I’m Rooting for AI to Blow Up in Everyone’s Face

Good evening. I’m pissed off tonight.

My guess is if you work in tech, or even around tech, you’re probably pissed off too. So let’s talk about it.

No script. Just notes and conviction.

The State of Tech Is All Messed Up

The last few months in tech have been chaos.

All of a sudden, people with zero background are shipping code with tools they barely understand. People who have never opened an IDE are now running around with Claude Code and Cursor like they’ve been doing this for years.

There are no guardrails.

And honestly, I’m rooting for consequences.

In 2026, I’m rooting for a giant, painful reality check. Data breaches. Leaked credentials. Teams learning the hard way that you cannot blindly hand over critical systems to people who do not understand the underlying technology.

If you think that is gatekeeping, I don’t care.

This is not gatekeeping. This is refusing to hand fully automatic weapons to toddlers.

If I have an electrical problem at my house, I do not start twisting wires because a chatbot gave me a confident answer. I call an electrician. Why? Because I respect expertise. Because I know what I don’t know.

It’s Not Just Programming

This is bigger than coding.

It’s product. Project management. Engineering leadership. People management. Whole roles are getting dismissed by people who think the only thing that matters is raw code output.

You can hear the bitterness in it. A lot of this is not about productivity. It is resentment dressed up as innovation.

“Managers are useless.” “PMs are useless.” “Everyone is replaceable.”

Cool posture. Very edgy.

But reality is still reality. If I botch plumbing, I flood my house. If I botch electrical, I get hurt. In software, damage is less visible, so people pretend it is less real. It isn’t.

Broken software can limp along long enough for people to miss the risk. Then one day it does not limp anymore. It detonates.

And yes, I think some of that pain is necessary, because people need to learn that technological progress is real, but paradigm shifts still have a cost.

The Transhumanism Delusion

Part of this comes from a recurring Silicon Valley fantasy.

The fantasy is that replacing humans is inherently good, and that wanting to be replaced is somehow enlightened.

I keep hearing people say things like, “I can’t wait for AI to do my job.”

What is wrong with you?

Seriously. If you feel that way, explain it. I genuinely want to understand the worldview where erasing human contribution is treated as progress.

The Hard Thing About Hard Things

The industry changed overnight. Everyone can see that.

Now you have people with minimal experience, or rusty experience from 10 to 20 years ago, suddenly presenting themselves as authorities.

What they miss is this:

Most of the hard things about hard things have very little to do with the visible hard thing.

If I pay an electrician a flat hourly rate and he fixes my issue in 30 minutes, I am not paying for 30 minutes of wire-twisting. I am paying for 10 years of apprenticeship, pattern recognition, scars, judgment, and responsibility.

I am paying for lived experience.

AI does not have lived experience. It does not have a body. It does not have stakes. It cannot get shocked. It cannot go home to kids after a dangerous day and think, “I need to be precise because my family needs me.”

Humans carry context, memory, fear, care, and consequence into their decisions. That matters. A lot.

The Sad People

There is also a social layer here that nobody wants to talk about.

A lot of people are deeply unhappy, and AI is amplifying that unhappiness. It is making people more of who they already were.

If someone already thinks all value can be reduced to standardized outputs, certifications, and scorecards, AI gives them a weaponized version of that worldview.

“Why does expertise matter if ChatGPT can do PhD-level work?” “Why does craftsmanship matter if the output looks fine?”

These are not smart questions. They are flattening questions.

And the people asking them often seem less interested in building anything than in devaluing everyone else’s work.

What You Get When You Optimize for Numbers

I once worked for someone who was a genuine genius. I do not use that word lightly.

He had kids. When I asked him for fatherhood advice before my first child was born, his answer was basically metrics. Put them in front of the iPad. Boost test performance. Quantify progress.

That was his definition of success.

Not connection. Not tenderness. Not whether your kid runs to you when they’re hurt. Not whether they feel safe with you.

Just metrics.

And that is exactly the trap we are walking into at scale. Everything reduced to numbers because numbers are legible, even when they miss the point.

What I Actually Want for My Kids

I have one daughter and another child on the way.

So I think about this all the time. What do I really want for them?

Do I want elite test scores, awards, and external status? Sure, those things can help.

But more than that, I want peace. I want them to be grounded in who they are. I want them to live with integrity, not just achievement.

Money matters. Intelligence helps. Life is easier with resources.

But if we optimize only for measurable outcomes, we can win on paper and still lose the human plot.

The Resistance

So how do we push back?

Not by grinding harder inside the same broken value system. Not by chasing more views, more money, better toys.

We push back by refusing the cult of optimization.

We push back by owning the parts of ourselves that are imperfect and human. The weird edges. The emotional intensity. The scars. The contradictions.

Those are not bugs to remove. Those are sources of identity.

In a world that increasingly rewards sameness, your imperfections are part of your signature.

The Beige Homogeneity

At the core, LLMs are probability engines. They choose the next most likely token. Yes, there is more complexity around that now, but the gravitational pull is still toward the statistically expected.

Expected language. Expected opinions. Expected style. Expected behavior.

That is how you drift into beige homogeneity. Everyone sounding similar, thinking similar, presenting similar.

But real culture is not built by statistical middle lanes. It is built by people and communities with distinct values, histories, and rough edges.

You should be allowed to say, “I’m imperfect, but this is who I am,” without apologizing for not matching the template.

Just Be Yourself

The best resistance right now is simple.

Be yourself.

Self-improvement is good. Worshiping self-optimization is not.

I got heated in this video. Fine. That is me. I am not interested in sanding myself down into a polished generic sphere to satisfy an algorithmic taste profile.

You do not have to do that either.

Tech used to attract people who were deeply different from each other, and that was part of the magic. Different minds, different methods, different obsessions, all colliding to build new things.

If we replace that with interchangeable outputs from interchangeable workers, what is even left?

The Beauty of Broken Things

I miss messy tech.

I miss strange systems made by strange people. I miss rough edges. I even miss bugs, sometimes, because they remind you there are humans in the loop.

Human creators are flawed. Sometimes seriously flawed. That reality is uncomfortable, but it is real.

The question is not whether humans are imperfect. The question is whether we still believe imperfect humans can make meaningful things together.

I do.

And I think tech is one of the last places where broken people can still build beautiful systems in public.

That is worth defending.

Thanks for reading.