Where to Draw the Line Before AI Reaches Professional Users
Patrick Neeman · 22 October 2026
Read the originalPatrick Neeman argues that chat boxes slow down expert users of business software, and that routine work belongs in screens, rare questions in chat and monitoring in AI agents.
Patrick Neeman, a user experience designer who has spent decades observing expert users of business software, sees a chat box on nearly every business application's roadmap. Claims adjusters, lawyers and supply planners already know their screens and filters, and a prompt makes them describe that knowledge in sentences. He offers five questions for deciding where a conversational feature belongs.
- Chat helps people who do not know a system, because it removes the need to remember menus, but it adds a translation step for experts.
- A filtered list shows the count and the outliers around an answer, which a chat reply does not.
- He cites a study in which purpose-built legal research tools gave false answers between 17 and 33 per cent of the time, so associates still check every citation.
- His ticket test compares chat with the alternative: a request to an analyst that takes days, or a click the user already makes from habit.
Product and design teams can sort each proposed AI feature by whether the task is routine, rare or continuous before building a chat surface. Neeman also recommends keeping the old path beside any new one and tracking which one users choose.
Keep the existing interface beside any new chat feature and track which path users choose.
Collect 90 days of requests that customers sent to analysts and use them as the list of questions a chat feature should answer.
Derived by Working Surface from the article; more in the Playbook. Source line: In professional work an answer isn’t finished when it appears.
24 October 2026: teams keep the time AI saves only when they decide in advance what an AI agent may do alone, what waits for a person and who maintains what it builds.