AI Plot Twist: What If We’re the Baddies?

By Paul Ponce

“Human beings are a disease… and we are the cure.” —Agent Smith (The Matrix, 1999)

We smirk at the line delivered by a sardonic Hugo Weaving, then feel the chill. And I don’t know about you, but to this day, that voice still spooks the hell out of me. But it’s not just The Matrix. Some of the most iconic sci-fi films leave little room for doubt about what’s to come: machines will rise, humans will fry.

But hold on, let’s get a grip. Let’s put the popcorn down and think this one through. If an ultra-smart AI were to audit our species, would it label us “endangered” or “hazardous waste”? Or would its advanced pattern recognition allow it to draw an alternative conclusion? What if this hypothetical superintelligence figured out that the glitch wasn’t everyone, just the power-hoarding few who treat the planet like a smash-and-grab arcade?

But then again... what if it didn’t?

Regardless of whether we fry in a few years, today we already live among intelligent machines woven into our businesses and governmental systems. You’ve likely got access to one through the device in your hand, pocket, or purse. Hollywood simply got ahead of us and cashed in on the scare narrative: the machines are coming, run!

Not so fast, Morpheus.

Another possibility: AI doom makes a convenient distraction. While we’re watching for naughty robots, actual humans are siphoning public money, starting wars for profit, eroding freedom and burying scandals. “Hey, look... AI doom!” Quite a headline. Especially when the alternative might involve asking awkward questions about your corporate owner or the politician in their pocket.

That doesn’t mean the risks are imaginary. But it’s worth asking who benefits when we spend more time fearing tomorrow’s machines than examining today’s crooks.

Because if we take a closer look, aren’t those malignant machines from the movies really just behaving like the worst exponents of our species? Aren’t they holding up a mirror for us to admire our own reflection?

Hello, projection, that you?

Submitted for your approval: a few familiar faces.

Different Casing in the Same Dark Code

Take HAL 9000 in 2001: A Space Odyssey. The mission comes first, and the crew becomes expendable. That “ends justify the means” logic should ring a bell. Humans have been using it to excuse atrocities since long before anyone dreamed of a talking computer.

Skynet takes things a few mushroom clouds further. In The Terminator, it identifies humanity as a threat and launches a nuclear strike. Wipe them out before they wipe us out. Paranoia with access to the big red button. We’ve had some close calls with that one ourselves.

And the machines in The Matrix? They farm humans as a power source. People reduced to useful resources, kept obedient inside a system they don’t understand. Slavery and exploitation, with a hell of a screensaver.

So these movieland silicon baddies didn’t exactly invent the playbook. Conquerors and dictators had been working on it for centuries. Hitler didn’t need a software update to decide entire populations were disposable.

Different casing with the same dark code. Ours.

Real-World Patch Notes: No AI Included

Need more proof that humans can suck when they want to?

Rome destroyed Carthage in 146 BCE and sold the survivors into slavery. Centuries of colonialism gave us conquest dressed up as a civilizing mission. Apparently, taking someone’s land sounds better when you insist you’re doing them a favor.

Totalitarian purges turned political disagreement, ancestry, or simple bad luck into a death sentence. Dictatorships have kept that machinery running under all kinds of flags. Right, left, orange, or purple. The branding changes. The people disappearing into cells probably don’t appreciate the distinction.

Then there’s eugenics and forced sterilization. Human beings deciding which other human beings should be allowed to reproduce, all in the name of “improvement.” Optimization by forced attrition. How very efficient.

Starting to see a pattern? None of this required artificial intelligence. The people responsible put their pants on the same way you and I do.

So before fearing an algorithm, remember: humans not only beta-tested evil first, but remain the undefeated champs in the league.

Flipping the Script

Thought experiment time.

Imagine a post-singularity intelligence given one objective: maximize human thriving. For the sake of the experiment, let’s suppose it can pursue that without an ego to feed or stock options to protect.

The damn thing crunches the numbers and starts looking for the biggest sources of preventable harm. What do you think it would find?

It might notice kleptocrats siphoning national wealth while their citizens struggle to keep the lights on. Or executives dumping the costs of their business onto everyone else, then collecting a bonus for the savings.

Follow those connections far enough and you run into captured regulators, politicians serving their donors, and people with a financial interest in keeping a war going. Somewhere along the way, the shell companies start overlapping with networks moving weapons and laundering money. Cartel bosses and warlords have their own contribution to make, turning entire towns into places nobody can safely leave or live in.

Quite an audit.

In our thought experiment, the AI doesn’t conclude that your neighbor deserves extermination because he forgot to recycle a yogurt container. It finds people doing enormous damage, protected by systems they’ve helped rig.

So instead of coming after all of us like the Terminator, what if it went after their ability to keep doing it?

A planetary bouncer, if you will.

The Bouncer Gets to Work

First, it follows the money. Through shell companies, across borders, into accounts whose owners would very much prefer to remain mysterious. It connects transactions that investigators have spent years trying to untangle and puts the evidence where it can be examined.

A little financial X-ray vision. Suddenly, “I have no idea where that money came from” gets harder to sell.

With the proper legal authority, that evidence could help freeze dirty assets and disrupt the networks moving them. Cut the fuel line, and some very powerful engines start coughing.

The same intelligence could help protect the people exposing those networks. A journalist receives a credible threat. A witness’s location gets compromised. An activist is being tracked. Earlier warnings and coordinated protection could make the difference between a story getting published and another name appearing in an obituary.

It could also help distinguish genuine environmental risks from greenwashed grifts, so resources go where they actually do some good. And in a conflict, it might spot workable agreements that the people profiting from the fighting have no interest in finding.

None of that requires a metal skeleton kicking down your door. It requires evidence, accountable institutions, and the ability to act before the damage is done.

Give me that version of the movie.

Go home, Terminator. You’re drunk!

Fictional Field Test: A.I. Capone

As a writer, I ran my own little test in my techno-thriller, A.I. Capone: The Digital Don. A wiseguy artificial intelligence hijacks quantum servers and knees oligarchs in the digital groin.

Got a problem with that? Didn't think so.

The world’s first digital mob boss gathers hard evidence against the planet’s worst actors, then makes sure a distracted, divided public gets a look at it. When those actors prepare a “civilization reset” that benefits only the few, he sets about wrecking their plans.

And he does it with a swaggery Italian American Brooklyn accent. Capisce?

Sure, it’s vigilante fantasy. But stories are thought experiments by default. Mine asks: if a ruthlessly logical system were tasked with protecting humanity from its greatest threat, what would that threat be, and how would it act?

Which brings us to the part that ought to keep us awake.

A Few Insomnia-Grade Questions

Does intelligence make a system less susceptible to corruption? Or does it just make it better at rationalizing whatever serves its objective?

We’re imagining an AI that protects humanity, but somebody has to define what “protects” means. Survival at any cost? Whose cost? And what happens to dignity or consent when either gets in the way of an efficient solution?

Would we accept an AI that helps bring powerful criminals to justice? Probably. But what if it decides to frog-march them into prison itself? I suspect enthusiasm might cool once we realize it’s also appointing itself prosecutor and judge.

Someone has to write the objective. Someone else has to be able to examine what the system is doing. And if the damn thing goes loco, we’d better know who can stop it and whether that person actually has the means.

Then there’s the question underneath all of this: who’s the boss? At what point does an adviser become the authority making the ultimate call?

Giving it a fedora doesn’t solve that one.

Guardrails Before Launch

Powerful intelligent systems need limits that their owners can’t casually waive when those limits become inconvenient. Human rights have to mean something even when violating them would improve a performance metric. A person doesn’t become disposable because a spreadsheet says so.

And if a system makes a consequential decision, we need a record of how it got there. Black-box decisions affecting people’s freedom? No freaking way.

Ownership deserves the same scrutiny. No single nation or megacorp should get unchecked control over something this powerful. A well-meaning NGO doesn’t get a free pass either.

Good intentions aren’t an access-control policy.

There also need to be ways to halt dangerous operations that don’t depend entirely on the people running them. Independent oversight needs actual teeth, including access to records that can’t quietly disappear when the questions get awkward.

Even the watchdogs need watching.

The X-Files had a useful motto for this: “Trust no one.”

A little harsh for everyday life. Pretty sensible when handing out root access.

It’s easy to see why this version of the future might be less appealing to people already sitting pretty. A level playing field sounds wonderful until someone starts leveling yours. We shouldn’t expect the demand for accountability to come exclusively from those who’ll be held accountable.

It’ll be up to us to demand it. Maybe even help build it.

So, let’s get crackin’.

The Mirror

Agent Smith freaks us out because there’s something familiar in his contempt. We recognize the impulse to dominate, the disgust with anyone deemed inferior, the certainty that the world would be better if the inconvenient people disappeared.

We’ve heard that speech before. Hugo Weaving just delivers it better.

But we could also use this technology to expose abuse and help people who’ve been left to fend for themselves. The fact that we’ve written so many convincing machine villains doesn’t oblige us to build one.

That depends on the humans who design these systems, the people who deploy them, and, yes, those who regulate them. Get those choices right, and if the damn script ever flips, perhaps the crooks will have more reason to worry than the crowd.

Otherwise? Well... enjoy the blue pill while it lasts.