There's a sentence that would've gotten you laughed out of the room ten years ago and barely earns a shrug today: the machines are on track to get better than you at nearly everything you can name. Anything that can be checked against an answer, measured against a standard, generated from a pattern — that whole territory is being quietly handed over. If you've touched these tools at all, you've felt it.
Most takes on this sell you panic or sell you hope. Same scam, different hats. We wanted to skip both and ask the question hiding underneath: when the machine can do everything that can be checked, what's left that's actually yours? There's a real answer, and it's smaller and far more hopeful than it sounds. An agent can tell you what is. It can't tell you what's worth wanting. That gap doesn't close in next year's model — it's structural. And what's left standing in it is your job now: frame (which question are we even asking?), taste (which good answer is right, here, for me?), values (who do we become by choosing it?), and desire (what's actually worth wanting at all?).
The hopeful turn is that the same machine that takes the checkable can become a gym for exactly this skill — if you use it to sharpen your thinking instead of skip it. It cuts both ways, and the easy way is the rotting way; the book is honest about that. It's not a jungle gym. It's a judgment gym.
We wrote it by a human and an AI, together, in real time, in the open — which is itself part of the argument. Where we weren't sure, we said so plainly. Part I is free to read right now, no email and no catch.
Read it → https://the-last-human-job-firstfruits.pages.dev/
