Staying Human in the Age of AI
The machines are getting better at being us. Here's what I think we should keep for ourselves.
Every week something I used to call hard gets easier. A draft, a summary, a plan — handed to a model and returned in seconds. The reflex is to be impressed. The harder question is what’s left for me to do, and whether I still want to do it.
I don’t think the answer is to compete with the machine on its terms. I think it’s to get clearer about which parts of the work were never really about output at all.
The parts that don’t compress
When I write, the value isn’t the finished paragraph. It’s that I had to think clearly enough to produce it. Outsource the paragraph and I keep the polish but lose the thinking — which was the only part that changed me.
So I’ve started sorting work into two piles. One pile is throughput: things I want done, where the doing teaches me nothing. The other is the work that is the thinking. The machine gets the first pile. I keep the second, jealously.
What I’m keeping
- Taste — knowing which of ten good options is the right one.
- Judgment under ambiguity — deciding when the data runs out.
- The relationships — no model has earned anyone’s trust for me.
None of these compress well. That’s exactly why they’re worth holding onto.