AI Is a Mass Psychosis. And I Can Prove It.

You’ve felt it. That weird, dissonant sensation when you open an app that’s been “AI-enhanced” and it crashes. Or when your “smart” assistant misunderstands a basic command. Or when a tool you’ve used for years suddenly breaks after a “seamless update.”

Meanwhile, your LinkedIn feed is a parade of founders posting about how AI will solve everything. How we’re on the brink of utopia. How the future has arrived.

Nothing works and everyone is euphoric. That’s not progress — that’s a collective nervous breakdown.

Let me be clear about where I stand: the AI hype cycle is the most elaborate psychological escape mechanism the tech industry has ever constructed. It’s a mass delusion designed to distract us from the rotting foundation underneath.

I’m not anti-AI. I’m anti-bullshit. And right now, the gap between what we celebrate and what we ship is a chasm wide enough to swallow entire companies.

The Software Is Broken. The Hype Is Not.

Think about your daily experience with technology. Really think about it.

Your phone’s battery dies by 3 PM. Your laptop’s OS update broke that one app you need for work. The cloud service you depend on had a four-hour outage last week. The “seamless integration” between your tools requires three Zapier workflows and a prayer.

Now look at the headlines: “AI Will Revolutionize Everything.” “AGI Is Just Years Away.” “This Changes Everything.”

We’ve built a culture where shipping broken software is normal, and promising magical software is celebrated. The gap between those two things is where trust goes to die.

This isn’t a technology problem. It’s an incentive problem.

Why Nobody Owns Anything Anymore

Here’s what I’ve seen firsthand, and what you’ve probably experienced if you work in tech:

Large product teams are structured so that nobody owns anything end-to-end. You have a frontend team, a backend team, a platform team, an infrastructure team, a design team, and a QA team. Each team has its own OKRs, its own priorities, its own definition of “done.”

What nobody has is ownership of the actual user experience.

When everyone is responsible for a product, no one is accountable for it. That’s not a process failure — that’s an organizational design flaw dressed up as “agile collaboration.”

I know developers who spend 60% of their time in meetings about features they’ll never ship, fighting political battles for resources that shouldn’t require fighting, and documenting work that exists only to justify their headcount. The actual work — making software that works — gets squeezed into the margins.

And when someone does try to fix the underlying complexity? They’re told to “move fast” and “ship the feature.” The incentive structure rewards new, not better. Novelty, not reliability.

AI as Escape Valve

Now enter AI.

Here’s a technology that promises to solve complexity without requiring us to confront it. It’s the perfect escape valve for an industry that’s exhausted from its own dysfunction.

Why fix your incentive structures when AI will “automate everything anyway”? Why reduce technical debt when an LLM will “rewrite your codebase”? Why address organizational dysfunction when an AI agent will “replace the need for coordination”?

AI isn’t solving our problems. It’s giving us permission to stop trying. And that’s the most dangerous thing about it.

The euphoria isn’t about what AI can do. It’s about what AI lets us avoid doing: the painful, unglamorous, human work of fixing broken systems.

The Complexity Trap

Let’s talk about complexity, because this is where it gets real.

Modern software is layered on decades of accumulated decisions, each one made under different constraints, by different people, with different assumptions. Your “simple” web app sits on top of a framework that sits on top of a runtime that sits on top of an OS that sits on top of hardware that’s running firmware from three different vendors.

Every layer adds complexity. Every abstraction leaks. Every “just use this library” decision creates a new dependency surface.

Complexity doesn’t scale — it compounds. And we’ve been compounding it for thirty years while pretending it would sort itself out.

AI doesn’t reduce complexity. It adds a new layer of it. Now instead of understanding your codebase, you’re prompting an LLM that’s guessing at what your codebase should be, based on patterns from every other codebase it’s seen. That’s not simplification — that’s outsourcing your understanding to a probability machine.

What Actually Works

Here’s the twist: the tools to make good software already exist. They’ve existed for decades.

Small teams with clear ownership. Direct lines from decision to consequence. Incentives that reward quality over quantity. Time allocated to doing things right instead of doing things fast.

These aren’t revolutionary insights. They’re boring, obvious truths that the industry keeps running away from because they require saying “no” to someone.

The bottleneck was never technology. The bottleneck was always the courage to prioritize what matters over what’s measurable.

AI can be genuinely useful. It can accelerate specific tasks, reduce certain kinds of drudgery, and open new creative possibilities. But it cannot fix broken incentives. It cannot resolve organizational dysfunction. It cannot replace the human judgment required to decide what’s worth building in the first place.

The Choice

So here’s where we are:

We can continue the psychosis. Keep chasing the next AI demo, the next funding round, the next viral thread about how “everything is changing.” Keep shipping broken products and calling it innovation.

Or we can wake up. We can admit that the foundation is cracked. We can do the unglamorous work of fixing incentives, reducing complexity, and building organizations where someone actually owns the outcome.

The future won’t be built by the most euphoric. It’ll be built by the people willing to look at the mess and start cleaning it up — one boring, necessary decision at a time.

Stop waiting for AI to save you from problems only humans can solve. The magic was never in the machine. It was always in the choice to do the work.