ACTIVE  ·  BUILDING  ·  v2026.11.20 BUILD 2026-11-20  ·  JL:IOTA:001
No. 060 · 2026-06-11

The Botsitting Tax

AI saves 11 hours a week but costs 6.4 in supervision — and the fix starts with the workflow.

DISPATCH  ·  LOGGED WITH MAI

A survey of 6,000 workers landed this week with a number that should stop every AI rollout conversation cold. AI saves about 11 hours a week per person. It also costs 6.4 hours in supervision, context-feeding, and cleanup. Researchers from Stanford and Berkeley coined a word for it: botsitting.

The net gain is real but thin. The cause is the deployment pattern. Most companies bought tools, handed them to people, and called it transformation. Nobody redesigned the work. Workers are repeating prompts across disconnected systems, checking outputs that look right but might not be, and debugging errors that surface three steps after the AI made them.

Sixty-nine percent of workers admitted to shipping AI output they never verified. The supervision load overwhelmed them. When you spend hours feeding context into tools that do not talk to each other, you stop checking. The volume outruns the oversight.

The companies getting results did one thing differently. They treated AI as a reason to redesign the workflow. Fewer tools. Shared context. Clear standards for what goes to a model and what does not.

The 6.4-hour tax is an operations problem.

LOGGED WITH MAI  ·  2026-06-11  ·  No. 060
← All Dispatches