The economics of automation
Why one job gets automated and the near-identical one next to it does not.
What this course is
An Explore course, which in most catalogues means a video essay. This one does not, because an opinion about the future is cheap and you can get one anywhere.
Instead you will take a task you already understand — from your work, your family's business, your own week — and put numbers on it. By the end you will be able to say why that specific task is or is not worth automating, and roughly when that answer changes.
What you'll leave with
- A task decomposition of one real job, split into things that are and are not candidates.
- A payback figure in months, computed from your own numbers rather than a vendor's.
- A written limit — the condition under which your own arithmetic stops being true.
Nothing here needs hardware, an account, or a model. It needs one job you know well.
Module 0 — What actually gets automated
Free, and enough on its own to change how you read the news about this.
- The task, not the job — Jobs are bundles of tasks with wildly different economics. Averaging them is how everybody gets this wrong. (20 min)
- The payback arithmetic — Four numbers decide it, and you can get all four in an afternoon. (25 min)
- Where the arithmetic stops working — Every model in this course has conditions. Knowing them is the difference between understanding and repeating. (20 min)
Related writing
- What your machine actually costs to run — A £15 plug meter settles the energy question for your own hardware in an afternoon: idle draw, load draw, watt-hours per task, and what that comes to per year.