If you haven't noticed, Mark Zuckerberg is having a midlife crisis, and unfortunately for the rest of us, he's got billions of dollars to work through it.

After fumbling Llama — Meta's answer to ChatGPT that landed with all the impact of a jab from Joe Biden — and watching OpenAI's ChatGPT become a household name while his chatbots gathered digital dust, Zuck is now throwing nine-figure salaries at anyone who helps usher in superintelligence. In other words, godlike AI. The kind that will apparently save humanity from itself.

The warning signs were all there. First came the pivot to jiu-jitsu. Then the hair. Out with the North Korean intern bowl cut, in with a tousled look that whispers, “I read emotions now.” And then — God help us — the gold chains. Jewelry. On a man who once dressed like a CAPTCHA test for “which one is the tech CEO.”

We're likely looking at AI trained on the digital equivalent of gas station hotdogs — technically edible, but nobody with options would choose them.

Call me a skeptic. I've been called much worse. The same man who turned Facebook into a digital landfill of outrage bait and targeted ads now wants to control the infrastructure of human thought. It’s like hiring an arsonist to run the fire department, then acting confused when the trucks keep showing up late and the hoses are filled with gasoline.

Diversifying dopamine

Facebook's transformation from college networking tool to engagement-obsessed chaos engine wasn't an accident — it was the inevitable result of a company that discovered outrage pays better than friendship. While Google conquered search and Amazon conquered shopping, Meta turned human connection into a commodity, using Facebook, Instagram, and WhatsApp to harvest emotional reactions like a digital strip mine operated by sociopaths.

The numbers tell the story: Meta's revenue jumped from $28 billion in 2016 to over $160 billion today, largely by perfecting the art of keeping eyeballs glued to screens through weaponized dopamine. The algorithm doesn't care if those eyeballs are watching cat videos or cage fights in a comment section; it just wants them watching, preferably until they forget what sunlight feels like. Now, Zuckerberg wants to apply this same ruthless optimization to artificial intelligence.

The pattern is depressingly familiar: Promise connection, deliver addiction. Promise information, deliver propaganda. Promise intelligence, deliver ... what, exactly? Given Meta's track record, we're likely looking at AI trained on the digital equivalent of gas station hotdogs — technically edible, but nobody with options would choose them.

The growth trap

Zuckerberg's AI pivot reveals a fundamental truth about modern tech giants: They're trapped in their own success like digital King Midases, except everything they touch turns to engagement metrics instead of gold. Sure, Meta still owns three of the most used platforms on Earth. But in the age of AI, that’s starting to feel like bragging about owning the world’s nicest fax machines.

Relevance is a moving target now. The game has changed. It’s no longer about connecting people — it’s about predicting them, training them, and replacing them. And in this new arms race, even empires as bloated as Meta must adapt or die. This means expanding into whatever territory promises the biggest returns, regardless of whether they're qualified to occupy it. It's venture capital Darwinism: Adapt or become irrelevant.

RELATED: Mark Zuckerberg is lying to you

  Photo by Alex Wong/Getty Images

When your primary product becomes synonymous with your grandmother's political rants and your uncle's cryptocurrency schemes, you need a new story to tell investors. AI superintelligence is that story, even if the storyteller's previous work involved turning family dinners into ideological battlegrounds.

The Altman alternative

Comparing Zuckerberg to Sam Altman is like asking whether you'd rather be manipulated by someone who knows he's manipulating you or someone who thinks he's saving the world while doing it. Altman plays the role of philosopher-king well. Calm and composed, he smooth-talks AI safety as he centralizes power over the very future he's supposedly protecting. Zuckerberg, by contrast, charges at AI like a man chasing relevance on borrowed time: hyperactive, unconvincing, and driven more by fear of obsolescence than any coherent vision.

The real question isn’t who is worse. It’s why either of them — men who have already reshaped society with products built for profit, not principle — should now be trusted to steer the next epoch of human development. Altman at least gestures toward caution, like a surgeon warning you about risk while sharpening the scalpel. Zuckerberg’s model is simpler: Keep breaking things and hope no one notices the foundations cracking beneath them.

Zuckerberg's real genius (if you can call it that) lies in understanding that controlling AI isn't about making the smartest algorithms. It's about owning the infrastructure those algorithms run on, like controlling the roads instead of building better cars. Meta's massive data centers and global reach mean that even if its AI isn't the most sophisticated, it could become the most ubiquitous.

This is the Walmart strategy applied to AI: Undercut the competition through scale and distribution, then gradually degrade quality while maintaining market dominance. Except instead of selling cheap goods that fall apart, Meta would be selling cheap thoughts that fall apart — and taking your society with them.

The regulatory void

The most alarming part of Zuckerberg's AI crusade isn't his history of turning every good intention into a cautionary tale. It's the total absence of anyone capable of stopping him. Regulators are still trying to untangle the damage social media has done to public discourse, mental health, and America itself, like archaeologists sifting through digital rubble. And now they're expected to oversee the rise of artificial superintelligence? It's like asking the DMV to run SpaceX: painfully unqualified, maddeningly slow, and guaranteed to end in catastrophe.

By the time lawmakers figure out what questions to ask, Zuckerberg will already own the answers and probably the lawmakers too. The man who testified before Congress about data privacy while reaping user info like a digital combine harvester now wants to build the systems that will make those hearings look quaint. It's regulatory capture with a time delay.

Zuckerberg's AI venture will likely follow the same trajectory as every other Meta product: promising beginnings, rapid scaling, quality degradation, and unintended consequences that make the original problem look like a warm-up act. The difference is that when social media algorithms prioritize engagement over accuracy, people share bad takes and ruin Thanksgiving dinner. When AI systems optimize for the wrong metrics, the collateral damage scales exponentially, like going from firecrackers to nuclear weapons.

The man who promised to "connect the world" ended up fragmenting it like a digital sledgehammer. The platform that pledged to "bring the world closer together" became a master class in division, turning neighbors into enemies and family reunions into MMA fights. Now he wants to democratize intelligence while building the most centralized cognitive infrastructure in human history.

Mark Zuckerberg has never built anything that worked as advertised. But this time is different, he insists, with the confidence of a man who has never faced consequences for being wrong. This time, he's not just connecting people or sharing photos or building virtual worlds that nobody visits. He's building artificial minds that will think for us, decide for us, and presumably share our private thoughts with advertisers.

What could go wrong?

Everything. And if and when it does, there won't be a "delete account" button. The account will be your mind, and Mark Zuckerberg will own the password.