Output identity
O = H × P
Weekly output equals human hours multiplied by productivity per hour.
The Saved Hours Doctrine · Working paper
When AI saves hours, who gets them back?
AI and automation promise to save time at work. But saved time does not automatically go anywhere useful. It can disappear before anyone names what was saved.
Make saved time visible.
The minimal model
The gain appears before the distribution decision.
Hold output and quality constant. Count all new supervision, verification and correction work. Then ask what happened to the difference.
Saved hours
S = H0 − H1
Old human hours minus new human hours, for the same output and quality.
Break-even multiplier
P* = H0 ÷ H1
A move from 37 to 20 hours requires 1.85× productivity. Anything above that threshold is surplus.
The Saved Hours Doctrine
If technology saves human work time, those hours should be visible, measurable and negotiable.
The existence of saved hours is technical. The distribution of saved hours is political.
Six destinations
A saved hour does not stay empty.
Once the gain exists, it is allocated. Sometimes that choice is negotiated. Sometimes it happens silently through budgets, targets and systems.
Worker captures time
Shorter required working time with stable income and output.
Firm captures output
The same hours produce more goods, services, cases or documents.
Capital captures profit
Lower labour costs, higher margins, valuation or shareholder return.
Consumer captures prices
Lower prices, wider access or a better service for the same cost.
State captures capacity
More public service, administration or institutional capability.
The metric captures it
The output definition expands. The gain becomes a new target before anyone can claim the time.
The least visible destination
The pauses inside the job disappear.
The system captures intensity. The worker keeps the job. The worker keeps the wage. But every saved minute is filled with more demand, more monitoring or a denser day.
AI policy measures jobs lost and GDP gained. It does not measure hours saved, hours captured or hours returned. Without that measurement, saved time is allocated by default. Defaults are rarely neutral.
Current research baseline
The Saved Hours Doctrine
The working paper develops the measurement model, audit protocol and capture framework behind this page. V0.94 is a public review draft, not a peer-reviewed empirical result.
Project status
Foundation stage
Saved hours must be visible before they can be negotiated.
This page is the web form of the first physical Saved Hours flyer. It explains the public question and points to the frozen paper baseline.
No survey or consultation is open here. Hivemind / DeepAsk v2 is being developed separately and will connect only after its participation, privacy and evidence boundaries are stable.
Corrections or substantive comments can be sent to contact@hedegreenresearch.com.