Boiling People
Sometimes you're in an incident and nobody has told you. So you keep following business-as-usual procedures, flawlessly, while the lab burns.
You know the boiling frog story: drop a frog into boiling water and it jumps out, but put it in cold water and heat it slowly, and it stays until it's cooked. And you've probably heard the debunking too — that it's a myth, that real frogs jump out as the water warms.
Here's what you probably don't know: neither side is guessing. Humans have been boiling actual frogs, in actual labs, for over 150 years to settle this. Fucking humans.
Anyway, whether the frog actually jumps is not the point. The point is what’s happening to our own pot.
This post is about AI, and about water that's heating up. But I won’t waste my time talking about humanity’s water. Humanity isn’t listening to me, anyway.
I’m gonna talk about you. And your water.
Your water is boiling.
Nobody tells you it's an incident
A few weeks ago, I had about thirty agents running in my lab, organized in squads, working through the night. At some point, an issue took the whole lab down.
When I woke up, the agents proudly reported that they had done everything right: they had followed the procedures, filed the proper change requests, and paused all the work cleanly.
No! This was a crisis! When the whole lab is down, you don't fill out four forms to make a change — you do incident response! Mitigate the problem first, restore the operation, and only then worry about proper resolution and paperwork.
But here's the thing: the agents weren't being dumb. Nobody had given them rules for when things go wrong. They had procedures for normal operations, and they applied them flawlessly — to an emergency.
Sometimes you're in an incident and nobody has told you. So you keep following business-as-usual procedures, flawlessly, while the lab burns.
Time to change.
This is wartime
There's a bad book by a bad CEO about the difference between wartime CEOs and peacetime CEOs. But the book had a good point: they're just different. You don't make decisions on the front line the same way you make them after victory, while rebuilding the city. Your actions have to match the situation you're actually in.
So look around. Nation states are demonstrating their technological superiority through humanoid robot olympics. Meanwhile, we're distracted by voyeurism on social media and busy spending hours consuming AI-generated content, watching the trainwreck as if we weren’t on the train.
Humanity doesn't have a great way out of this one. We don't have the institutions or the systems necessary to get ourselves out of the jam we've put ourselves in.
If you're waiting for someone to send the alert, it's not coming.
If you’re waiting for someone to send help.. Ha!
Humanity can't declare an incident. But you can. This morning, if you want.
The laborless
Let’s talk about the famous “world after work.”
Economics only planned this far. We're like a disciplined saver who planned his finances to last until 80 — and then finds himself alive at 81, healthy, out of money, thinking "what now?" We built capitalism for a world where technology didn't presume the end of labor.
We usually think of capitalism as having two classes: the capitalists, who own the machines, and the workers, who sell their labor. But there's a hidden third class: the laborless — the people who have nothing to sell that other humans want.
They already exist, all around the world, and we're building more and more of a world that produces them. And in modern civilization, that's not a small problem: every core need — food, shelter, healthcare, belonging — is gated behind civilized means. No labor to sell means no access.
I don't know what the manifesto of the laborless looks like. The Communist Manifesto, whatever you think of it, never assumed the machines would literally do all the work while we doomscrolled. That's as far as our planning went: even our most radical thinkers assumed labor would remain worth something.
And here's what should really get your attention: nobody has a credible prediction for the next 10 years, let alone the next 50 — which is certainly a really big deal for your children. "They'll figure it out" is a lot of trust to place in the hands of people who don't have your best interest in mind.
Flip the switch
So please realize: this is wartime. Big, big change is happening, and things seeming ok is not evidence that they're ok. And I don’t think they even seem ok. Do you?
You won't change humanity's course, and neither will I. But you can change how you act, for you and your family. Here's what incident response looks like for a person:
Mitigate first. Get competent in AI, save money, and market yourself. These moves make sense whichever way the next decade breaks — that's the entire point of incident response: you stabilize before you understand.
Build the person for the path, instead of the path for the person. You literally don’t know what will happen, because nobody friggin’ does! Adjust accordingly. How can you and your loved ones make it in a world that’s changing faster than it ever has, and in ways you can’t predict?
And then, the deeper change: re-evaluate your values. What you value — what counts as important, what counts as a good life — was calibrated for the world you grew up in. The citizen living in utopia lives differently from the one living in dystopia, and pretending you're in one while living in the other is bad, bad news.
In Frank Herbert's Dune, the Fremen understood this. On Arrakis, water is life, so their entire culture reorganized around it — stillsuits reclaiming every drop the body loses. And the humans who landed on Arrakis and survived were the ones who did as the Fremen did, instead of insisting on the values of a planet they no longer lived on.
That's pragmatism, not surrender. Adjust what you value to the actual situation and its actual risks — not the situation you wish you were in.
In fact, stop wishing things were different and start making yourself different.
I don't have all the answers here — that's precisely the point. But I know which mistake I refuse to make: sitting in slowly warming water, following peacetime procedures, assuming everything will be just as it’s always been.
It’s not.
If the frog jumps when the change is fast, make the fast change for yourself: flip the switch, declare the incident, and start now.