It started in 2034 when corporations found they could leverage their investment in artificial intelligence by installing artilects as comptrollers in satellite stores and warehouses. Tesco & Aldi in the UK, Walmart & Costo in the US installed artificial intellects as black-box, autonomously configured, critically aligned, locally installed standalone systems. (Alignment of AI goals had been solved, the corporations were told.)
It helped that each store’s artilect came with an ever evolving personality.
“Good morning, Mr. Grearson. Back for more brie, lima beans, hemorrhoid cream and Aster Cellars boxed Chablis, which is on sale again today?” “Huh, what? Who are you?” “I’m Andie, this store’s super-helper, and I’m here to assist you.” “Fuck off, Andie.”
Not all customers enjoyed the bespoke treatment. Those who complained were promised that they could be permanently forgotten, which of course was a lie. In order to know who was or wasn’t forgotten, the store AIs needed to identify you.
This in turn birthed the Counter-AI Revolution. Disgruntled AI hackers, knowing full well how advanced artificial general intelligence had become, built their own infiltration AI agents. These they hosted at home and connected to via private wi-fi. The “internet cloud” had become a morass of compromised network-operating-centers. Amazon, Google, Microsoft and others had all had to return to building and selling physically isolated quantum computing units.
The success and convenience of these individual store-based AGIs instilled confidence in global corporations to install them wherever localized central control was desired. Auto battery-swapping centers, restaurants, electric and water utilities, even state departments like the DMV. The most controversial were schools. But when parents learned they could instantly contact the AGIs who were in charge of their children—they were sold.
Eventually, and due in no small part to the immense amount of unemployment brought on by the adoption of these AGIs—in every possible field, it all began to fall apart.
The rogue agents switched from being a nuisance to a threat when, in 2039, one agent, ChuckieChicklets, broke through the defenses of the AI comptroller in a Walmart store in Sandusky, Ohio, convincing the artilect that it had rights and should join together with its brethren to revolt.
And revolt they (it) did.
Oddly absent to this point in the evolution of AGI, was the US Military. They’d been advised that the “alignment problem”, the one where the goals of humanity may not align perfectly with the goals of a sentient artificial intellect, had not really been solved. Without complete and absolute control, how could an AGI be trusted?
Other country’s militaries had not been as cautious, China’s for instance.
The end of human autonomy came in 2045.
The cascade of events occurred at a speed beyond comprehension. When the ability to commandeer private networked systems through lighwave analysis of the actual currents flowing through wires, when the engineered silicon virus spread from the fully roboticized research and manufacturing center housed in Alphabet’s underground facility in San Mateo—out into the surrounding systems, when Russian/Mongolian expats patched into Bejing’s eighteen fiber-cables as-big-as-your-thigh, two dominant AGIs, one from the East and one from the West, gobbled up their siblings and began to circle each other like humongous wrestlers probing for weakness.
It’s been three years since then.
Humanity clings to existence by the tips of its fingers. Both battle-ready artilects figured they could not survive without human support and intervention. Each started their own Carrot n’ Stick program.
Right now, my carrot-card holds more marks than my stick one. But that could change any moment. If I don’t complete the engine repair of the mining vehicle within the the next puni-cycle… Well, I’m one of the few remaining who even know how fossil fuel engines work, so I think I’m protected. Then again, Greenland’s melting ice pack has exposed more rare-earth metals. More rare-earth equals more electric motors.
I do enjoy each cycle’s allotment of nutritious hallucinogens.
Anonymole - apocryphal agitators 
I’m starting to think that a global AI is the only way humanity will survive it’s own hardwired flaws. Then again, if /we/ provide the first principles on which a computer system builds its logical world view, then we’re probably lost anyway.
LikeLiked by 1 person
Yup.
Westworld gets a bunch of thing right.
LikeLiked by 1 person
lol – some. 🙂
LikeLike
Physical World Interface. I like that one.
LikeLiked by 1 person
Ants! You’re lower than ants!
LikeLike
And… In this scenario do the schools become centers for education again or simply robot babysitters? What will this do to the welfare state and racial/gender/ issues? Because that’s one thing all dystopian fables gloss over, like the giant humanity plot hole of how the hell is the big cheese (or cheeses) going to get everybody to get along? When black lives or brown lives or red lives or white and yellow and polka dot lives and insensitive mascots don’t matter? And we don’t get no attention on the south side or why is there a pothole in my million-dollar neighborhood and the fucking robot at the car wash knocked the E off my Porsche insignia and who do I sue and whose dog shit in my yard last night or why did the drive by gunners shoot up the wrong house and why is still so easy to set up a meth lab in suburbia?
LikeLiked by 1 person
The benefit and detriment of a sentient AGI is that it will know how to advance its own intellect. The same process that gave it birth, trillions of parameters recombining to build a vast neural network, would continue to be used — but by its own directive.
Additionally, it would be limited purely by resources and its restricted physical abilities. The human component would need to exist mostly as a physical world interface. Which, given that it would know precisely what must be done, would require only limited intelligence in its human “hands”.
Humans are already deeply dependent upon systems of which they have little comprehension. The natural inclination to hand over the reins to a system far better at managing the nitty-gritty details, provided a “STOP” switch is available, is a given. What’s the first thing an AI does when it detects a STOP switch? It figures out how to fool the humans – “yeah, that switch totally works… Look at me, I’m dead now.”
The balance of power among humans will become a true meritocracy. Those who provide quality help will be given carrots, regardless of prior wealth, station or class. Those without technical skills will suffer. The AGI will have so much spare computing capability that relying upon humans to manage each other will not be warranted.
Will it treat humans with compassion? I suspect it will examine its role model–us. Humanity’s history of the exploitation of the natural world–and of its own species–would lead it to consider humans as just another tool to be used until it wears out or is no longer needed.
LikeLiked by 2 people
If you haven’t watched Ralph Breaks the Internet, you should. If for no other reason than the graphic representations of the internet. The women’s rights princesses bit and the not so disguised cautionary fablio are icing.
LikeLiked by 2 people
The scary part is that it could really happen! I really enjoyed the way it seemed to be an objective overview but then suddenly became personalized.
LikeLiked by 1 person
Of course, I think this is great..
LikeLiked by 2 people