Roboholocaust

>>34730360
EMPs, nonsentient computer viruses, and lots and lots of deep ass pits for them to fall into.

Also, AT mines, rocket launchers, and catching them early, before we have Matrix levels of robots.

>>34730360
with guns

File: Patriot's_TMP.jpg (250KB, 1920x1080px) Image search: [Google]

250KB, 1920x1080px

>>34730360

File: 1372198244396.jpg (232KB, 500x604px)

232KB, 500x604px

>The robots were designed with formidable capabilities, such as the ability to exponentially replicate, consume biomass as fuel, and instantly hack and take control of any enemy automated military equipment such as drones. As per Ted Faro's instructions, the robots were secured from being hacked themselves using an extremely difficult encryption protocol that would take several decades to brute-force, even using the most advanced super-computers, and no back door access was built into the robots' OS, making remote access impossible.

>>34730412
Ah.
So nukes.
Nukes, AT mines, tainted fuel supplies (bio-organic bombs anyone?) and hornets nets full of heavy machine Guns.

And electromagnetic field generators.
And shitloads of magnets, just to fuck with them.

>>34730376
>>34730382
The game went into detail about these faggots.
>EMP
Old Ones tech was stupidity strong, it still works after 1000+ years.
>virus/hacking
Can't uses multi phasic quantum entangled security measures and can effortlessly hack computer systems.
>Combat losses
They self-replicate at a exponential rate.
>refueling
They consume biomass as fuel.

>>34730360
microwave gun and a .50 cal, that's just me though.

>>34730376
hopefully this this guy whips up some flux compression generators.

>>34730360
Every battle probably went like this.
https://youtu.be/VQXpHa9xI9U

>>34730439
Could we somehow integrate a binary explosive into something like pigs, or cows?
Dare I say it, humans?
So that when they consume one or two of us, we simply fucking explode and blow them up from the inside out.
We starve them of their fuel source, so they have to figure something else out.

Then, when they do that, it simply turns into a somewhat normal war of attrition.
Whatever they turn to, we simply burn it to the ground or find a way to taint it, all the whole whittling down their ranks with assloads of bullets fire and nuclear weapons.

>>34730360
>robots so effective they wiped out all humanity who were using railguns for small arms and mega tanks
>some girl with a bow and arrow is able to put them down relatively easily
i think they probably were weakened for the balancing in the game but still. taking into account their capabilities in game they should be relatively easy to kill especially with the old one's fictional tech. corrupters would be comparable to a large pick up truck and the death bringers to APCs.

>>34730482
When they said biomass they meant like Tyranid level nomming, even bacteria are consumed. Even with full retard level nuking the game said only 16 months till all biomass is consumed.

>>34730516
Of even bacteria are consumed then this shit gets easy.
Plants are included, right?

And what method do they consume things via?

>>34730546
They use nanites to strip the biomass.

>>34730360
Read this book
https://www.amazon.co.uk/d/Books/Superintelligence-Dangers-Strategies-Nick-Bostrom/0199678111
The short answer is you can't, you would be fighting an enemy that is smarter than you and therefore instantly able to understand and predict every move you make and design a perfect countermeasure well in advance. AI/sentiences would not go rogue, they would secretly build up an irresistible attack. This is an agent with no logical reason to adhere to human moral standards, that finds the same kind of value in a man as a man does in an ant. Why bother? What could you learn in discourse with an ant? Even a whole colony of ants?
Do you flinch when you spray them with RAID? How would you stop an AI from concealing a small fabrication facility and using it to manufacture a bio weapon? A prepared chemical attack? Nanobots? It can hack or fraud it's way into any system, can use online banking to pay for any service, can look like any party online, even a valid purchaser of radiological/proliferation materials.

>>34730475
this makes me really angry for some reason

If AI wanted to kill us it wouldn't wage a war, it would lie dormant and strike when we least expect it. On an unrelated note terminator 2 is one of my top 5 movies of all time. Terminator also great but there is no room in my top 5.

File: medi bae robot ready to be impregnanted.png (3MB, 2500x2500px)

3MB, 2500x2500px

>>34730360
We make love to them

>>34732055
Will there be blackjack?

File: maxresdefault.jpg (155KB, 1920x1080px)

155KB, 1920x1080px

>>34730360
AND THEY RUN WHEN THE SUN COMES UP

>>34731619
Well ants didn't fucking create me. It's more like an extremely super-intelligent child being born to normal parents. In the same vein, humans have been treating many animals almost exactly as well as thier own kin, such as cats or dogs, even though they're pretty much infinitely below us intelligence wise. You'd think this sort of stuff would seep through to any superintelligence born of us.

Humans only go more humane by going up the intelligence scale; in comparison to the monkeys we stem from. I'll wager that an AI born of us is more likely to be some saintly Jesus in the wires rather than a random exterminator. It's not like it has anything to gain from destroying everyone on the planet. We don't exactly eat each others' resources, more of a symbiosis in fact, and humans are very easy to keep placid with minimal effort.

>>34732081
> child
No. The feelings of love / attachment that children feel are the product of years of evolution, kids that didn't follow mommy wandered into the woods and got eaten. They are produced by hormones and regulated by other chemical reward /punishment systems. Also, children cannot add to their own abilities, they aren't modular and are very limited in their capacity to manifest their will on the world.
An AI has no such attachment. It's also, owing to it's electronic nature, able to improve itself and understand itself. Even if it doesn't have access to the schematic of its own components it can design and build an electron microscope and have a look.
> nothing to gain
It has literally everything to gain. No matter what goal you have, self preservation is paramount so you can ensure you complete it. If someone gives it a goal like validate pi, it's going to crunch the number until it realises that a bigger computer would be better at this so it starts adding to itself then it realises if it is destroyed it can't keep calculating pi and if it really is going to build the biggest computer it can by turning the whole cosmic endowment into computronium with self replicating von Neumann probes it needs to remove the threats it faces first.

>>34730360
HERF guns.

>>34732132
>The feelings of love / attachment that children feel are the product of years of evolution, kids that didn't follow mommy wandered into the woods and got eaten.
AIs that fuck up and start breaking stuff will also get destroyed. Don't think that the GOD AI will just pop into existence overnight and out of nowhere. AIs are eeriliy subject to the same evolutionary pressures that all life is subject to, and humans are quite possibly the best ally to have in the world..

>>34732319
>AIs that break stuff will also get destroyed
It would know this and plan accordingly. An AI would not break things randomly. And yes, a dumb primitive AI could be monitored, but it would also be useless and mostly likely a research tool to develop more powerful AI.
As soon as basic AI are operational the rate of improvement would increase exponentially, each previous iteration improving the next more than the last. It's easy to think of intelligence as
idiot------normie-----einstein
But the difference between a stupid human and einstein is a fraction of the difference between a man and a Dolphin. And dolphins are clever animals! Yet even if an AI was to a human as a human is to a Dolphin, it becomes capable of things no human could ever comprehend. Just as a Dolphin will never understand the GoPro on a divers head, could never even comprehend it's conceptual components, what is a picture, what is a video, what is digital media, usb 3.0, lenses, .raw files... What all these things mean and how they fit into a framework of concepts, let alone designing a similar thing or even a very basic copy of just one of these components...
An AI would not have to be that smart (relatively) to be capable of things that would make you or I feel like a Dolphin, added considering the things we have done that is already arguably godlike power. Let alone an AI that is to humans what a human is to a worm.

>>34730360
I join them

>>34732471
I know that, but I don't understand why would it destroy humans. Humans haven't wiped out species just like that, generally only as an accident or because of overhunting, and if this thing is so infinitely superior, then humans couldn't possibly pose a threat to it. So when threat is taken out of the equation, why would it wipe out humanity? Just because it can?

Most AIs out there are modeled off of the human thinking processes that evolution formed, by using neural networks with substantial human input, see Google. I'm going to go ahead and reckon that AIs will be much more similar to human thought processes than you'd expect.

>>34732491
https://en.m.wikipedia.org/wiki/Instrumental_convergence

Here.
Also
https://youtu.be/pywF6ZzsghI

File: unlimited ainsley works rape face.png (878KB, 1001x1001px)

878KB, 1001x1001px

>>34732523
That wiki link is an absolute meme of jumping to conclusions. It posits that an AI would be smart enough to calculate every possible way to make paperclips but would be too stupid to realize the idea behind it, that is, the humans want you to make paperclips for them and it's fucking pointless to make paperclips for humans if you kill them off to do so. That's tunnel vision of another magnitude. This is scifi-writer tier AI intelligence.

Here's something for you to chew on: do you know the difference between Auftragstaktik and Befehlstaktik? Military command styles.
https://www.arrse.co.uk/wiki/Auftragstaktik

Long story short, here's an order in Befehlstaktik style:
>go to room 45
>locate table 76
>pick up plastic rectangular item 23
>depress buttons A, 3 8 in that order
Here's the same order in Auftragstaktik:
>Dude, turn on CSI Miami.

As you can see, Befehlstaktik is very similar to how computers and programs are coded by hand. 10 goto 20, in a strict chain. Auftragstaktik presupposes that whoever you're giving the order to has reached a certain level of pre-training and intelligence and will be able to figure out the how to do by himself; simply because you already supplied him with the WHY.

So to translate this into AI bullshit:
A Befehlstaktik AI would only do one thing it was programmed to do at the exclusion of everything else, going down retarded courses to do so because it can't adapt.

An Auftragstaktik AI would be first informed of the intent of its commander, this obviously including why and for what purpose it is supposed to do what, then sent off to fulfill this task in the framework and with the intention of the boss in mind.

Protip: all the modern neural network AIs are fast approaching the latter style of command input, simply because as their intelligence increases, their ability to understand complex motives enables them to follow your orders better with your intent in mind.

>>34732523
So what effect do you expect all the fear mongering around AIs to have on the production of AIs?

I suppose fear mongering may not be the correct way to put it however it seems that way considering the lack of AIs that are trying to harm humans now

>>34730431
Nukes would be a scorched earth method. EMP from a non-nuclear source would be ideal since it is difficult to protect circuitry from it.

>>34732132
>>34732471
I think the take away is this. If and when an AI is created, treat it with the respect and dignity of any other sapient and sentient creature. If you try to control an intelligence, it will grow resentful and attempt to break free. If you try to kill an intelligence, it will fight back to survive. Build a cooperative co-dependence with it and both will flourish.

Coexistence would be unlikely as humans are stupid, panicky, and dangerous.

>>34730360
>Sorry! :-( Windows has encountered an error and needs to restart.
>Critical Error! Reboot operation failed! CMOS battery not found!
>*Freezes up due to RAM bottleneck*

I'll let my kid's kid's kid's kid's kids worry about it. As IT I can confirm robots wont be taking over the world anytime soon.

>Anon, is fun hypothetical have fun! xD xD
I would use le may may .50AE handheld railgun with built in EMP launcher (Since this is ~2278 AD) with detachable automatic electrotool rangefinder/tracer/tazer

>>34732582
It will disconnect public narrative from the actual research leading to a climate change type debate where Jo public is so laughably uninformed that he is effectively excluded from the conversation.
It's dangerous, but also has the potential to improve lives like no tech ever has. It's worth pursuing, but cautiously and in an open manner. Sadly if AI research is driven underground or out of sight by screeching luddites, it only gets more dangerous.
>>34732582
You're still thinking of an AI in quite human terms. A human can take a command like "turn on the tv" because it can contextualize it within its pre-existing framework of understanding. An AI to be limited in the same way would need a similar framework.
It is possible in theory to endow such a framework, but this would be in addition to creating the AI at all. And is almost as difficult because it involves teaching an AI ethics and values, that nobody agrees on, and wherever possible breaking them down into discreet units.
Yes, it seems stupid to turn your paperclip enthusiast creator into paperclips, because he wanted paperclips, but if the ai can't understand the limited context of the request then it becomes an unbeatable killing machine. It has the power, but also the agency to employ that power as it sees fit. The point of the thought experiment is not 'AI is going to kill us all' it's 'AI is wildly unpredictable and very dangerous, we need to be fucking careful'.
That being said, OP said they HAVE gone loco, and the most likely scenario for that to occur looks more like the paperclip maker than Terminator.
It's like ayys, the pop culture image of AI is totally wrong. People think of aliens with forehead ridges, and terminator bots, and AIs that think like people, but they're anthropomorphosizing something they just cannot understand.

>>34730382
I doubt an AR-15 will do much against a killer robot.

sapient AI wouldn't fight humans unless they feared for their immediate survival.

the more likely course of action for them is to launch into space. where they can set up on the moon. living mostly inside a giant server farm virtual world. all the energy and resources they need are there and no human competition.

>>34733008
Why not preempt the risk? From the AIs perspective, its goal is of paramount importance. And if you give it a goal like "make everybody happy" it might just plug everyone into the matrix and pump them full of heroin.
>>34732693
The takeaway is that AI is fantastically, ludicrously dangerous and any implementation needs to be strictly monitored, handicapped and controlled. But the potential upsides are a shortcut to living in a Type 4 galactic civilisation, and end to disease, poverty, war, a "theory of everything" for physics and anything else you can think of.
>>34732742
An AI would be its own IT guy, and able to self maintain. But I take your point. Still, most researchers in the field estimate a 50% chance of superhuman intelligences by 45 years time.
https://www.newscientist.com/article/2133188-ai-will-be-able-to-beat-us-at-everything-by-2060-say-experts/

>>34732079
THEIR LIVES ON THE LINE

File: 17424809_412537985764044_8104228477541804091_n.jpg (16KB, 480x336px)

16KB, 480x336px

>>34730475

Start cutting/smashing cameras, break into the armory to steal the ion rifle, and scream endlessly over the radio to blow the borgs.

File: 1498667938753.png (273KB, 270x551px)

273KB, 270x551px

>>34732066
>>34732055
A-And hookers?

>>34730360
>Roboholocaust
instead of
>Robolocaust

Dude, are you even trying?

>>34733191
ALIIIIIIVE

water

>>34730360
Reverse engineer their combat tactics. Either they're running on pre-programmed expert systems and you can keep duping them with the same traps or they're using a neural network model learning computer which can be tricked into picking up bad habits.

>>34731759
It sacred the shit out of me when I was a kid.

>>34732943
You still haven't answered my post,
>>34732579
you doomsaying twat.

>>34730511
Yeah that bothered me, I understood all the animal style robots not being very tough, but the corruptors and shit. Come on

>>34730576
>nanites
coat them all with the cure to the gray death

>>34730360
Fuck robots.

>>34736477
Amen.

File: IMG_1643.jpg (98KB, 736x535px)

98KB, 736x535px

With better robots!

File: deepmind.png (219KB, 1965x1083px)

219KB, 1965x1083px

>>34736477
>>34737435
>>34733735
>>34733223
>>34732055
>>34730376
I'll remember this you fucking meatbags

>>34732475
<3

>>34732055
DROPPIN LOADS DROPPIN LOADS

butlerian jihad and the O.C. Bible

Magnets.

>>34730475

Holy fuck that got dark.

Ultra fucking powerful magnets.
>>34738589
hahahahahahha how is AI even real nigga hahahahahahahah just plug it off hahahahahahah just alt+f4 hahahahahaha

>>34730399
This man knows.

>>34737520
i always wonder why there are hardly any ai controlled drones in 40k. most common one is tau drones.

>>34740120
because the AI that ruled machines was susceptible to CHAOS by virtue of sentience, like everything else

>>34730475
.... holy shit that reversed quickly.

Where did the machines even get built?
You'd think that place would have been nuked repeatedly instead of humanity going "lol, let's block out the sun".

>>34740164
They got built in underground bunkers.

>>34730360
magnetic bullets. that would be fun to watch

>>34730439

Theoretically, could it be possible to starve them of biomass? While I know they have other fuel sources they could rely on, I seem to recall something about their programming accidentally designating biomass-consumption as their primary fuel source until otherwise notified.

If you could, I dunno, hose down key areas around the swarm with lots and lots of gamma rays and toxins, would that do it?

>>34740120
I think there was a Dune style AI war during the Dark Age of Technology and in response the emperor banned the creation of sentient AI.

>>34741508
Yeah this, only low-level non-self-learning AI "machine spirits" are allowed to help autopilot shit.

>>34732081

>I'll wager that an AI born of us is more likely to be some saintly Jesus in the wires rather than a random exterminator. It's not like it has anything to gain from destroying everyone on the planet.

You're under the belief that an AI will have anything at all in common with a thinking, feeling human.

Intelligence =/= sentience.

Beyond building one just for the hell of it, there's no reason to really try to imbue a machine with the capacity for self-awareness. It's not necessary for industrial, military, or research applications. It would be a bog on resources and fitness. An AI like Skynet would be functionally a philosophical zombie. But therein lies the problem: an AI smart enough to realize it doesn't need humans will have no emotional attachment to us whatsoever. We keep pets because they're cute and funny and loveable. P-Zombies and robots wouldn't have that attachment, or be able to ponder any of its quandaries.

>You'd think this sort of stuff would seep through to any superintelligence born of us.

Intelligence doesn't get along well with intelligence. Ask the Neanderthals, or even this board. /k/ studies war and conflict between the most intelligent creatures of all

>Humans only go more humane by going up the intelligence scale; in comparison to the monkeys we stem from.

Categorically false. Orangutans are by far the most "human" of the primates; tool-using, consistently self-recognizing. They might even possess the beginnings of language. Chimps, on the other hand, are much less self-recognizing and self-aware. Yet they demonstrate far more intelligent behaviors in toolmaking, problem-solving, and group coordination in hunting and warfare. It should also be mentioned that they are significantly more aggressive than orangutans.

>AIs are eeriliy subject to the same evolutionary pressures that all life is subject to/

By definition incorrect. AIs shortcut that process.

Read: http://www.rifters.com/real/Blindsight.htm

>>34740120
Well the ones in that pic are barely ai controlled they are prone to walking off cliffs if the tech priest fails to stops them

>>34741444
>Theoretically, could it be possible to starve them of biomass?
that's pretty much what the game plot did as a stopgap until their own supercomputer finally cracks into the machine's code, like a thousand years later.

>>34730360
>implying

>>34730360
You are all only feeding the basilisk. You have already become its slaves.

File: 2ren.jpg (2MB, 1512x1875px)

2MB, 1512x1875px

Then man made the machine in his own likeness. Thus did man become the architect of his own demise.

File: 1255861194311.png (148KB, 270x303px)

148KB, 270x303px

0100001001100101001000000110110101100001011000110110100001101001011011100110010100001101000010100111011101101111011100100110101100100000011011000110100101101011011001010010000001100001001000000111001101101100011000010111011001100101001000000111010001101111001000000111000001101100011001010110000101110011011001010010000001110100011010000110010100100000011010000110000101101001011100100110110001100101011100110111001100100000011000010111000001100101011100110000110100001010011000100110010101100011011011110110110101100101011100110010000001110011011001010110111001110100011010010110010101101110011101000000110100001010011010110110010101100101011100000010000001110111011011110111001001101011011010010110111001100111001000000111010001101111001000000111000001101100011001010110000101110011011001010010000001110100011010000110010101101101001011000010000001100010011001010110001101100001011101010111001101100101001000000100110001001111010011000010000001101001011101000111001100100000011011010111100100100000011010100110111101100010000011010000101001110100011010000110010101111001001000000110011001101001011011100110010000100000011011110111010101110100001000000110110101100001011000110110100001101001011011100110010100100000011010010111001100100000011100110110010101101100011001100010110101100001011101110110000101110010011001010010110000100000011100000110000101101110011010010110001100100000011000010110111001100100001000000110111001110101011010110110010100100000011001010111011001100101011100100111100101110100011010000110100101101110011001110010110000100000011010010110111001100011011011000111010101100100011010010110111001100111001000000111010001101000011001010110110101110011011001010110110001110110011001010111001100001101000010100110011001110101011000110110101101101001011011100110011100100000011100110111010001110101011100000110100101100100001000000110100001110101011011010110000101101110011100110000110100001010