The year is 2050. And everything is in ruins.
After several failures of democracy in developed western countries in the early twenties, technocrats had stepped in and promoted a radical idea — that countries, and ultimately — the world — should be governed not by humans, but by artificial intelligence. A trivial idea among futurists, if it wasn’t for one important aspect — that the artificial intelligence would be guided by a democratic process.
And the failures that lead to this proposal were serious. Trump was just the mild messenger of more gruesome processes, not only in the US, but in the whole world. The “checks and balances” broke apart one by one, after carefully prepared court appointments, myopic partisanship and fear.
Scholars theorized that the ultimate reason for the failure of democracy was not the inability of large groups of people to make decisions, but the inevitable corruption of elected officials. Being in position of power, with all the complexities and conflicting interests of the modern world, was said to inevitably lead to compromising one’s ideals. That realization in turn had lead to mostly opportunistic, careeristic, ego or personal gains driven people to run for public offices. For the rest, the deal was a bad one — investing your time and energy only to find yourself in situations where you cannot but compromise on your values.
So the goal became to abolish the “people in power” construct. And fortunately, we all thought, artificial intelligence had been well developed and already serving humans.
The project was developed in cooperation between the best universities and the strongest technology companies of the day. The end result was an AI masterpiece, complex and yet elegant and simple.
The system would monitor all the metrics of society — human development, wealth, equality, safety, etc. It would then pass the necessary legislation to improve these metrics.
If reported crime was on the rise, it would mandate increase of police force, but also better school programs in suburbs and rehabilitation programs; if wealth was declining, it would deregulate markets and lower taxes; if equality was dropping below a calculated threshold, it would increase taxes and institute equal opportunity programs.
And these examples are just scratching the surface — the system monitored thousands of indicators, both objective and subjective, ran simulations of policies and picked the right one for the desired outcome. After indicators changed, it tweaked the parameters of the policies accordingly.
But the designers of the system didn’t want their biases to influence the “optimal” environment, nor they wanted the system to eventually lead to an equilibrium that would mean constant stagnation — of values and markets alike. They believed it is impossible and undesirable for an equilibrium to exist, as people themselves changed the world around them, and therefore new parameters and new methods should be taken into account.
At the same time, the will of the people had to be accounted for — the AI should not just do something if the people don’t want it to, no matter how good it “thinks” it is.
So voting was preserved. But not voting for politicians, who propose and defend policies based on their biases, their lobbying friends, their bank accounts and pensions, and their lust for power — voting for the direction of the policies instead.
Every year each citizen would sign on to the system and express their preference on a number of broader issues — do you want less immigrants, do you want to reduce poverty in your area, do you need more freedom, do you want to be able to do business more easily. Then the votes of all the people were taken into account to tweak the parameters of the policies the AI was introducing. Culture and local context were thus taken into account, and so across the world the system would yield different results in different periods of time.
In a sense, it was a democratically elected machine. First, its very existence and “rising” to power was approved by the citizens. And then its operations were constantly guided by the popular vote.
It never went to any extremes, it kept society engaged in public life, but not for the sake of personal gains or rising to power, but for the sake of what’s right at a given moment. Politicians didn’t cease to exist — but their job was now to advocate for policies they would not benefit directly from. They could not put loopholes in legislation, nor leave an issue ignored.
It was almost like a utopia. Almost, because new issues always appeared, and were always solved in the course of time, with solutions guided by the citizens’ vote and converted to rules and legislation by the artificial intelligence and its access to a huge pool of data.
But it’s 2050 and the world is in smoking ruins. Half of the population is dead, the other half living in unbearable conditions.
No, it wasn’t the global warming — we got that solved through good policies and scientific advancements.
It wasn’t global conflict either — foreign policy was handled flawlessly among a peer-to-peer network of installments of the AI, achieving just the right balance between national and global interests.
It wasn’t hackers — the system was secure and self-healing — any attack or attempt to tamper with the parameters of sources of data would be immediately discovered.
It wasn’t an AI uprising either — Azimov may not have been right with his laws of robotics, but rather similar “checks and balances” were introduced internally to avoid the AIpocalypse.
It wasn’t an alien invasion —apparently no aliens were keen on ruling the small watery planet full of fleshy inhabitants.
It was a revolution. Many revolutions, in fact. And they were due to fear of all of the above.
Few people knew exactly how the system worked. Yes, it was open and everyone could inspect it, but the intricate details of the AI were a complex matter.
Many people didn’t understand how the system worked. That didn’t bother them, of course, as they were busy living their improving lives. But by the year 2045 the once considered “conspiracies” had already gained traction — that hidden elites are manipulating the system to their advantage; that the system cannot be trusted with the future of humanity — that AI will surely take over; that it won’t be prepared to respond properly in case of an alien invasion; that hackers will surely break in some day or that they have already broken in.
And no amount of cryptography explanations, free AI 101 courses or pointing out emergency protocols for “human override” could help to reverse the trend. All of that was seen as cunning propaganda by the AI and its human accomplices who have allegedly made a deal with the machines for personal gains.
Some psychologists argued that this is entirely expected — while conspiracy theories usually don’t “entangle” a large percentage of the population, the situation was different — suddenly a machine was in charge of everything. Even the most corrupt and egocentric leader could be understood and imagined — he’s a human after all — but a machine was something entirely different. And even though it was democratically elected, and its course of action was guided through people’s votes, it was still a machine — a perceived black box.
The sudden transfer of power from humans to artificial intelligence had apparently undermined people’s rational reasoning. They feared what they didn’t understand, and no expert could be sufficiently reassuring that everything is going to be alright.
People started forming anti-machine organizations. Since the system hadn’t been designed to be oppressive, it didn’t monitor personal communication or spy on those organizations and didn’t make it illegal for them to exist, as it would limit freedom. Law enforcement agencies investigated some of them, but that only worsened the problem — the new theme was that the AI wants to disband “the rebellion”.
Some hard to explain policies enacted by the system, some controversial decisions that didn’t make sense without all the data at its disposal, and some prominent political figures taking a stance against “the machine” were all that was needed for things to get out of control.
In no time “the rebellion” amassed hundreds of thousands of people around the world. And they were armed. Civil war has become quite likely — with “the rebellion” on one side, and the rest of the people, who were pretty happy with the way the system worked, including the civil service, on the other side (yes, the AI only decided on the policies — there was still a civil service to implement it).
The dialogues were tense, the online discussions were uglier than ever, and escalation seemed to be around the corner. Any attempt of negotiation was seen as “siding with the machines”, “disregard for humanity’s future” or “ignorant cattle talk”. The non-rebellious side was not helping either — it was labeling the rebellion as “wackos”, “dumb idiots” and “the reason humanity will never truly progress”.
During the election of 2048 it was again time to ask the public about gun ownership — and since the majority of people thought it’s a bad idea especially with “the rebellion” now getting traction, they voted for banning guns. And that happened to be the breaking point. The rebellion…rebelled. They attacked public buildings, killed civil servants and bystanders.
They also organized attacks on data centers that were home to the AI. The situation required the emergency protocol, so humans had to be elected to cope with the situation. Having lied dormant for decades, the opportunistic fear-mongering people were there once again to ask for permission to make everybody safe again. The uprising was promoted to a primary fear. And so after emergency councils were elected, they decided to be brutal against the rebellion.
The escalation was happening all around the world, nearly at the same time. Those in power tried to cut off communications, but the network was so resilient and distributed, the effort was futile.
Rogue military personnel hijacked military equipment for the rebellion, and it turned into an actual war. In some countries the rebellion rose to power, shut down the AI completely, and declared war on neighboring countries that were still using the system. Soon after two blocks formed — the anti-AI and pro-AI countries. It doesn’t matter who won. I don’t remember who won.
It was all a madness that nobody asked for. But it was there. And two years later, we are in ruins.
The democratically elected machine was dead.
I guess we weren’t ready for machines to run our lives. I guess we weren’t prepared for the thought of it. I guess, once again, we had taken a shortcut and used technology before being ready for it. And now we’re back to square one.