When AI makes coworkers less necessary, who keeps the team talking?
When AI makes coworkers less necessary, who keeps the team talking?
A question that once went to a colleague now goes to an AI assistant. The answer comes faster, and nobody has to admit they did not know. The task gets done. What quietly disappears is the conversation around it, and with it much of how people in a team learn who knows what.
What the article argues
In a new paper in Social Science and Management, Jen-Chieh Juan Lu calls this a fall in the “relational dependence” of knowledge work. The author argues that:
- AI replaces some reasons to go to a colleague more than others. Information, reasoning, feedback and wording move to AI first. Coordination, authority and high-stakes decisions still need people.
- People turn to AI most where asking a person feels costly: the fear of looking incompetent, not wanting to interrupt, a steep hierarchy.
- The task gets faster, but the side effects of those conversations fade: knowing who can do what, weak ties across teams, informal mentoring.
- The management job is not to save every old conversation, but to decide which ones matter and build them into work on purpose.
Why this matters to us
On cross-functional work the paper is direct: “Cross-functional projects may need deliberate sensemaking sessions.” When everyone works with their own AI, shared understanding no longer happens by accident. Someone has to make room for it. That is what the Sensemaking Club is for: a small group of innovation leaders from different companies, comparing real cases around one table.
A question for you
Which conversations in your team has AI quietly replaced, and which of them do you miss?
Source