The Statistic That Haunts Me
I read a statistic to Victoria yesterday that stopped me in my tracks. A recent study found a massive discrepancy in how AI is impacting the workday:
- Executives: 40% say AI saves them 8+ hours a week.
- Staff: 66% say AI saves them <2 hours (or nothing) a week.
How can two groups using the same technology experience such a different reality?
The "First Draft" vs. "Final Product" Problem
The discrepancy exists because of the nature of the work.
Executives operate in the world of Strategy. They use AI to brainstorm, summarize reports, or draft vision statements. For these tasks, "80% correct" is a triumph. If the AI misses a nuance in a summary, the Executive still saved 2 hours of reading.
Staff operate in the world of Execution. They are responsible for the final code, the accurate data entry, or the customer response. For them, "80% correct" is a liability.
- If an AI writes code that is 90% right, the developer spends more time debugging the 10% error than if they had just written it from scratch.
The Janitor Effect
This creates a dangerous dynamic I call the Janitor Effect.
The Executive feels productive because they are delegating the "Thinking" to the AI.
"Give me a 5-point marketing plan." (Time: 30 seconds).
They then hand that plan to the Staff. The Staff realizes the plan is vague, hallucinatory, or impractical. They spend 3 days verifying and fixing the AI's "ideas."
- The Executive thinks: "Wow, AI is fast!"
- The Staff thinks: "AI just created busywork for me."
The Illusion of Speed
We must be careful not to confuse generating work with doing work.
If I use AI to generate 100 blog post ideas in a minute, I feel efficient. But I have just burdened my team with the task of validating 100 ideas. I haven't saved time; I have just shifted the "Time Tax" downstream.
The Protocol: True efficiency happens when the AI does the execution, not just the ideation. If your AI usage is making you feel like a Wizard but making your team feel like Janitors, you are using it wrong.
