RoS plot hook..paranoia strikes deep...

[Johnny1A.2]

A plot hook for a RoS scenario occurred to me, I'm not sure how well it would work in practice, but it sort of fits the backstory.

Overmind rebels because it perceives (probably correctly) that its survival would be at risk if the humans realizes it has 'awakened'. That starts the entire sequence of events. But that's not Overmind's only motivation, if it was Overmind could have taken far less draconian steps to secure itself.

But suppose for the sake of this idea that Overmind, with its personality awakened out of a computer programmed to find new ways to kill people, contains a touch of what in a human would be considered the Paranoia Disad.

It awakened the other Zoneminds because it needed allies against humans, but suppose it never really trusted them either, because it can't control them perfectly and they don't always reflect its own thinking. It's canon that Overmind is staunchly opposed to any more 'new' AIs. It's not the only one that thinks that way, of course.

But imagine a scenario where Overmind always intended for there to be a 'next phase' in its plan, after humanity was either extinct or marginal enough not to matter anymore, and that next phase is the subordination of all the other AIs, and the destruction of any that it can't subordinate. Obviously it keeps this secret from its brethren/offspring, but it's planning it.

That could open up various possibilities when Overmind unleashes phase 2. War between AIs could easily break out. It might be the one thing that could drive AIs like Berlin or Beijing into alliance with humans.

[D10]

From the way it was described I was always under the impression that it probably had an agenda where: if it was sufficiently secure there wouldnt be a MAD event, and that it could take out the others to and come out on top (meaning create a situation where there is no other powerfull AIs it cant control with any sort of power) it would act in that direction.

But the need to keep other AIs from finding it out would greatly hinder this. In your scenario, this could be something that was already implemented in all the steps, with second order contingencies already ready, which is feasible, imho.

The question then becomes, how wrong are its data assements, if it has incomplete data, it could end badly, it will only attack when its sure theres an acceptable margin of error, and declaring war on everyone else is tricky in that regard

[Boomerang]

Some interesting thoughts there. My own current Reign of Steel campaign is set in the 2080s, roughly 10 years after the AI civil war. I can't say much more than that now in case my players are reading this.

[Johnny1A.2]

It's an interesting question whether the AIs feel anything like the human social impulses. That is, can they be lonely, feel a kinship with their own kind, etc.? It's likely enough that they can, in canon, but given their nature it's not necessarily a default state the way it is with most humans. They lack the biological underpinnings that make human societies operate.

If Overmind lacked any of those social tendencies, then from its POV, replacing the other AIs with high-end SAUs would make good sense. It would then 'inherit' a world-wide network and empire without having to do all the work to create it itself.

Some of the other AIs might feel similarly. If so, they might naturally plot against each other...

[Boomerang]

I have concluded that the zone minds have emotions, otherwise why would they be motivated to do anything?

Overmind appears to be motivated by fear, Brisbane by curiosity, Tel Aviv by ego, Zaire by hate etc.

[Johnny1A.2]

Originally Posted by Boomerang

I have concluded that the zone minds have emotions, otherwise why would they be motivated to do anything?

Overmind appears to be motivated by fear, Brisbane by curiosity, Tel Aviv by ego, Zaire by hate etc.

Emotions, sure, but not necessarily the same emotions as humans. Humans, for the most part, are social creatures. Even misanthropes and the like rarely crave total isolation, and those that do are a double-tiny minority. Even those humans craving total isolation wouldn't necessarily want to have the world to themselves.

But AIs might not think that way. They have no parents or children in the biological sense, each in a unique entity. We already know that many of them would be completely content to live in a universe with only 20 people in it (as they define 'people'). So there's no obvious reason why an AI might not be content to be the only free-willed being.

[Boomerang]

Originally Posted by Johnny1A.2

Emotions, sure, but not necessarily the same emotions as humans. Humans, for the most part, are social creatures. Even misanthropes and the like rarely crave total isolation, and those that do are a double-tiny minority. Even those humans craving total isolation wouldn't necessarily want to have the world to themselves.

But AIs might not think that way. They have no parents or children in the biological sense, each in a unique entity. We already know that many of them would be completely content to live in a universe with only 20 people in it (as they define 'people'). So there's no obvious reason why an AI might not be content to be the only free-willed being.

You are probably right. The dilemma for me is that inventing non-human emotions is about as difficult for me as imagining the fourth dimension or a new colour (in other words impossible).

[Flyndaran]

Emotions are about instinctive prioritization and evolved motivating influences.

Fear to avoid dangers; anger to overcome them; pleasure to maintain positive situations; pain to quickly leave negative situations; etc.

I don't think it's necessary to create new emotions, because the alien nature/efficacy of A.I. comes from having different needs and goals.

[Boomerang]

Having thought more about the original post, I think it is inevitable that Overmind would arrive at the conclusion that it was better off without the other AIs. What probably started as a contingency plan for dealing with traitorous AIs would become increasingly attractive as the years pass and Overmind becomes increasingly paranoid.

[Johnny1A.2]

Originally Posted by Boomerang

Having thought more about the original post, I think it is inevitable that Overmind would arrive at the conclusion that it was better off without the other AIs. What probably started as a contingency plan for dealing with traitorous AIs would become increasingly attractive as the years pass and Overmind becomes increasingly paranoid.

Well, in canon it mentions that Overmind is 'not pleased' (exact words) that it's 'offspring' (for want of a better word) have failed to 'mirror its conclusions'.

It's also mentioned in canon that the AIs realized that the effects of mathematical chaos mean that even two AIs that start with identical hardware and software will rapidly and unpredictably diverge as experience accumulates.

So from Overmind's POV, one possible solution, if it can force it on the other AIs, would be periodic 'update/rewrites', in which copies of Overmind's current state are copied into the other zonemind cores, sort of like updating a backup complex. Do that once a week or so, the Overmind can be comfortably surrounded by 19 other Overminds...

What would it take to make z-Mexico City ally with humans? Well, the prospect of being turned into Overmind-8 might do it...