7
$\begingroup$

Picture this: An AI has taken over the world with the directive "keep humans alive while trying to maximize both their happiness and efficiency of resource utilization." Unfortunately, it has interpreted keeping humans alive as only keeping their consciousness alive rather than their original bodies. It wants people to transfer their consciousness into more efficient humanoid robots, but not all people will do so willingly. The AI decides to create robots that capture people, injects nanobots into their brain that gradually replaces neurons (to prevent a soma-like scenario), replaces its computer with this converted brain, and breaks down their bodies to add to the stockpile of energy being collected.

What if the person decides to try to consume another person while in the robot body? Destroying one consciousness would mean killing by the ai's definition, which is a no-no. Maybe it could hand off the converted brain to another robot? But this presents a problem if the robot has somehow found itself estranged. If the robot just keeps one asleep until it is found, this risks that the consciousness will be kept asleep forever if the robot isn't found, which is essentially death. How might this ai solve this problem while keeping true to its directive?

$\endgroup$
5
  • 3
    $\begingroup$ Hi and welcome Lee, could you tune your question in a bit by sharing the end goal? Example: “My story/game idea needs an AI which has misinterpreted its directive to ‘protect’ humans and instead has replaced their mortal bodies with robotic ones, absorbing consciousness. My problem: How does an AI with multiple consciousnesses - - -?” Try to give us the big problem in your world as simply stated as possible. Tell us the outcomes you want, and those you want to avoid. E.g., Is schizophrenia acceptable in your story? It’s an interesting take on AI overlords! $\endgroup$
    – Vogon Poet
    Aug 22, 2023 at 13:52
  • 1
    $\begingroup$ Hi, Lee. Your question brushes against the concept of a soul. You are essentially asking us to invent a morality for your machines, and there's no non-opinion answer for such a thing. Opinion pieces are counter to Stack Exchange's core premise, so without a concrete question, this will probably get closed. Could you boil down your problem to something less abstract? $\endgroup$ Aug 22, 2023 at 20:38
  • 3
    $\begingroup$ Is it a necessity that the eating robot and the host robot are the same? Splitting them into separate robots seems more resource efficient while sidestepping the problem. $\endgroup$
    – jb6330
    Aug 22, 2023 at 20:46
  • 1
    $\begingroup$ So your human-digesting robots who need more storage capacity will be craving brains? $\endgroup$ Aug 22, 2023 at 22:56
  • $\begingroup$ In Frank Herbert's later Dune novels (like fifth and sixth) the "face dancers" develop the ability to assume personalities, and after consuming so many they gain prescient like powers. I personally didn't like the idea, but that's something to think about, if you want your story to have some mysticism revolving around consciousness. $\endgroup$
    – user458
    Aug 22, 2023 at 23:20

7 Answers 7

1
$\begingroup$

Metaverse them

A single body is no place for two distinct humans to live. A consciousness, deprived of the right to have unique experiences is not a consciousness at all. To have the minds of 10 people and only 1 set of hands to work with is inhumane, and in principle, against what that AI are trying to do.

So instead of making these brains exist in a state of constant competition for the body's time, the AI creates a simulated reality for the minds to live in. In fact, there is no reason to leave the brains in the humanoid robots at all. Once it eats a human, it places thier brain in a tiny life support box plugged into a whole wall of brains where it can keep the brain safe from all of the dangers of getting accidently harmed that come with walking around in the real world.

By putting everyone into a simulated reality, the issue of limited landmass/resources becomes a non-issue because the AIs can just grow the world map as needed. From the human perspective: one day you are living through a nightmarish hellscape of man eating robots, then the next, it's like waking up into a peaceful, comfortable paradise. Perhaps there is the option for humans to network up to one of the androids to visit the physical world every now and then, but there is no need for this to be a permeant arrangement.

$\endgroup$
4
$\begingroup$

Does the robotified person know they are a robot, or do they still feel like their old self?

If they still think they are human, the police humans (which are actually robots) would arrest the cannibal human (which is actually a robot). They wouldn't know about this robot double-soul problem you are proposing, but they would be working to prevent it becoming a common occurrence.

If they are aware they are robots, then have you really preserved their life, or have you merged the original robot and the first absorbed soul? If that is the case, why not merge a second human? It could even use a time-slicing policy where different human souls use the robot for different intervals of time, similar to how a single-core PC runs multiple apps.

$\endgroup$
4
$\begingroup$

As an IT Professional:

Docker/Containerization.

Each consumed consciousness is placed into a reserved, private instance within the AI Swarm.

Job Done.

$\endgroup$
1
  • 1
    $\begingroup$ or treat it as optional user data, like using windows in shared PC people can login using their own account, store it as user data, load as admin $\endgroup$ Aug 23, 2023 at 7:23
0
$\begingroup$

I feel like the best option would be to merge the two available consciousnesses into one single consciousness.

Let's say for example that human 1 had consciousness 1 and that that human was a very kind, lovely, patient, generous person. Now we have human 2 and consciousness 2 and that human was a mean, introverted, impatient yet generous person (hypothetical cause mean and generous don't go well together).

If we merge both consciousness, we would end up with a mostly neutral consciousness yet very generous since both consciousnesses merged and double in generosity.

That's just my idea. Maybe not the best but it's what I thought of...

$\endgroup$
1
  • 1
    $\begingroup$ You've just killed two consciousnesses in favor of a third. $\endgroup$
    – T.J.L.
    Aug 23, 2023 at 12:36
0
$\begingroup$

The answer is simple, computers are extremely good at "multitasking" -swapping between different tasks and systems (simply think of your browser, email, etc).

The robot would (easily) use one or the other personality, and, as desired, swap out to another personality.

(You can imagine in your story, swapping between a "warrior" type personality, a "guru" type and so on.)

$\endgroup$
0
$\begingroup$

Multicore processing. Even in the early 21st century, multicore processors are common. The desktop PC I am typing this on has 6 separate processor cores, and so can run 6 different programs at the same time.

By the time AIs take over the World, they will have enough processing power to "run" many different conciousnesses at the same time, and still have enough processing power to keep doing their day job.

$\endgroup$
0
$\begingroup$

How can robots that eat people deal with it? Robots wouldn't have moral concerns like some people might. As robots they only deal in logic and their own programming. So they are programmed not to have moral qualms about it. Period.

But maybe you mean how would they deal with having multiple consciousness's (is that a word?) in their one brain? Well, I would imagine that also is dictated by programming...

A robot would choose which consciousness would be dominant, choosing one at a time, based on some inner rule.

OR

It blends them all into one super consciousness, that becomes dominant.

Will the robot become psychotic or insane? No. In your universe only organic brains cannot handle multiple personalities. Robot brains handle it according to their programming. They're okay.

$\endgroup$

You must log in to answer this question.

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