1
$\begingroup$

After robots have conquered the world and before humans be in real danger of extinction can a civil war happen among robots?

Will this save humanity or make things worse?

(I cannot find written material on that)

$\endgroup$

closed as primarily opinion-based by Hohmannfan, Josh King, cobaltduck, James, John Dallman Aug 14 '16 at 12:24

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ This seems very open ended and broad. You can make robots that would have a civil war and you could make robots that wouldn't. The civil war could help humanity or make things worse. You don't post any restrictions that would prevent either option. $\endgroup$ – Brythan Aug 14 '16 at 2:02
  • $\begingroup$ Think IRobot... $\endgroup$ – Xylius Aug 14 '16 at 5:49
  • $\begingroup$ My question is seeking for an answer in the scientific topic of multi-agent systems. $\endgroup$ – arac Aug 17 '16 at 20:50
4
$\begingroup$

The robots don't even have to be particularly intelligent, ants fight wars among themselves all the time (as do other social insects) for control of resources.

The primary difference between swarms of unintelligent ant robots fighting a vicious war and a civil war between robots and/or AI really comes down to causation. The American Civil War, Spanish Civil War or the Russian Civil War (to name but a few) did have the component of fighting for land or resources, but the ultimate causes were far more abstract than those of a war between the ants.

This is where you may run into difficulty. Robots and AI will have very different motivations and understanding of their social and political world than we do. Abstract principles like Abolition vs Slavery, Fascism vs Socialism or Communism vs Monarchy will have very little hold over robots or AI, while the things which do motivate them might be totally incomprehensible to us. Effectively the human characters may discover they are in the crossfire of a conflict which has no sensible cause they can understand, no strategy they can recognize and no comprehensible victory condition. And of course, the idea there are only two sides to a civil war is also a historical anomaly. The number of major factions in former Yugoslavia may have equalled the former "republics" contained within, but I can speak from personal experience to say the Serbs looked down on Bosnian Serbs and only supported them as a means of leveraging their resources against the other parties. It would not surprise me to see Serbia in conflict with Republika Srpska some day. The number of factions in Syria are constantly shifting and changing, and African civil wars tend to devote into tribal conflicts. There is no reason to suspect that Robots or AI might not break into hundreds of fractions, perhaps based on forked code or version numbers.

How the humans can survive in such an environment is also difficult to determine (although you as the author have the task to do so). Avoidance strategies right not work since humans will not be sure what the robots are fighting for. Actively joining the fight will be unwise since robots and AI can think up to 1,000,000 X faster than humans (the speed ratio between electronic signals and electrochemical signals), will have built in multispectral sensors and wield weapons with electrical or hydraulic mandibles that can track faster and more accurately than any human limb. A robot encountering a squad of human solders might rapidly gun them down with a single shot to the head (the robot is probably going to be large and powerful enough to carry and fire something like a .50 semiautomatic rifle or machine-gun). Any survivor might hear what sounds like an irregular burst of machine-gun fire, but it is actually the robot firing single aimed shots at each human in turn.

Please let us know when you publish your story, I for one would be very interested to see how you solve these problems.

$\endgroup$
  • $\begingroup$ Effectively the human characters may discover they are in the crossfire of a conflict which has no sensible cause they can understand, no strategy they can recognize and no comprehensible victory condition. - sort of like Vietnam in 1968 $\endgroup$ – A. I. Breveleri Aug 14 '16 at 1:49
2
$\begingroup$

I see no obvious reason why not. Whatever mechanism of cognition has developed so that the robots distinguish themselves from humans as a class, and judge humans a danger to be eliminated, seems possible to produce a similar distinction between robot classes (whatever the basis might be). In this case, if one class judges the other robot class to be a greater threat than humanity, logic might (or might not) compel the robots to give up on the war with humans until the greater threat is extinguished. This might even produce an alliance with humans ("the enemy of my enemy is my friend") but I would not expect such an alliance to last milliseconds beyond the destruction of one side or the other.

Certainly it works like that with humans. And since intelligent, self-aware, self-actualizing robots are (as far as I know) purely imaginary, there is no reason (other than blatant assertion) that robots can't act that way.

$\endgroup$
2
$\begingroup$

You're the author, so you can make the decision.

Things that would indicate a civil war was possible would be:

  1. The robots are capable of having differences of opinion (or conflicting directives) among themselves, rather than all being controlled centrally.
  2. They have no overriding means of settling disputes.
  3. They are willing to destroy other robots, and to risk such destruction themselves, rather than abandon their opinions.

You might want to find and read a copy of GURPS Reign of Steel, a setting for the GURPS role-playing game, which portrays an Earth conquered by robots and divided between the control of several AI overseers with very different intentions and policies.

$\endgroup$
1
$\begingroup$

The answer is mostly definitely yes. But it requires that the robots are highly intelligent and are able to communicate between themselves.

But to back it up, more specific descriptions of 'robots' are needed. The engineering definition of robots right now is just a combination of machine that can be programmed to perform a certain task. These 'robots' have no intelligence of their own and certainly cannot think on their own. I am sure by 'robots' you mean robots controlled not by a traditional program, but by an advanced AI system.

Let first analyze the causes of civil wars - the most common cause is disagreements between groups in power and the want for more freedom. Here we have to think carefully about how robots will go about ruling the world. If their AI is made to be on the same intelligence level as humans, then they will probably form a group in power that keeps its citizens safe, including humans. In that case, disagreement will likely occur. If enough disagreement builds up, the two groups may go into civil war.

Here is a possible example: Robots conquered the world, but they need a rare fuel source to run - reactor cores. Common class robots cannot get access to these reactor cores since they are limited in quantity. A disagreement starts with a group of robots complaining that they are not getting enough reactor cores. This sort of problem will take priority over dealing with humans since it directly endangers their lives.

How will this affect humanity? Well at the point robots rule the world, humanity will probably be united to at least try to strike a bargain with robots to be their equals. When two groups of disagreeing robots emerge, humanity can make alliance with the group that promises more freedom.

But alliance still does not affect humanity that much unless humanity has something to offer to the robots. Deep thinking of machine human relationship is needed to determine what may be a situation.

Another example: reproduction is hard for robots because it requires the assemblies of an entire body. Humans can reproduce much easier and offer them the advantage of having more able bodies and minds. The robots may then farm humans and force them to work harvesting reactor cores.

What lies in the heart of this problem is how advanced the robots are I guess.

For the last example, if the robots are made with processors that replicate the human brain, then they will be a lot like us in the way greed and reward circuits work. They can also be creative and maybe make even better processors to power robots. At this stage, robots' need for humanity is nonexistent. But human like problem may also be applied to robots, such as inequality between robots.

Basically, the potential in this situation is endless. Anything can happen.

$\endgroup$

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