The Difference Between Automating Work and Improving Work
Automation makes an existing process faster. Improvement asks whether the process should exist. The two get confused constantly, and the confusion is expensive.
By Al Hutchinson
Automation takes a process as given and makes it faster. Improvement asks whether the process should exist in that shape at all. The distinction sounds academic until you have automated something that should have been removed.
Automation preserves assumptions
Every process encodes decisions about how work should flow — many of them made years ago, under different constraints, by people no longer present. Automating that process preserves those decisions and makes them harder to revisit, because now they are also expensive to change.
A short diagnostic
- Who consumes this output, and what decision does it inform?
- What would break if this step stopped tomorrow?
- Does this step exist to create value, or to compensate for an earlier failure?
- Would a person new to the work invent this step from scratch?
Steps that survive those four questions are worth automating. Steps that do not are worth deleting, and deleting is cheaper.
Why this matters more with AI
AI lowers the cost of automating almost anything, including work that should not be done. When automation is cheap, the discipline of asking whether a process deserves to exist becomes the scarce skill — not the implementation.
More insights
Where Artificial Intelligence Can Actually Help Skilled Nursing Facilities
Most conversations about AI in long-term care start with the technology. A more useful starting point is the parts of the day that consume time without improving care.
Why Healthcare Professionals Need Practical AI Education
Clinical staff are already using these tools. The open question is whether they are using them with any framework for judging when the output can be trusted.