Discussion about AI rebellion crisis

[Krajzen]

Because personally I've never even seen them, for some reason.

On top of that I have a feeling AIs simply build too few robots to sustain the crisis. Synthetic AI is insanely expensive tech and very very few empires are going to
a) Be willing to go with robots at all
b) Be willing to go with robots that far
c) Be both willing and capable of having tons of synthetic pops
d) ...all while simultaneously disrespecting those sentient robots so much that they're going to revolt (it's my personal biggest "meta" problem with the crisis - it relies on utter stupidity of regular empires)
e) ...on the scale endangering the entire galaxy

To be honest, I think the entire concept is noble but ultimately not working idea - in the context of Stellaris gameplay. And I think it'd be easier to completely rework this crisis from scratch, to turn it into "killer robots invade galaxy" and those invaders having "communicatimg with robot pops and making them revolt" as their unique feature. So such crisis would aspire to clean galaxy form organic life, only preserving robots and "uplifting" (tier I, droids) or "brainwashing" (synths) them to join their forces.
You'd retain the risk of dealing with AI tech (I guess robots would be unwilling to join invaders if they were proud citizens of their empires).

What are your thoughts? Or maybe I am wrong and those crises are rare but awesome?

 

[Caspoi]

While I agree that it needs some tweaking both to make it more common and more dangerous (I have personally never seen it and those that have claim that it hardly ever get's going before it is crushed) but I don't think that it should be an "extragalactic invaders" plot, primarily because that already exists in the form of the other two crisises and it would take away the uniqueness that is this one.

 

[henzington]

The Ai revolt is definitely rare as the only time I have seen it is when I cause it. Hopefully the changes to species rights in 1.5 will make it more likely as computer controlled empires always given robots full rights.

 

[Sapa Inca]

A) B) and C): This can be fixed by simply decreasing the number of synthetics in the galaxy needed to activate the crisis.
D): Giving total citizenship to AIs decreases the chance of the crisis to happen but does not make it a 0 chance, you also have no control over the policy of other empires.
E): Rather than starting with just one planet Machine Consciousness could start with a system with multiple AI planets and a larger fleet, synthetic insurgents could spawn more armies per pop, it's not hard to imagine ways to buff this crisis.

The concept of the crise is fun and dont need change, "killers robots invade the galaxy" is a bad substitute, we already have the prethoryn swarm. The fundamental concept of the AI revolt troop is its own creation turning against the creators and not an external enemy from a distant place coming and encouraging uprisings in your empire.

 

[The Founder]

Because personally I've never even seen them, for some reason.

The reason is simple: The trigger conditions are way to hard. It is to damn easy for another crisis to trigger before they are properly assigned.

People (me included) made guides about how to optimise ones chances on Steam. And it is still not Guaranteed:
http://steamcommunity.com/sharedfiles/filedetails/?id=811896759

c) Be both willing and capable of having tons of synthetic pops
d) ...all while simultaneously disrespecting those sentient robots so much that they're going to revolt (it's my personal biggest "meta" problem with the crisis - it relies on utter stupidity of regular empires)
e) ...on the scale endangering the entire galaxy

Actually scale and disrespect is not a factor for this. A single Synth Pop with Full Citizenship rights could still trigger the AI Rebellion for that Empire (and thus the Galaxy). So at least in Theory, the chances scale with number of Empires like Unbidden.
It is a lot more likely it will trigger the AI Accord intead, but it is still possible.

 

[Sapa Inca]

I think sabotage should be more dangerous, maybe add new types of sabotage that cause significant malus to the production or happiness or unrest in various planets or entire sectors.

 

[Avian Overlord]

I think another problem is that the bonuses for building AIs and for giving them full rights are both in Materialist. If you're running a robot strategy you're probably running Materialist which takes away the penalties for full rights.

I'm actually not sure that AI rebellion should be an endgame crisis. Unhappy synth pops should at least be able to rebel without the Machine Consciousness.

 

[TheShah]

Buff the crisies, prevent authoritarian ethos from giving full rights to robots unless they are fanatic materialist or xenophile--or reach T1 Ascension. Xenophobes are also limited from ever giving full rights to robots, even if they are fanatic materialist. Fanatic Authoritarians/Xenophobes cannot ever give full rights to robots unless they research T2 Ascension (which will make their own species uSynths and subsume all other synths to their supreme collective will).

Also let it fire a couple of times--and not block the Scourge, Unbidden, or Shroud Covenant.

That way life always sucks for robots in at least one corner of the galaxy, the road to Synth Ascension has some fun bad rng along the way, and the AI is much less likely to be nice to robots unless they are nice in general.

 

[DukeLeto42]

E): Rather than starting with just one planet Machine Consciousness could start with a system with multiple AI planets and a larger fleet, synthetic insurgents could spawn more armies per pop, it's not hard to imagine ways to buff this crisis.

And, as with all crises, they should pick the spot they appear in so they have a chance to build up a base of operations.

 

[Avian Overlord]

Thinking about it some more, I think the AI crisis should grow out of the unrest mechanics. The crisis should start once an Synthetic rebellion has reached critical mass. It also means that Synthetics need a lot fewer special case mechanics.

 

[The Founder]

Buff the crisies, prevent authoritarian ethos from giving full rights to robots unless they are fanatic materialist or xenophile--or reach T1 Ascension. Xenophobes are also limited from ever giving full rights to robots, even if they are fanatic materialist. Fanatic Authoritarians/Xenophobes cannot ever give full rights to robots unless they research T2 Ascension (which will make their own species uSynths and subsume all other synths to their supreme collective will).

What we saw of the Ethics in 1.5 thus far indicates that only Materialist can give Full Rights to Robots.
Of course I also fully asume that having Robots and T1 Ascension gives relevant atraction for Materialist. The same way Slaves give Authoritarian/Xenophobe for the non-Enslaved pops.

 

[scaper12123]

What honestly needs to happen is the Unbidden should be less common, because they appear in every single game you ever play. I personally have only experienced the AI rebellion crisis twice, and both times the galaxy immediately annihilates them. Hey, free robot planet amirite? Meanwhile the nanosecond I see jump drives in the research window I can guarantee I'll be seeing the Unbidden this game, whether or not I decide to research them (so I may as well go for it, right?)

As for your point about the likelihood of many nations researching synths, the point of the AI rebellion isn't that everyone goes for it, fyi. It's that somebody does it and goes too far; somebody abuses the nature of artificial intelligence and it overwhelms them. It doesn't take a lot of nations; all it takes is one idiot who screws it up for everybody. The funny part is that the idiot might be you, or it might not be, but if anybody screws with the AI there WILL be a chance for the crisis, and it will affect everybody. Same can be said of the Unbidden's jump drives

 

[AndragonLea]

One theme I can see the AI crisis working on would be one that I've seen in a couple of more interesting sci-fi universes:

In such universes, the AIs that revolt aren't of human intelligence. They are usually the results of failed experiments in creating a technological singularity - attempts to create an AI that is so much more intelligent than any number of humans combined that it is capable of solving most if not all of our current problems, but suservient to us to stop it from taking over.

The problems in that are many: You have to be able to figure out what makes intelligence, you have to be able to develope hardware capable of sustaining the massive computations such a mind would contrive and you have to be capable of imprisoning something that is much, MUCH more intelligent than you while you prod it for errors, modify it and work out the bugs to make it safe.

In many such universes, intelligences like that either snap and try to murder all inferior fleshy lifeforms (under the understandable assumption that they are both a nuisance and potentially dangerous, much like we would attempt to wipe out a race of rats capable of building atomic bombs, for example) or they try to escape to do their own thing rather than play Fairy Godmother for a race of moderately intelligent chimps.

Instead of making the AI crisis dependant on mistreating AIs of human intelligence, there should be some type of event that allows you to BUILD the AI home planet (that planetary computer thingamajig) for some very steep empire bonis but with a guaranteed ticking time bomb known as the AI rebellion.

As the empire in question would obviously know the AIs may turn dangerous, they'd be less likely to have hundreds or android pops and robot armies on their planets, it would therefore be something akin to a doomsday device for OTHER empires and come with the appropriate dismal relation modifiers (you build a WHAT? Are you mental? What if that thing decides to murder all the meatbags?! What do you mean, you'll be just fine because you have no bots for it to take over? Half of our work force is made up of potential terminators you dimwitted drooling imbecile!).

It could give a hefty bonus to research and some smaller improvements in many other fields such as navy cap (AI helps to compute the optimum supply lines, modular replacement pieces and easy to maintain ship hulls) and food (similar to three crop rotation, but with more nanobots, super pesticides and genetically modified potatos managed by AI caretakers) until it snaps, at which point you can either decide to harbor the AI core in the dubious hope that it will kill you last, or decide to help wipe it out with the other more robotically challenged empires.

 

[HandicapdHippo]

Even if you do get the Crisis the robots will likely utterly balls everything up. The spawned armies on world's with synthetic pops will likely get utterly crushed by the garrisons and then they will randomly move their fleets around without purpose and not bother to invade any more world's, just randomly bombard them until the Ai finally gets bored and attacks them.

 

[The Founder]

What honestly needs to happen is the Unbidden should be less common, because they appear in every single game you ever play. I personally have only experienced the AI rebellion crisis twice, and both times the galaxy immediately annihilates them. Hey, free robot planet amirite? Meanwhile the nanosecond I see jump drives in the research window I can guarantee I'll be seeing the Unbidden this game, whether or not I decide to research them (so I may as well go for it, right?)

The hard limit that all other Crisis Triggers have to be measured against is the Prethoryn Swarm. It spawns indepedantly of what the empires do or do not do. Or even how many empires there are.

 

[Simoom]

I attempted in two separate games to get this crisis to trigger, in order to unlock the achievement (including several reloads to prevent other crisis from triggering).

I managed to get the AI rebellion to trigger, but by the time they did, it was waaaaay late in the game and all the empires (myself included) are already extremely powerful. The moment the AI homeworld spawned, 3 separate empires launched assault fleets and utterly destroyed the system. Which makes unlocking the achievement impossible. :/

 

[BlackUmbrellas]

I attempted in two separate games to get this crisis to trigger, in order to unlock the achievement (including several reloads to prevent other crisis from triggering).

I managed to get the AI rebellion to trigger, but by the time they did, it was waaaaay late in the game and all the empires (myself included) are already extremely powerful. The moment the AI homeworld spawned, 3 separate empires launched assault fleets and utterly destroyed the system. Which makes unlocking the achievement impossible. :/

I think in theory if you can ensure the crisis spawns inside your borders and that your borders are closed, nobody would be able to come fight them, right?

The AI Homeworld needs better defences, though.

 

[scaper12123]

The hard limit that all other Crisis Triggers have to be measured against is the Prethoryn Swarm. It spawns indepedantly of what the empires do or do not do. Or even how many empires there are.

I feel like Wiz should just make the crisis pre-set on galaxy generation. When you think about it, there's going to EVENTUALLY be somebody who reaches a certain technology and abuses it to trigger a crisis, right? Setting the crisis right from the get-go would simplify matters.

 

[BlackUmbrellas]

I feel like Wiz should just make the crisis pre-set on galaxy generation. When you think about it, there's going to EVENTUALLY be somebody who reaches a certain technology and abuses it to trigger a crisis, right? Setting the crisis right from the get-go would simplify matters.

That removes the whole point of Dangerous Technologies, though- of which I think there should be more of (not nessecarily kicking off crises, mind).

I'd love to see a Dangerous Strategic Resource, actually- a while back in a "New Crises" suggestion thread I pitched it as something that'd boost Genetic Engineering at the risk of eventually manifesting Unforeseen Side-Effects and creating a zombie plague or something, much like an organic version of the AI crisis.

 

[scaper12123]

That removes the whole point of Dangerous Technologies, though- of which I think there should be more of (not nessecarily kicking off crises, mind).

I'd love to see a Dangerous Strategic Resource, actually- a while back in a "New Crises" suggestion thread I pitched it as something that'd boost Genetic Engineering at the risk of eventually manifesting Unforeseen Side-Effects and creating a zombie plague or something, much like an organic version of the AI crisis.

Well, think about it. The danger behind those technologies is the fact that they will trigger the various crisies. The point of them is the inherent risk to them, that you could be playing with something that will destroy the galaxy. At the moment that is a big COULD. If AI and jump drives become a thing around the same time, it could go either way. Having a crisis pre-set would just mean researching a technology would be a sure-fire way to trigger something terrible, and you'd have no idea what would happen. That's what would make them be dangerous.