In this issue:
→ How AI changes the risk of being wrong
→ Why people may speak less, not more
→ One audit that surfaces what your team is holding back
AI creates a new kind of pressure.
When answers feel instantly available, and often correct, being wrong starts to feel more visible. More avoidable. More costly.
According to Google’s Project Aristotle research, psychological safety is the number one predictor of high-performing teams.
But here’s what’s changing: the definition of risk is shifting.
The subtle behavioral shift
When the right answer feels accessible, people hesitate to share incomplete ideas, early thinking, and unstructured perspectives.
Not because they lack insight. Because AI has raised the perceived bar for contribution.
And when that happens, thinking moves from out loud to internal. Collaboration becomes filtered instead of exploratory.
The hidden risk
Leaders lose access to raw thinking.
They only hear polished ideas, refined conclusions, post-processed input.
That’s Cognitive Atrophy at the team level. Less visible thinking, less shared intelligence.
The leadership shift
Psychological safety is no longer just about permission to speak.
It’s about permission to think in public again.
That’s a different leadership muscle.
THE UNFINISHED THINKING AUDIT
Five statements about your team, not about you. Score each 1 to 5, where 1 is rarely and 5 is regularly. Three minutes.
Ideas arrive in my meetings already structured, formatted, and complete.
I can’t remember the last time someone said “I’m still working this out” in front of the group.
Written updates from my team read more polished than the people who wrote them sound in conversation.
Disagreement, when it comes, comes privately and afterward rather than in the room.
I rarely hear a half-formed idea that turns into a better one during the discussion.
Your score
5 to 10 — Thinking is still visible. Protect it. This erodes quietly.
11 to 17 — Filtering has started. You are getting conclusions rather than reasoning. The intelligence is still there; it just isn’t reaching you.
18 to 25 — Fully post-processed. You are seeing output, not thinking. Every idea reaching you has already survived a private filter you cannot inspect.
The three moves
Reward unfinished thinking. Say: talk me through your early thoughts.
Model intellectual vulnerability. Share what you’re still figuring out, not just what you know.
Normalize iteration over accuracy. Insight comes from progression, not perfection.
And one thing to say out loud this week: I want the version before you cleaned it up. Say it in a meeting, not in a message. It only works if people watch you mean it.
Question of the week: Where might my team be holding back ideas because they’re not fully formed?
Lead what AI can’t.