I'm not anti-automation. I build systems for a living. I want the machine. I want the workflows that run while I sleep, the agents that handle the repetitive work, the infrastructure that means a decision only has to be made once. If something can be systematized, systematize it. That's not the enemy of good work. That's what frees you up to do it.
But there's a quiet failure mode that nobody warns you about, and I've watched smart people walk straight into it. You get so good at building the machine that you start running yourself like one. Every hour optimized. Every conversation a transaction. Every decision routed through whether it scales. And one day you look up and realize you automated the busywork and then turned the same logic on your own life.
You automate the busywork. Then you turn the same logic on yourself.
The same thing happens to companies. They get efficient. Ruthlessly, beautifully efficient. And somewhere in the optimization they sand off the exact thing that made anyone care about them in the first place. The weird voice. The founder who answered emails at midnight. The product that was a little too generous because nobody had told them yet that it didn't pencil out. Efficiency is supposed to serve the thing. Too often it eats it.
So here's the line I keep coming back to. Build the machine. Just don't become one.
What that means in practice is that you have to decide, on purpose, which parts you refuse to automate. Not because you can't. Because they're the point. The judgment about what something means. The choice of who it's for. The moment of actually caring whether it lands. The relationships you'd never run through a scoring model. Those aren't inefficiencies to be eliminated. They're the reason the machine exists.
This gets more urgent every year, not less. As the tools get better at generating everything, the generated stuff becomes the floor, not the ceiling. Anyone can produce the average now. What's scarce, and getting scarcer, is the human underneath it. The taste. The conviction. The willingness to make a call the data can't make for you. The more the internet fills up with machine output, the more the un-automatable parts of you are worth.
I think the people who thrive in the next decade are the ones who get this balance right. They'll automate aggressively, with no nostalgia for work that didn't need a human. And they'll be fierce about protecting the small set of things that did. They'll build the machine to buy back the time, and then spend that time being more human, not less.
Use the tools. Build the systems. Let the machine carry what it can carry. And then, deliberately, keep the part of the work that only you can do, the part that was the reason all along. Build the machine. Just don't become one.