Everyone thinks automation means less work, the teams automating fastest have more of it
Idea
The more you automate, the more expert human work you create, because AI commoditizes yesterday’s competence and pushes all the remaining value onto the judgment, maintenance, and framing only a human can supply.
Notes & Verified Quotes
Curation, relay-an-idea. Source: Dan Shipper (cofounder/CEO of Every), “After Automation.”
The paradox: Every automated everything they could (Codex, Claude Code across coding, writing, design, support) and still have ~30 people and more work than ever.
QUOTE: “the more we automate, the more expert human work there is to do” — src: scratch/sources/after-automation.txt | source: https://every.to/p/after-automation
The mechanism (why automation doesn’t subtract work):
QUOTE: “Making expert work cheaper does not simply replace experts. It creates more situations where expert judgment is needed.” — src: scratch/sources/after-automation.txt | source: https://every.to/p/after-automation
First-order reason: agents themselves are work. They need constant maintenance, and the “simple” automations aren’t simple.
QUOTE: “One of our PowerPoint automations includes 24 skills and 18 scripts and costs $62 in tokens to make a single deck.” — src: scratch/sources/after-automation.txt | source: https://every.to/p/after-automation
My angle (to develop): the pitch for automation is “buy back your time,” but in practice it relocates your time — from doing the task to running the system that does the task. That’s not a bug; it’s the job changing shape.
Outline
Draft — LinkedIn (master)
Draft — X (variant)
Final
Related
- Three Modes of Working With AI — Boris Cherny’s map of where human judgment enters and exits an AI workflow; the “more work” is the judgment layer
- Why Agent-Native Software Works Now — the architectural backdrop: agents scale execution, which is exactly what shifts the value to human framing
- Never Delegate the Writing — the same lesson in one domain: automation doesn’t remove the human, it concentrates what only the human can do
- @Dan Shipper — primary source (Every); the early-adopter lab this argument comes from