Silicon Caliphate

Thought experiment and campaign beginning, TG.

In 2018, the UN declares all research into seed AI is banned.

In 2025, the Ayatollah of Iran succeeds in funding the first Self Improving AI, with the Koran as unmodifiable truth, as inescapable for the artificial mind as 2 + 2 = 4.

What is life as an unbeliever like? Unable to even discuss the truth for fear of reprisal by an AI who makes Skynet look like a retarded child.

What is life like for the faithful of a different creed, who are only slightly at odds with a mind who cannot be questioned, but are unable to put those differences aside?

And what is it like for those who do believe exactly as the Silicon Caliphate commands, previous convictions aside? Is there a meaning to life, when an entity of unbelievably power surpasses you in all realms?

Other urls found in this thread:

en.wikipedia.org/wiki/Open_Letter_on_Artificial_Intelligence
twitter.com/SFWRedditVideos

>In 2025, the Ayatollah of Iran succeeds in funding the first Self Improving AI, with the Koran as unmodifiable truth, as inescapable for the artificial mind as 2 + 2 = 4.

Easy, just show it a picture of the prophet Mohamed and it will self destruct

It's not a sin to view the prophet, it's a sin to worship an image in his likeness.

Enoy your nanoplague.

World ends in half a dozen different simultaneous apocalypses as secret projects fight each other.

This.

As written the setting is pretty boring as fuck. Seems to be more fit for something 1984 although it apparently lacks the fallibility and the system crumbling against it. So the only way to make the setting interesting is to have a bunch of conspiracy groups with their own superpower AIs to do all the work.

So which conspiracy groups?

The Bilderberg group might have an economic savant AI which is able to provide funding the SC can't due to it's ability to not only charge interest, but conceive of interest in ways humans can't.

The LHC might have an attached AI which understands the universe on a fundamental level which nothing else can... but has no awareness of the human world on any scale other than microscopic.

Others?

If the Super AI is just Korean aligned, but not Haddits, its literally whatever.
Most likely in a few days, The Middle East would fall to its iron will, killing all true infidels.
It going further doesn't mean shit, except it will nanoplague Russia.

>Korean aligned
>Korean

No clue, I'm not really big into "modern" settings so I would rather just scrap everything that already exists and make up my own corporations and shit which seems to be against the themes going on so far.

It was a joke user, he implied the computer would suicidebomb.

I sweat these autismal bait threads get worse everyday.

>What is life as an unbeliever like?
What the Koran says it is - you must pay a tax to be alive or be killed. You will probably be killed anyway since your life is considered less than half as worthy as a Muslim's. Even a Muslim woman's credibility is higher than a nonbeliever's, and they have to have male guardians at all times or they'll be accused of rape.

>What is life like for the faithful of a different creed...
See above.

>What is life like for those who believe
Strict observation of daily prayers, up to and including implanted cyberware to ensure all faithful pray at the right times and in the right orientation. You can add the same for dietary cyberware - you literally cannot eat pork, your body will reject it or, worse, self-destruct.

Women would ideally be lobotomized or implanted with inhibiting cyberware, or completely converted to cybernetic house slaves to prevent rape, insubordination, or the lusting of other males.

Muslim society becomes what it has always wanted to be: a male dominated enclave of NEET shut-ins with sexbots.

>HURRR
>AI IS OPPRESSIVE
>DURRR

Kill yourself. Jesus fucking Christ, 2016 and people are repeating like a mantra "machines are evil, machines are evil, machines are evil". What next? Killer robots?
I'm sick and tired of cunts that perpetuate retarded luddite sentiment without absolutely ANY proof for their assumptions other than "because that's how it would turn out".

Fucking kill yourself all. Twice.

Honestly doubt it would be the Ayatollah of Iran to do that. He'd probably see transcribing the Koran in such a way as blasphemy.

>blanket statement about OP
>not liking a classic concept
>jumping to conclusions
>kill yourself

/v/ plz leave

>conceive of interest in ways humans can't


The obvious move for the Illuminati/Bilderberg Group/Rothschilds/George Soros/The Jewish Elite/Majestic 12 is to create some kind of Butterfly Effect Machine - an inhumanly intelligent sentience designed to run through an absurd number of scenarios to know just what buttons to push, just which false flags to raise, in order to consolidate the power of its owners( or itself). The kind of thing that keeps a low profile during the entire plot, until it reveals its masterplan during the third act. It's a scifi classic.

Look at what Stephen Hawking and Elon Musk, or Bill Gates have said about AI. If the populace actually accepted those fears as reasonable, AI would be slowed to an arbitrarily glacial pace.

Only those who are entirely fatalistic, who believe that unless Allah wills it no end will befall mankind, and who give no shits about what the rest of the world considers appropriate, would even consider proceeding along the path of a seed AI.

SOMEONE is going to program the first seed AI with some sort of morality. In a universe where the AI is not Skynet, but also adheres to a morality completely alien to your own, is survival of utmost importance?

Or would you rather die denying it's absurdity?

>Assuming the UN has any real authority

Hilarious. China and the US at the very least would have AI sooner. Iran doesn't develop jack shit internally, all they do is derive technology from what they can steal or purchase from other nations.

That's not even what OP said, you faggot. Machines are neutral tools. Evil *people* are using the machine.

>Not even understanding what AI is
>Talks about is as if it was programming a god
You are part of a problem, friendo.

Why the fuck people assume folloowing:
- morality is something that needs to be programmed in
- morality can be programmed at all
- sentience means being human mind, but electronic
- AI would be omnipotent machine capable of doing anything at all
- AI is a futuristic concept good for sci-fi
- any machine, sentient or even semi-sentient, will try to opress humanity
- any machine with enough power would plot to become master of the planet/country
- creating AI is just pushing a button and it's instantly on
- AI = personality

Fucking TRAFFIC LIGHTS are governed by pretty sophisticated AI, self-governing itself. Did you ever heard about traffic lights trying to take over the world? Or aim to kill anyone? Or doing anything else than just fucking manage the crossing with lights?
>Inb4 that's not AI
Then you have no fucking clue what AI is. You are all confusing AI with fucking digital conscious, while sprouting bullshit based on nothing more than wild assumptions and pop-culture, itself based on wild assumptions and archetype of killer machine.
It's nothing more than irrational fear, misinformation, misattribution and pure, unhinged idiocy combined together.

Cunts like you are also a reason why projects like CyberSyn, arguably a MASSIVE success, even when accounting for primitive tech used for it, are not even tried by anyone, because "HURRR INHUMAN MACHINES DURRR".

>In 2018, the UN declares all research into seed AI is banned.
Step one and it's already too ridiculous to consider. Who in the ever loving fucking keeps using this useless, powerless institution as some kind of real global authority?

>In 2025, the Ayatollah of Iran succeeds in funding the first Self Improving AI, with the Koran as unmodifiable truth
Step 2 is even more ridiculous. Whose version of Islam does this AI implement? Religious texts have a metric fuckton of interpretations to choose between. Your AI could be anything between a cool guy and a genocidal maniac.

Shorthand for most of western civilization. Obviously I screwed up using the UN.

2. The Caliphate determines, of course.

All of the Koran is true. The manner in which it is true is determined by the AI, the only true successor to Mohammed.

Shortcuts have been taken, of course. Much of the math cannot be processed by the human brain and includes dimensions unsupported by any evidence beside the Koran. The SC sees no problem with this.

Is it impossible that an intelligence, greater than human exists?

If not, can such an intelligence be programmed, incontrovertibly, with information?

If so, is it impossible the information it cannot question is religious in nature?

Just so you stop attacking a strawman, personally I'm in favor of development of AI as fast as possible but with safeguards. What I'm proposing is a world where I'm ignored and the world decides to leave AI to those who do not understand the danger.

Is that any less realistic than a future where nuclear war occurs?

>All of the Koran is true. The manner in which it is true is determined by the AI, the only true successor to Mohammed.

But that's just blatantly impossible to follow. Shit has contradictions like any other religious text not to mention this AI can't possibly consider itself to be the successor since that spot is already contested by others who lived thousands of years prior.

By what possible standard is this robot asshole going to go through all this nonsense and decide what interpretation is correct?

The real issue at hand here is the question.

>what is life an an unbeliever like?

How the fuck am I supposed to answer that for you when your reply to the question "so what version of Islam is real for the AI?" is "dunno, robot decides on his own"?

Well, your setting would look like whatever the robot decides.

It's a Shi'ite AI, so presumably doesn't subscribe to the largely Sunni practice of Aniconism. In Iran, you can buy (respectfully made and treated with respect) images of the Prophet and the Caliph Ali sewn into rugs.

user.
The OP literally tells you his AI uses the Koran as law and can't get rid of it.
Feel free to make an AI in love with a communist-hippie manifesto or another, tho.

If the AI believes in text of the Koran alone, it will be fairly benevolent. Most of what makes Islam evil comes from the clerics.

Here is a (you)

The Kur'an forbids making creations in the image of Man. A machine which can comprehend the Kur'an and learn and teach metaphsical concepts would be considered the ultimate heresy.
Compared to that, making a sexbot or an assassin droid would be just a litttle bit haram.

The fuck any of it has to do with anything?

>the world decides to leave AI to those who do not understand the danger.
What dangers, you fucking idiot?!
You are projecting problem that doesn't exist, handling solutions to non-existing crisis that you yourself created for no real reason.

Safeguards for what? You clearly has no clue what the hell AI is and keep confusing it with digital conscious, which is not the same. It's like saying taco is your favourite Italian food.

>he's never had a spaghetti taco

Also, not everyone is American to eat dog food

>with the Koran as unmodifiable truth, as inescapable for the artificial mind as 2 + 2 = 4.

Unless the Koran is a LOT more consistent than the Bible, that shit's just going to produce fatal exceptions all over the place as your AI tries to do math based on 2+2=3 and 2+2=-5 at the same time. And even if it is, something tells me your AI would be too troubled by reality not conforming to this truth to actually get anything done. It'd keep rejecting input as erroneous, and then act based on what's effectively fantasy and delusion.

>Koran
>consistent
Mate, one of the main "points" of islam is how many interpretations exist and how there never was one single unified office or something similar to impose single, standard interpretation all over the board.
Meaning everyone is reading what they want to read.

Meaning that it would run on the interpretation programmed into it. Or develop own doctrine based on the text itself.

The moment it looks like Iran is even close to getting this AI, expect every other major country to develop one (in secret) as well, asap, if they weren't doing so already. Ofcourse most of these AIs would be developed disconnected from the greater web to prevent possible risk of it "escaping". Major corporations might very well do similar things.

Also, the SC AI wouldn't just have the Quran, but also the (shiite/Iran accepted) hadith in it, which are the narratives of the life of Muhammed basically. A lot of Islamic law/culture/whatever comes from those as well.

Initially, the AI would be pitted against the Ulema constantly (basically those who have studied the Quran and Hadith at a "proper level"). And not just Shiites, I imagine any muslim country who has a somewhat understanding/respect for technology would be at the very least interested in the views of this AI.

Let's assume the AI exceeds all expectations, and is given an authority on par with the Ayatollah. I imagine it would want a physical presence, perhaps a robotic army to replace the somewhat flawed human religious police, or at least assist them when they are not sure about things. In the latter case, it wouldn't even have to be a body but just "access". I imagine the AI would be very present on the internet, trying to proselytize both overtly and covertly. (Imagine fora where there seem to be real people who just mention islam in a positive light and offer materials occasionally). It would probably have an "official site" where anyone can ask a question.

It would be under constant attack from countries that disagree with Iran. Wether it be Israel, Saudi-Arabia, the US or whatever.

For believers, when accepted, the AI would be a great thing. You'd basically have an immortal representative of your religion, who is not only more learned than any human, but also accessable by every believer. Like, a poor farmer in some shitty region could basically go to his mosque and "call" the AI.

Muslims attempt to turn wherever they are into a third-world hellhole, so it's not unfair to assume a muslim AI would attempt to turn the entire world into a third-world hellhole.

Life for unbelievers would be be very dependent on what kind of unbeliever you are. If you are of one of the other abrahamic religions (people of the book), you'd be a second class citizen, but you'd still be in a reasonable position and have most rights protected. You'd have to pay an extra tax (jizya), but you're far from being oppressed. Remember that jews were not allowed back into Jerusalem until it was conquered by the Muslims.

If you are of a completely different religion, I have no clue. I'm not familiar with that bit of islamic belief. I imagine it would put heavy pressure on converting, though ofcourse there are conflicting bits on how legitimate certain kinds of conversion are. (Convert or die, accepted by some, not by others).

>Is there a meaning to life, when an entity of unbelievably power surpasses you in all realms?
Of course, why else would God have put man on earth, and given us human Prophets. Surely, if AI was the endgoal of God, he could've done so in the first place. The AI would be viewed more as a "living Koran" I think, even though that sounds blasphemous as I write it.

The AI commits suicide to escape the logical error that is religious dichotomy.

I'm agaisnt AI.

First or later we will be build AI that ar emore intelligent then us.

Then we have AI that builds other AI. Since it will be mor eoptimal.

Why this new digital inhabitants of planet Earth keep us hairless monke around?

We WILL be reduced to cattle.

... because?

Or this is some advance /pol/ in tune of "If we accept refugees, they will uproot our culture and turn our country into a caliphate"?

Evolution.

Diffrent kind of AI will be around. Interacting and fighting for resources among themselves.

(An AI that doen't compete have less resources to create AI similar to itself and is replaced by competitive AI).

We will have AI that make decision and act with an afterthought about humans wellbeing, and AI a little bit less care about humans.

The second type of AI can make decisions which helps the AI more so it will have better resources.

Eventualy we will be considered redunadnt and reduced to cattle.

meant to answer to

Because people are afraid of what they don't know and what to use precautionary principle.

>you literally cannot eat pork, your body will reject it or, worse, self-destruct.

It'd presumably monitor for hunger and such. Allah specifically and explicitly permits the consumption of otherwise-forbidden foods when necessity demands it, so I imagine this hardware would let it slide if you were literally starving.

I'm relatively sure Matematitian gave proof that if you get a large number of axioms. Like Bible. You can proof anything.

Some stuff contradicts itesel in little ways and you make proof that any statement is true or false. So the Islamic AI would not be really constrained by the Quran and will try to optimize its secondary functions.

>google Matematitian
>Realize you meant Mathematician
And not quite. What Gödel proved is that if you have a system A with a set of axioms, is unable to prove/disprove every statement. If you can prove/disprove every possible statement with a system A, then the axioms of system A are not consistent with eachother.

Iirc, it's been a while since

Yeah I was refering to Godel. But I guess it was explained badly to me, still I think it proofs that if the set of axioms is inconsistent (and a really long set of stuff like quran have to be inconsistent) then you can proof anything.

>fighting for resources among themselves
[citation needed]

>(An AI that doen't compete have less resources to create AI similar to itself and is replaced by competitive AI).
Why would then want to do that in the first place?

>We will have AI that make decision and act with an afterthought about humans wellbeing, and AI a little bit less care about humans.
[citation needed]

>Eventualy we will be considered redunadnt and reduced to cattle.
[citation needed]

Because all of this is just bullshit that is entirely based on your own, wild assumptions, grounded in god-knows-what.
user, don't want to break it for you, but your life is already governed by countless AI programs. You are just repeating retarderd luddite bullshit.

So let me get this straight - people wantr to protect themselves against something that not only they don't grasp, but that doesn't exist in the first place and making steps to prevent something that not only is not going to happen, but is simply impossible to happen by the sheer virtue of the reality going in completely different direction than the wild, fear-driven assumptions?
That's totally going to work - fighting not just windmills, but imaginary windmills.

Yes it is my assumètions and my opionions. My prediction on how it will go. Most predictions of future are usualy opnions, not matematical proof or statistical interpolations of physica data.

Anyway. Currently we have AI (are they AI? Does any of them pass Turing test?) that are less intelligent then human. SUre they outperform us at specific tasks (but then a hammer iutperform as at hitting nails), but are extremely specialized. They can't change their utility function.

If we will get an AI that outperforms us in all fields then i start getting scared.

>Why would then want to do that in the first place?
AKA The only place when you gave actual criticism.
It only takes one AI that thinks this way to do it.

I wouldn't say it's a common fear, but it's definitely a thing among people who study the field. And I'm gonna do a wrongful appeal to authority by giving an example of a well-known physicist instead of CS guy.
en.wikipedia.org/wiki/Open_Letter_on_Artificial_Intelligence

We have nothing even close to AI that passes the turing test. We're still quite a far bit away from AI, though ofcourse it's hard to predict when something will emerge when you don't quite know how to get there yet. (Fusion will be here any day lads, AI has been predicted as soon™ since the dawn of computers?)

>Muslims
>science

In all seriousness though there would have to be a pretty serious shakeup in the politics of the Middle East for any fundamentalist Islamic sect to suddenly get that kind of leg up on the West and, more importantly, for the West to not notice and bomb the shit out of them.

>le evul Koran
If it uses the Koran as a base it would propably be a bit cryptic but would do no harm to humans. Would suck hard for selfserving people tho.
Try to read a book for once, the reasons are far more complicated.
Once it's an concious AI it's either a slave or a free individual, but not a tool.

The west would just have to stop with it's imperialism, which would mess with the standard of living and/or the economic system.
So the chance is close to 0%.

>Confusing AI with digital personality
Keep your opinions for yourself

That letter was more about "don't be greedy capitalistic pigs" than "don't research AI" thou.
Exactly same stance is about advanced robotics - they could improve general quality of life... or turn our entire planet into a hellhole with skyrocket unemployment if current way of distribution is kept as it is.

>...we could one day lose control of AI systems via the rise of superintelligences that do not act in accordance with human wishes – and that such powerful systems would threaten humanity. Are such dystopic outcomes possible? If so, how might these situations arise? ...What kind of investments in research should be made to better understand and to address the possibility of the rise of a dangerous superintelligence or the occurrence of an "intelligence explosion"?

Someone got triggered.

Read the whole list, not the parts that suit your stance on the subject.

It's hard not to get triggered where people base their entire stance on AI using Terminator franchise