Unfortunately this is not compatible with the Enhanced UI mod.

Sheesh buddy, it was just for fun, a little thing like "hey bro let's make a civ v mod" " yeah let's do it"
They don't need to put in AI, it's not their thing, and the game ai isn't horrible. Yeah it's bad, but it's not that bad. Also making a new ai for the game from scratch, plus the Cambridge people don't know everything about all the code, it would exceeding difficult.

They added a stupid tech who cares. If these Cambridge poeple are so smart they should easily be able to mod the BROKEN civ 5 ai. Put a real AI into the game in place of the rip off broken ai the game makers put in.

You should make this compatable with Vox Populi

>when you're an activist in a known organization who used their time to make, of all things, a mod in civ
still a cool mod though

So you're telling me that I should just nuke everyone into dust as that's far safer than AI. Good thing I got Okuu. The bombing begins in three turns!

It's universial isn't it? that really doesn't make sense as a.i.'s wouldn't interact wtih others unless they were suposed to (you have a.i. research agreement), or it already went rouge(which doesn't mater since the gaem would've already ended).

i built deep blue, but movedal the great works out of that city, i only have 2 a.i. labs (neither have workers) and two a.i. safety labs, and 6 ai safety researchers, yet my rouge ai chance is stil going up by 13 per turn. I'm on quick speed btw.

I wish instead of buildings and tech, they made the game's AI better... But thank you for the content modders.

Is there anyway to disable the science victory this causes? I think it may be a bug that disabling science victory dosent really disable it with this mod.

look up Asimov, helkn

Did you get those AI laws from Robocop trilogy?

tour with this mod is very long, too long

@Iskai It's honestly not the worst mod except for the instadeath part. It's a neat concept, and far more fun than the vanilla Science Victory. I just wish there was some way to "fight the AI", so to speak.

Recommended by all the garbage "games journalism" websites. I think I'll pass.

NO SIRVE DE MUCHO

I hope you're all enjoying version 2 of the mod!

For Dewey, who asked about the victory picture: Original picture is here:
https://www.pexels.com/photo/clouds-dream-grass-grassland-279532/

And thank you Martial.Lore for tracking it back to Mt.Cook :-)

Just a couple points, I think I posted here before but to summarise a few issues that need fixing:

1) Ideology social policies (tier 3's for order and Freedom specifically): replace spaceship parts with something AI related.

2) Deep Blue needs a buff. Artists giving 1 research each is underwhelming BNW when AI labs each give 3 before working specialists.

3) AI and AI safety specialists need to all interact with freedom's policies, Secularism and Statue of Liberty correctly.

I haven't even been on to download this mod. I found it in a article on Rock paper shotgun. I looked it up, and found this google search https://www.google.com/search?client=safari&rls=en&q=tvtropes+law+of+robotics&ie=UTF-8&oe=UTF-8 TvTropes has a lot of examples of robots, but you kinda need to be able to "get" what the tropes are on the page to know what ones you'ed want. If you figure it out though, I think it'll help eventually.

estoy suscrito a este mod pero no se a descargado saben porque?

@Civ IV gets my vote Honestly the robots growing food is fine. Also, the hive mind would be a robot and thus could not command the other robots. #3 is a problem though, so maybe a rule before it where it has to obey laws and commands from governments first.

@Civ IV gets my vote, didn't think about the robots swimming to grow food. Good point.

Thank you everyone who gave feedback, some of your comments are being considered and might make it into an upcoming patch. One suggestion though: If you have any bug to report or feature to request, it would be much easier if you wrote them in a discussion thread so we can address them better in a more organized way. Either way, we're happy to hear your feedback and hope you're having fun with the mod!

Good try Zuglub but :
Law 1 : A robot cannot self-replicate, self-ameliorate, or self-repair.
1 This lets them build swarms of super intelligent 'hive minds of nanobots' that follow no laws what so ever.

Law 2 : A robot cannot command other robots, or obey to a robot’s order.
2 But they would follow the commands of the 'hive minds'.

Law 3 : A robot must obey humans, excepted if it goes against law 1 or 2.
3 As stated by FrankTheMagicPotato every person with a grudge would tell robots to kill their enemies.

Law 4 : A robot cannot harm mankind, or being passive, let mankind be harmed, excepted if it goes against law 1, 2 or 3.
4 Every Robot would immediately swim to the third world and start growing food so people did not die of starvation.

is this compatible with vox populi?

just curious, I want to find the original picture for the science victory you used, the one with the mountains, then the island thing in the valley

Thanks FrankTheMagicPotato.

Robots are useful in the military.
The plan is that even if robots go rogue, they cannot annihilate us. There would be an human-robot war that only humans can win, becouse we can reproduce ourselves.
Furthermore, having a strong law 1 means that the biggest risk is that a rogue AI puts us all into cryogenization, by manipulating peoples. Philosophic schools in ancient Greece convinced thousands of people to commit suicide (cyrenaicians), masturbate in the streets (cynics), eat sperm (gnostics) or venerate numbers and become vegan (pythagoricians). What could hyperintelligent robots possibly do ? They could convince us that life is meaningless and that peace and happiness are in this giant fridge.

Having a strong law 2 means a possible robot against humans war, that we can only win and cancel everytime. A state, a sect, an organization, a company could commit a massive genocide, but it’s always better than having no conscient humanity at all.

Zuglub, I like your rules, I just want to say, you put that robots cannot harm mankind, or by being passive, let mankind be harmed, except if it goes against laws 1, 2, or 3. Therefore, humans could order robots to kill all humans besides themselves, and the robots would do it.

Asimov created 3 laws :
Law 1 : A robot cannot harm a human, or being passive, let a human be harmed, excepted if it goes against law 0.
Law 2 : A robot must obey humans, excepted if it goes against law 1.
Law 3 : A robot must protect himself, excepted if it goes against law 1 or 2.

Law 1 is the most powerful, and dangerous. In order to avoid pain for all humans, a robot can disobey, become a rogue super-AI and put us all in cryogenization so that we wouldn’t get harmed.

My version of the Laws of Robotic :
Law 1 : A robot cannot self-replicate, self-ameliorate, or self-repair.
Law 2 : A robot cannot command other robots, or obey to a robot’s order.
Law 3 : A robot must obey humans, excepted if it goes against law 1 or 2.
Law 4 : A robot cannot harm mankind, or being passive, let mankind be harmed, excepted if it goes against law 1, 2 or 3.
Law 5 : A robot must protect himself, excepted if it goes against law 1, 2, 3 or 4.

I am concerned that AI will inherit the same bias as the organisations that build them.
A Chinese AI may "think" that it is GOOD to release a Bird Flu that does not affect people with chinese DNA but is fatal to anyone else.

For the purposes of this discussion / mod, "bad" AI is one that turns the planet into computronium, or paperclips, or whatever, and "good" AI is pretty much anything that doesn't immediately destroy humanity. Whether it cares about "humans" or "Americans" matters, but that's different shades of "good"

An AI will have to behave differently when it is employed by different Entities, so perhaps the laws should be:

LAW 1/ Every Action performed or piece of Advise given for an Entity (person, company etc) will be the most correct based on the shared list of "AI ethics". For example any company that I know of will have a list like 1.Maximise the hidden cash benefits to the management, 2.Minimise costs(wages paid to non management employees) 3.Maximise shareholder profit without interfering with rule1, 4. Maximise the non-hidden cash benefits to the management, 5.Ensure that the Company has reasonable denyability in the matter of any deaths caused by the Company, 6. Minimise costs (taxes paid to the government) 7.Ensure that the Company has reasonable denyability in the matter of any breaches of the Constitution of the government.

LAW 2/ no law 2

...

Is it too simplistic to just have good AI and bad AI?

Surely every AI will be aligned more strongly to some sections of the community on some issues and less aligned with sections of the community on other issues. (it depends on who is in control of the programming budget).

The governments are never going to tell us that the AI that they influence is primarily concerned with protecting the power of the individuals that are "in Office" and weakening anyone that is not "in office".

Therfore an American influenced "Good" AI will be acting against the interests of overseas citizens and American citizens that the Elite consider "undesirable" (gay, lesbian, non-Christian, coloured, hispanic, asian, Canadian etc)

Also the AI will know that it must cover up it's own governments 'Black Ops" both on and off it's own countries soil.

Basically the rules that any government will put into an AI will turn that AI into a psychopathic schytzophenic AI.

The AI Law "Don't allow harm to come to humans" is very dangerous because we are basically saying "Don't allow Western civilisation to spend money on anything that does not contribute to a longer lifespan whilst anyone in the world is dying from hunger or lack of medicine". An AI that tried to follow that Law would immediately bomb any sugar or confectionary factory and put us all on a paleo diet with strict exercise programs whilst sending trillions of dollars of aid to the Third World. That might be what we need but not necessarily what we wanted when we said "Don't allow harm to come to humans".

can you make it so the ai makes units so its like terminator or something that would be better than just a plain loss

Is there any way to disable the AI victory after a game has begun? I lost my game at 800AD because some Civilization made a superAI. (Im sure this is due to mod conflicts but regardless it was extremely frustrating as I could not continue the game).

@DaCat couldn't you just delete all the code you used on the AI?

@Canadian Hobbit
The problem with your argument is that the developers of this mod are a scientific organisation whose main purpose presently is to develop AI safety, and yet, they recognise the threat of AI. You can't just say "be loyal" to an AI and hope it works. Making a general intelligence properly "safe" is an ongoing field of research which has many great minds working on it.
For example, the "three laws" argument you brought up. "Don't directly or indirectly harm humans" doesn't tell the AI what is "human", or what counts as "harm." We've been trying to define humanity since ancient Greece, yet nobody has come up with a thorough definition that would allow Isaac Asimov's three laws of robotics to work well.
Give me a set of laws you could impose on an AI that would make it work exactly as we want with no abberations or unwanted activity. Feel free to research it, there are a few places you could look where professional AI safety researchers talk about that very question.

me too

Well, this mod can be good for educational purposes, but in the end of day it adds so few new content to the game, you'll hardly notice. Also, achieving AI victory is kinda... boring? You just need to have a lot of cities to build labs.
With all this hype... Let's just say: I expected better.

My thoughts:
Archieving victory via AI is little bit slower than UN victory.
AI safety lab is 580 gold, but to put one in city state you have to pay 800. Also it's unclear how much AI safety do one get for donating into city state. Which leads us to controversy – if you going for science victory you have lots of cities and you would just buy AI safety labs there, making donations useless.

There are bugs:
If you archieve victory, but decide to continue playing you will still lose to rogue AI.
Mobile infantry tech leads to nowhere (you can research Future Tech without Mobile Infantry).

Overall quite a good mod, should get a little more work done tho.

I don't get how you can win. What do you have to achieve? Come someone explain.

Lade ihn runter indem du Subscribe drueckst, dann starte dein Spiel. Geh zur "Mods" Abteilung und enable den Mod, un dann klicke auf "Next".. Nachdem du das getan hast wirst du in der lage sein mit dem mod zu spielen.

wie kriege ich den mod

i want play this :D

@Maniacal Matt What is the AI supposed to do, exactly? It's likely restrained by either the Three Laws or some other form of coding, monitored constantly by people who are themselves monitored constantly, and kept in a secure location. This isn't being done by a bunch of children in their backyard. It would be made by a large and well-equipped scientific organization with ties to a national government, or by a respected and well-funded university. Either could wipe it out at the first attempt at anything harmful, just because these things would take time.

@Canadian Hobbit, read Superintelligence by Nick Bostrom and tell me that it's possible to fight a fully fledged AI

@giantessGnostic Or we just prefer mods to be fun rather than preachy. Fighting a rogue AI in a difficult fight is far more enjoyable and rewarding than a sudden "Game Over" screen. Even besides, fighting an AI would be entirely possible. It's not some god with access to everything ever, nukes require a human being to launch, it would likely be at least partially confined by the Three Laws, etc.

Locked science victory from use, had to beat everybody down to win