Human only, human + AI, or safe to automate?
Getting these ideas on paper helped me decide how AI could augment or replace which tasks and workflows.
By Jon Harmeling First shared on LinkedIn
- Human only: A sincere apology, a termination decision, a coaching conversation, or recovering a guest.
- Human + AI: The human stays responsible while AI improves the work.
- AI + human review: AI handles things like routine reports, organized information, or processing data, and a human verifies, adjusts if needed, but always remains accountable.
- Safe to automate: The work is predictable, low enough in consequence, reversible, and easy enough to verify that AI can do most if not all of it.
This is from my book, Frontline AI Leadership, released October 27, 2026. 100% of net proceeds from every book sale go to the team members I lead.
The takeaway for leaders
Machines do. Managers decide. Leaders develop. Before you hand a task to AI, decide which of the four buckets it belongs in, and who stays accountable.
This idea is at the heart of the Frontline AI Leadership keynote and the book Frontline AI Leadership.
Read the original post and the conversation on LinkedIn · Follow Jon on LinkedIn