Since ChatGPT launched in 2022, workers have instinctively used AI for the most human parts of their jobs—writing emails, brainstorming, drafting strategy. Meanwhile, they still spend enormous time on mundane work. We've got the division of labor backwards.
The Problem in Numbers
- The average employee spends *~1/3 of the workweek* on tasks they consider "low value" (Slack)
- 42% of frontline AI users save a full workday per week with AI—but more than half don't redirect that time toward strategic work (BCG, 2026)
- Writing remains the most common workplace use of ChatGPT (OpenAI)
Why Creative Work Is the Wrong Entry Point
- Creative quality is inherently subjective. There's no objectively "correct" tagline or strategy memo.
- Two smart people can read the same AI paragraph and disagree whether it's insightful or "AI slop."
- AI can be a useful *copilot* for creative tasks—but it can't be authoritative, so you can't build critical business processes around it.
The Better Question
Don't ask: "Can AI do this?"
Ask: "If AI does this, how will we know whether it did it correctly?"
- Verifiable output → strong candidate for genuine AI transformation (reconciling spreadsheets, reviewing contracts, updating CRMs, billing, collections)
- Unverifiable output → keep humans in the lead
Where Humans Win
When decisions are strategic, contextual, or dependent on taste, humans have the upper hand. Conveniently, that's also the work people actually *want* to do.
The companies that get the most value from AI in the next few years won't ask it to replicate high-level thinking. They'll **redesign mundane workflows around verifiability**—letting AI handle the work we can check, so people have more time for the work only they can do.