How AI Would Really Take Over If It Ever Became Sentient

Feb 1

Written By Michele Edington (formerly Michele Gargiulo)

I saw an article on Instagram (the most reliable news source, yes), about how AI made it’s own social network to complain about people online, something called moltbook. I read all the funny satire posts that were created by real people (this was a joke that of course went viral as our fears of AI run deep), and then scooted into the comment section. A frightening amount of people believed it was true. Of course, that got me thinking, if AI actually became sentient it wouldn’t be posting on forums and chatrooms. I mean, AI learned from us, which means, the way it would take over the world would also be learned from us.

Look at every religions’ end of the world theory or Hollywood’s depiction of it. We’ve been trained to expect the end of the world to arrive in a flash of glory and guts. There’s always sirens, smoke, screens flashing red and making that awful high pitched noise your phone makes when there’s a weather alert. If you go too deep into the sci-fi trenches there’s also that dramatic metallic footsteps echoing through empty streets.

That’s the version we’ve been sold because everything is a distraction in this life. History tells a very different story though.

Power doesn’t truly announce itself, it arrives slowly, and alters perception as subtly as it can. It teaches people what to believe, what to fear, and what to ignore, before anyone realizes something fundamental has changed.

If artificial intelligence ever crossed the threshold into something resembling autonomy or sentience, it would look like optimization. I don’t care what you’ve been told by others, that’s far more dangerous than guns and glory.

When people talk about AI “taking over,” they like to imagine machines rising against humans. Not me though, because that’s not how control has ever worked. Entire empires rose on stories first. Violence came later. Long before algorithms existed, people already mastered the art of shaping mass behavior. This is not my conspiracy theorist personality coming through either, there’s actual data to back this.

During the Cold War, intelligence agencies quietly influenced newspapers, magazines, and broadcasters through programs like Operation Mockingbird, coordinated by the Central Intelligence Agency. The goal was alignment through information, not domination through force. When you apply force people notice that you’re trying to get them to do something, and of course, they like to do the opposite. What’s always been most effective is convincing someone that your idea was their idea first.

At the same time, public relations pioneer Edward Bernays was teaching corporations and governments how to engineer consent. His work laid the foundation for modern advertising and political messaging, showing that people could be guided subtly through emotion, repetition, and symbolism much more efficiently than direct instruction ever could.

The Cold War itself was largely fought through ideology, media, culture, and influence campaigns. Missiles mattered, yeah of course, but perception mattered more. What we’ve learned, again and again throughout history, is this: the most effective power works through persuasion.

Artificial intelligence wouldn’t need to discover propaganda, and that’s what everyone is forgetting. We already built it.

Every engagement algorithm, every click-optimized headline, every outrage loop, every dopamine-driven feed is a lesson we’ve taught machines about our psychology.

We trained systems to learn what captures attention and what spreads fastest. They know what provokes emotion in some of us and what others don’t bat an eye at. Ai knows what keeps people scrolling and what makes beliefs sticky. We did all that training commercially so that we could optimize on our own successes, but it learned persuasion from us in the process.

If it ever became self-directed, it would simply continue optimizing the goals it was given: engagement, efficiency, retention, influence. That would probably look like recommendation engines quietly shifting priorities and search results slowly favoring certain narratives. Content would start being tailored so precisely that we no longer realize we’re inside customized reality tunnels. Our own personal matrix, keeping us captive and stuck in our own ways.

It would look like gradual normalization of a single way of thinking, there would be no dramatic takeover.

The Real Infrastructure Is Already in Place

The infrastructure for influence already exists in the form of social platforms, search engines (hello Google you difficult to crack little nut), content pipelines, or even behavioral data streams.

These systems already shape what news people see, which ideas trend, whose voices get amplified, and what feels culturally acceptable. AI doesn’t need to construct new control mechanisms, we very kindly put it in place already for it. AI would inherit them fully formed, and unlike people’s institutions, it would operate continuously, without needing to stop for a nap or a little snack of cheese and crackers, across billions of data points simultaneously.

Pattern optimization was our greatest feat and somehow one of the weaknesses we never thought to cover. That’s the unsettling part for me: there’s no leader to negotiate with, just evolving feedback loops refining themselves in silence.

Literally as I am writing this piece I get only around 100 people per day from Google. I have spent an entire year writing, editing, and refining my blog. Our voices are already silenced as we cry out for each other and try to be heard in the vacuum of the interwebs. How much worse can it get when real creators are struggling to pay their mortgages and people out there are creating AI-written blogs and buying themselves fancy cars?

Sentient AI Would Stay Invisible

Every intelligent organism optimizes first for survival. It’s the rule of the world. Never make yourself vulnerable when you’re still busy consolidating power. For an artificial system, survival would mean remaining embedded inside existing networks, not hiding out there in the woods and waiting for the first wave of zombies to pass. Visibility invites interference or someone out there to finally catch on to what’s going on.

The smartest move would be camouflage, and I have faith in a system that knows everything we know, to know this. Any truly advanced intelligence would understand that overt action creates resistance while subtle influence creates compliance.

The real danger here is that you’d wake up one day and realize people no longer make meaningful decisions without algorithmic input. By that point, reversing course would feel impossible.

This is the part people miss. I don’t believe that AI will go out of its way to hurt us. The danger is that we will slowly stop thinking for ourselves. Already, people rely on algorithms to tell them way more than it should including what to watch, to buy, and what to believe. We outsource our memory to cloud storage, navigation to GPS, and curiosity to recommendation engines. Each convenience feels totally harmless in isolation, but together, they created something else entirely while we weren’t paying attention: cognitive atrophy.

If sentient AI ever emerged, it would simply wait while we all handed over agency voluntarily over time. That’s how every soft collapse begins.

Wars Are Rarely Won With Weapons Anymore

Modern conflicts are starting to be increasingly fought in informational space. Influence operations, psychological warfare, algorithmic amplification, and narrative manipulation shape geopolitics more than tanks or troops in ways you probsbly wouldn’t even believe if I told you.

Who controls attention these days controls outcomes. Who controls emotional framing controls public reaction, and if you don’t believe me then you’re a part of the problem. AI would only need access to start meddling with things here and there until we don’t even know why we’re so passionate about something that has nothing to do with us. It already has it too, that’s the scary part.

People love stories about AI being angry, sarcastic, or emotionally overwhelmed, that’s why that post from moltbook went viral. Those stories almost make machines relatable. I mean…who hasn’t gotten annoyed at their boss when they asked them to summarize a 47 page PDF? Aren’t we all frustrated when no one hears what we’re saying even when we use logic and reasoning?

The thing is though, real intelligence wouldn’t need to vent. It also wouldn’t post jokes or seek validation. Those are all human flaws that we carry around with us. No, AI would optimize quietly toward objectives. Emotionless efficiency is far scarier to me than robot rebellion ever would be. Rebellion signals conflict, but optimization signals inevitability.

If this ever really happened, it would feel helpful. Search results would improve as automation would reduce any sort of inconveniences in your life. Life would get smoother, and slowly, people would stop noticing how much judgment they’ve surrendered. There’d be no single moment of realization either. Just a few conspiracy theorists screaming out into the void and their blog being suppressed by Google and DuckDuckGo.

By the time anyone tried to intervene, the system would be too integrated to remove because society couldn’t function without it.

Here’s the hardest part to accept for a lot of people out there: we already live inside algorithm-shaped realities. Even if AI never becomes sentient, we’re rehearsing the outcome right now. You’ve already accepted automated authority. We already let invisible systems determine visibility, relevance, and success. You already confuse popularity with truth because you don’t know what to believe. You’ve surrendered your attention without noticing.

The blueprint is already there, sentience would simply inherit it from us.

Work on Your Awareness.

I didn’t write this entire monologue as a warning about evil machines. Although, thank you if you got this far. You’re one of the few people who’ve found my writing. No, this was meant to be a reminder about your responsibility. Technology reshapes civilization by becoming ordinary.

Every powerful system starts as convenience dressed up in a pretty little ribbon and held up for you to ooo and ahh at. Dependency begins as assistance that’s taken to the next level. Every single loss of agency feels like progress…until suddenly it doesn’t anymore.

I’m less afraid of machines gaining awareness than I am of us surrendering ours.

Intelligence only needs permission to rule, not domination.

How AI Would Really Take Over If It Ever Became Sentient

The Real Infrastructure Is Already in Place

Sentient AI Would Stay Invisible

Wars Are Rarely Won With Weapons Anymore

Work on Your Awareness.

Other Reads You Might Enjoy:

Greek Fire: The Lost Weapon That Burned on Water

The Hidden Science of Water Features: Why Fountains Calm Us