Description: Finish early and get more work. Suggest AI and hear “security says no.” How workplaces can teach their best problem-solvers to stay quiet.
“We could automate this step and save hours.”
“Fine. What exactly will you do with those hours?”
As a planning question, that's perfectly fair. But if the subtext is “So you want to slack off?”, something has gone wrong.
Let's abolish email, then. Write every message by hand, walk it to each recipient, and read it aloud. Everyone will look incredibly busy. Actual results may have to wait.
1. Motivation can fade before the real job begins
One new employee starts with genuine excitement. Training sessions repeat the same material and different instructors give poorly coordinated explanations. By the time the actual assignment begins, expectations have already fallen.
Then come duplicated entries in spreadsheets, documents sent back for endless revisions, and suggestions that struggle to get approved.
The sequence matters. The initial decline in motivation came before any improvement idea was rejected. Later experiences merely reinforced a sense that the organization treated people's time carelessly. Saying “the employee lacks motivation” skips the mechanism worth examining.
2. The strange reward for working faster: more work
Productivity isn't measured by how exhausted someone looks. The OECD describes labour productivity as output per hour worked.[1]
Saved time can go toward better customer service, error prevention, learning, further improvements or fewer late evenings. A manager can reasonably ask how it will be allocated.
But when every saved hour attracts fresh tasks with no workload ceiling, staff learn a different lesson: “Showing my ability makes my life harder.” So they stop advertising efficiency. The company shouts about productivity while quietly teaching people how to hide it.
3. Security should be a design constraint, not a conversation killer
Someone proposes AI to help prepare reports. The answer is “Information security. Completely prohibited.”
The risks are real. Even a language model running inside the office needs access controls, careful data handling, logs, license checks, updates and scrutiny of inaccurate outputs. Local deployment is not a magic safety certificate.
Yet options exist between reckless deployment and indefinite prohibition: synthetic data, a tightly scoped task, documented risks, approval and stop criteria, and a small controlled pilot. NIST's AI risk guidance encourages managing risks in context.[2]
A well-reasoned decision not to deploy may be responsible. Refusing to investigate safer approaches is a different decision.
4. How capable people are taught to keep quiet
Improvement proposals go nowhere. Finish early and another assignment arrives. Nobody protects the upper limit of workload. Management still asks for “more ownership.”
Research on employee voice and silence examines why people share useful concerns or withhold them; silence can cost an organization valuable information.[3] Research on effort–reward imbalance also connects persistently lopsided work exchanges with health risks.[4] These findings cannot establish what caused one company's problems, but they invite a closer look at the incentives.
5. Type faster—or ask why we're typing at all?
“How can we fill in the spreadsheet faster?” is different from “Why do we fill it in again and again?”
Organizational learning research distinguishes fixing errors within existing rules from reconsidering the assumptions behind those rules.[5] Some procedures serve legitimate needs, including safety, audits, law and cost. Strong improvement ideas recognize those constraints and look for a safe experiment.
6. The hidden bill is paid in years of human life
After years at work, someone may be excellent at complicated internal procedures without building skills transferable elsewhere. A job can consume time while giving little extra freedom for the future.
So don't only ask about salary and prestige. Ask whether improvements are tested, whether workload has boundaries, and whether the skills gained will matter beyond this employer.
7. What a workplace serious about productivity would do
Measure time and quality before change. Define safe conditions for a trial. Start small and count errors and maintenance alongside minutes saved. Agree where saved time goes, set workload limits, and share the rewards of improvements with those who make them.
Then “What will you do with the time you save?” becomes a constructive question.
The point: Before suspecting someone who finishes early, inspect the system that wastes their time. And if efficiency is truly offensive, deliver tomorrow's office announcements by handwritten letter. Just let the messenger go home on time.
