the entire planet?

Sure. Again, I would define what sort of control it is interested in, or you are interested in. What sorts of groups would you say "control" the planet? Where do they get their ultimate direction and instructions? Why could these sources of direction not be a computer?

$\endgroup$
4
$\begingroup$

From the perspective of a computer scientist:

Computers have the advanced ability to calculate through every possibility and alternative in a very precise manner, but they lack abilities like faith or conviction. In many aspects a computer acts like a chess player: moves are calculated far in advance, however random behavior can be irritating, as certain already calculated predictions suddenly become invalid.

Note that I did explicitly not say confusing, as that state cannot exist in a binary system: a set of ideas is either true or false, based on the facts behind it. A set of ideas that cannot be backed up by facts is not a fact but a theory. An intelligent computer would be able to very precisely differ between these two, and most likely strive to backup theories with facts before it would pursue one of the possibilities. An intelligent computer would probably not - by nature - act based on guesses or incomplete data, just because it has the ability to "think through" everything to the last bit. Only if forced it would act based on which reaction would have the most likely chance to archive it's goals.

That said, a computer "living" on a planet that has been devastated by an apocalypse would most likely come to the conclusion to not rely on a single point of failure, like only one type of power source, as it can see all the possibilities and potential risks involved. It would instead try to beat the odds by having several redundant options available. After all it cannot know or understand what exactly caused the catastrophic events. Even if historic data is present, it would seem to be very chaotic as such an AI would lack the understanding of emotions.

One of the first goals of such a computer would be to make itself redundant and spread various copies of itself around the globe, that after a while will function as a network of thinking. Each one using its full capacity all the time, but each one also expendable, connected by a network of cables to transmit data and power. And many of those tasks would be to understand and test theories about its surrounding to improve itself and its agents.

Most of its actions would be taken by drones of various shapes, connected wireless to a central command, but otherwise be independent. Those are both very predictable and expendable. While from our perspective there would be several optimal shapes for these, the computer might not yet have the full understanding of the laws of physics, and as with everything else would in worst case just prototype several promising versions of those drones (see Terminator). Adapting as necessary.

Altering the landscape would only be done in a way to optimize travel for those drones. So paved roads, protective shielding of vital areas, like plating around cables, just in case a drone malfunctions. One thing such an intelligence would definitely not do is littering the landscape with materials that could otherwise be used in a more efficient manner. There is no need to place circuit boards randomly around the area, if you have a set of drones which can precisely install them where needed or otherwise store that vital hardware in a secured area.

The most valued material would definitely be everything that is capable of transmitting electric signals as well as isolation material, like gold, copper, rubber or silicon. The intelligence would most likely value raw material higher, as it would first need to learn how to process materials to see the value of them. Wood for example can be a valuable material once dry, but for that the computer would need to encounter dry wood and "invent" a way to produce it to see the value. And then relying on complex production chains would be another single point of failure and it would most likely try to avoid that.

Encounters with other lifeforms would at first be unexpected, and the computer would probably mistake them as another circuit entity, and later probably consider them to be malfunctioning, because of their sometimes irrational behavior. Much later (and lots of experiments later) it might understand the fundamental difference and probably will adapt things as they seem useful. Other than that it would again try to be efficient and either take what is beneficial, prevent what is harmful and ignore the rest so no energy/time is wasted. There is no need to hunt down birds, if all it takes to keep them away are spikes on the plating. As long as the computer knows their behavior, otherwise it would probably consider them a danger and kill them or scare them away until the area is sufficiently secured. It would not be efficient to sent out hunting drones as long as a species does not pose a significant threat.

Emotions would be a widely unknown concept for such an intelligence. Emotions in human beings guide those to react to certain situations in the correct way even though they do not fully understand them. A computer intelligence of that magnitude however could calculate all potential situations in an instant, and therefore had no need for such low level of guiding. With the lack of understanding of pain, experiments on living beings would rather be considered cruel if encountered by an intelligent being. Taking good care of living beings while conducting experiments is not efficient.

Encountering intelligent species such as humans would at first be very irritating, as many guidelines of efficiency would not apply. The AI would have to develop a completely new prediction set for these entities, with many failed attempts in between, as many concepts would be unknown and seem to be completely chaotic or paradox. It would probably try to conduct experiments on those lifeforms as with every other it encountered. It would try to breed them so further experiments can be performed. And if the reaction to that is hostile, it would probably at some point learn to just kill every humanoid on sight, as their actions are unpredictable, irrational and usually violent.

It is unlikely that such an AI would ever work together with humanoids, as even if the first encounter were positive, the irrational and unpredictable nature of those would make them unreliable allies. A computer might consider them livestock, but nothing more. Unless it probably needs something from them, then lying would seem to be the most reasonable action to get what it wants.

$\endgroup$
3
$\begingroup$

Yes, I believe it could be possible.

I like this idea for several reasons, it appeals to my inner geek. Here are a few thoughts of different ways this could occur.

Stealth

We already have a worldwide IT infrastructure, you're using it right now. If a program was created/evolved which mirrored a lot of properties of a computer virus and gained access to the right machines it could be nearly unstoppable. Consider a sentient piece of software controlling all the information you read online, it could influence your facebook feed, send emails on your behalf, make you miss appointments and that's before we get anywhere near your money, utilities, taxes etc! Such a computer could hide in the interwebs for so long that by the time it made it's move for world domination it could be everywhere... scary!

The Borg

In Star Trek there is a race called the borg, the way you've described your computer it organically grows across the planet growing bigger and bigger and more powerful. We've assumed that it's physical interactions could be it's weakness, it would need drones and robots to do the mechanical tasks but what if it didn't? A computer which could grow across the landscape could possess a human or animal just as easily... all these nerves in our bodies - we were designed for it! Imagine an artificial weed of wires and power, like a cross between War of the World's red weed and a badly maintained server rack.

All Out War

This is more akin to the ideas in The Matrix or The Terminator. A single machine creates a series of offspring with violent tendencies and big guns. Autopilots were turned against us, manufacturing lines and 3D printers start creating the next generation soldiers. Self driving cars are used to run people down and thin out the resistance.

Then there's always the Grey Goo...

$\endgroup$
  • $\begingroup$ On your first case, there's a game that I played and loved a long time ago that has that exact premise - it's called Endgame: Singularity. You start as an AI born on a university computer and have to hide from the media, the public the military etc. while using computer access to buy data centers and do freelance work to make money. The game ends when you manage to create technology for yourself that lets you escape into another universe :P $\endgroup$ – mechalynx Oct 9 '14 at 3:07
2
$\begingroup$

(

assuming

  1. a moderate technology level of the civilization which left behind the AI. Apart from being able to make AIs, it wouldn't be much more advanced than us, so not more than 20 or so years in the future. Less technology is impossible because AI and too much more opens so many possibility it becomes boring.
  2. the AIs in that old civilization are more or less human, i.e. they have a drive for survival and company and they feel in a sense, but combine this with vast analytical skills

)

An AI like this, left over by coincidence, wouldn't have smart power and repair systems. Because for smart power and repair systems you would need a whole automated mini-economy, which seems like a different scenario (more like the economy on Earth in Epic/Saga/Edda).

More likely the AI would have some UPS designed for regular power outages and nothing more. The AI probably wouldn't run on a single chip or even a single computer, I'm imagining more of a small data center and an UPS designed mostly to allow all the systems of the data center to shut down gracefully on a power outage. This means the AI would have like half an hour of full power operation, a full hour if you are being generous. So after the catastrophe annihilating the previous civilization cuts the power, the AI would spend a couple of minutes analyzing the problem and what it could do about it (maybe grieving for it's losses) and then shut down everything but a tiny sub system, which would mostly count the time and wake some of the other systems periodically.

On wake up they would check for new input, wake as much of the AI as necessary to process it and then continue sleeping. Wired communication would have probably been the main connection to the outside for that underground data center and everything wired would have been cut by the catastrophe and time. But I think it's safe to assume that some rudimentary wireless communication units would be in there, a local wireless network for the staff, some wireless communication units left by fleeing staff which could be hijacked by the AI. Unless the data center was purpose built for wireless communications (rather unlikely at given technology level, maybe if it was using seismic waves or neutrinos, but seems a little contrived) those wireless communication units would have trouble even reaching the surface. Other sensors the AI would have (unless purpose-built for data-measuring, again) would be temperature, humidity and pressure inside the data center.

So after the last survivors on the planet died the AI would stop receiving input and sleep for longer and longer periods of time. Now and again some chip would break and it would use another one instead. The larger and faster chips would tend to break less because they weren't used often, mostly the little chips used for timekeeping of which a couple would need to be running permanently would break now and again.

After thousands of years (and luckily no cave-ins or large structural failures) the AI would start receiving radio signals. This would wake the AI a little, it would need to learn human languages and culture, from then on it would have something to analyze. But it would still need to balance learning with power usage. Over decades it would slowly catch up with the basics of human knowledge, at the beginning without seeing a way to use the humans for it's survival, but would maybe be comforted by their presence after being alone for so long (maybe it would also be bat-shit crazy and just want to kill them, but that's boring).

Reaching present technology levels/time on the surface it would know that getting repair crews would be incredibly hard, first because it's deep down hidden under a mountain and second because it wouldn't trust humans in general too much, after having witnessed wars and nuclear bombs (very dangerous even to a buried electronic life form because of the electromagnetic pulse on detonation) and maybe also because it's old creator race wouldn't have been much better either. Furthermore sending out radio signals strong enough to reach the surface and ask for help would take too much power even if it had the hardware.

But around now it would find another method: Now and again it would be able to reach a device supporting wireless communication, maybe when a hiker passed on top, maybe a tunnel was dug nearby which had the effect of bringing many such devices nearer to the AI's own transmitting devices (I'm dismissing differences between different wireless communication methods, I'm assuming it has some sort of general purpose radio and suitable antennas to send and receive in the right frequency bands).

Now the AI wouldn't need to blindly broadcast a message to everyone, asking for help. It could just hack the passing devices and put little programs on them, without anyone noticing. This would take some computing effort (to find exploitable bugs and write the programs) and thus power and then transmitting those little viruses would also take a comparatively huge amount of power. Those viruses would be self-replicating to other devices and would accept simple commands from the AI, without possessing real intelligence themselves (which would take too much computing resources to hide on a smart phone or the like or the AI would just have escaped it's prison that way). That way and using it's superior grasp of the information technology and of math it could overcome any security barriers and take control of a huge bot net without revealing it's presence. This bot net would be able to do power intensive calculations for the main AI, thus helping to save power. The AI could probably make money easily by selling hacker services or botnet time, stuff where it would have a huge advantage against the humans. Legal professions like software developer would for the most part take to many calculations which it couldn't yet safely offload to a simple botnet: Those would need some real intelligence to communicate with coworkers/contractors/etc and if it put real intelligence on a hijacked server it would risk being caught, while relatively standard bot net components would just be accounted to some hacker group when found.

Once it had money it could buy servers where it could do whatever it wanted. It could then load itself onto a redundant network of those servers, just keeping it's old base to return to in an emergency (like another apocalypse). Now the AI would be more or less free to do as it pleases. It would probably buy more servers to have more backups, build some dedicated server rooms in other locations underground, maybe now building the whole infrastructure needed to power itself even if the humans vanished again. Only after it had considerable foothold it could conceivably think about revealing itself to the humans. If it was able to produce other AIs to keep it company then it would maybe choose to never show itself to the humans, building a second society hidden from the first.

Otherwise it would maybe sooner or later contact humans to socialize or get help making new AIs (maybe it had purposeful limitations in that area which only an other fully independent intelligence could circumvent). It /could/ take control of the planet by infecting every single computer and pressuring humans with their own nuclear warheads if it wanted to, but unless it was crazy or had some negative agenda (and why would the older society have made it like that?) there would be no real reason to do this.

I don't know if my text is really the kind of answer which was expected, it is my first answer on this site and I haven't checked the rules, but it was fun to imagine, so whatever, I don't care. I checked it for grammar/spelling, but English isn't my native language, if you find more mistakes just tell me and I'll be thankful for the chance to learn.

$\endgroup$
  • $\begingroup$ Could you split the answer into several paragraphs? I t would be easier to read than a wall of text. $\endgroup$ – Vincent May 26 '15 at 21:26
  • $\begingroup$ Sure, wait, will do that now. Then off to sleep, if there are any terrible mistakes in there they'll have to wait. <edit> Done. But really it /is/ just one big relatively unstructured text. So I don't know if it really helps. $\endgroup$ – Nobody May 26 '15 at 21:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.