Every morning, you tend your crop of prompts. Water them with context. Prune the outputs. Harvest what's useful, compost what's not. The knowledge worker is now a knowledge farmer.
You used to write. Now you cultivate writing. Plant a prompt, watch it grow, harvest paragraphs. Some crops fail. Some need replanting. Some produce unexpected yields. It is agriculture, not authorship.
The Agricultural Patience
The developer who once coded now farms code. They seed AI with specifications, tend the growing functions, debug the pests. Good farmers know their soil—which models grow what code best, which prompts yield clean harvests.
This explains why AI feels like more work: farming always is. You can't just plant and leave. You monitor growth, adjust conditions, cross-pollinate ideas. Every prompt is a seed that needs the right environment to flourish.
The good farmers are emerging. They know which fields to let lie fallow, when to rotate crops, how to read the seasons of model updates. They're building prompt greenhouses, output silos, verification workflows that feel suspiciously like irrigation systems.
We traded direct creation for agricultural patience. Plant, tend, wait, harvest. The immediate became seasonal. The simple became cyclical.
We're all AI farmers now, whether we know it or not. The only question is whether you're subsistence farming or building an estate.