Humans only want to talk to humans when it's bad news.
The line surfaced in a conversation with an ex‑McKinsey partner now running an AI boutique. It explains, neatly, why the senior end of consulting has barely been touched by AI even as the analyst end has been fundamentally re‑shaped.
Draft in progress
What this short note will cover.
- The conversation it came from — and why it landed harder than it should have.
- The asymmetry — clients welcome AI for the easy delivery (good news, status updates, the analytical body of a deck) but recoil from it the moment the message turns difficult.
- What "bad news" actually means in a consulting context — write‑downs, redundancies, missed quarters, "your CTO is the problem", "this acquisition is destroying value".
- Why the model can't do this work — not because it can't generate the right words, but because the audience refuses to receive bad news from anything other than a person whose career is on the line for it.
- The implication for senior operators — the work that survives is the work where someone has to put their name to the call. AI accelerates the analysis; humans still own the verdict.
— Full note coming. Quotable as‑is in the meantime.
"Consulting is still a relationship game. Trust, empathy, walking the client's walk — that's the moat." — Managing Partner, AI boutique; ex‑McKinsey
— Lucas, December 2025