On AI
The work is what we already do. Psychology-based culture development with senior leadership. Reading the unspoken dynamic. Holding the room when stakes are high. Naming what nobody else will say. None of that is something a language model is trained on. None of it will be.
Most companies bring in change management when AI is already failing. By then the speak-up culture is already gone. We come in earlier, applying the same work at the moment AI is being introduced, not after the rollout is losing momentum. Either way, the variable is your culture.
The shift in consulting
For four decades, the Big Four sold analysis. Reports, benchmarks, frameworks, board packs, transformation playbooks. The output was the deliverable. The deliverable was the value.
That output is now generated by AI systems at 10 to 100 times the speed, at a fraction of the cost, with output quality that improves every quarter. Most of the analytical work that justified a Big Four engagement no longer requires a Big Four engagement.
Whatever survives will not be a faster version of what was. It will be a different category of work entirely.
AI/01 · text-domain
Analysis at scale.
Synthesize a thousand interviews, benchmark against industry data, generate a 60-slide deck in twenty minutes. Faster, cheaper, and increasingly indistinguishable from what an associate would produce.
AI/02 · text-domain
Pattern recognition across documents.
Spot inconsistencies across a corporate archive in seconds. Cluster themes from years of board minutes. Surface what was always there but nobody had time to read.
AI/03 · text-domain
Generate a framework.
Write a methodology. Produce a maturity model. Draft an OKR template. Frameworks are abundant. Frameworks were already abundant before AI.
AI/04 · text-domain
Deliver a report.
Compile findings, format the executive summary, render the PDF. A deliverable that lives in a folder and changes nothing.
AI/05 · text-domain
Roll out across the org.
Buy seats, generate enablement content, schedule the training. The technical implementation is the easy part.
INS/01 · in the room
Read what is happening in the room.
The pause before someone speaks. The eye contact that does not happen. The conflict everyone is working around. None of this is in the data. All of it is in the room.
INS/02 · in the room
Name the pattern out loud.
Naming what is happening, in front of the people it is happening to, in a way the room can absorb without rupture. That is human work. It cannot be delegated to a model.
INS/03 · in the room
Stay until something moves.
Hold the room after a moment of candor. Process what surfaced. Repair what needs repair. The change happens because we are present while it happens.
INS/04 · in the room
Leave a different team.
The change is not in a deliverable. It is in how people speak to each other on Monday morning. The artifact is the team itself.
INS/05 · in the room
Make the rollout actually take.
The rollout doesn't fail at the tool. It fails at the team culture. Trust ambiguity drains adoption while the dashboard still says "deployed."
The split
Anything that can be put in a deck, AI will write better and faster. Anything that cannot, we do.
What AI does to teams
When AI gives confident answers that turn out wrong, people stop knowing whether to trust their own judgment. They go quiet. They work around the rollout instead of through it. That silence doesn't show up in the dashboard.
Almost a third of employees admit to actively sabotaging their company's AI rollout. Many are not malicious. They are uncertain.
By the time the dashboard says "losing momentum," the speak-up culture is already gone.
48%
of employees hide their AI use from colleagues and managers.
Henley Business School
29%
admit to actively sabotaging their company's AI rollout. 44% among Gen Z.
Writer / Workplace Intelligence
22%
have hesitated to lead an AI project for fear of being blamed if it misfires.
MIT Technology Review Insights
30%
cite fear of AI-driven job displacement. The number nearly doubled in a year.
KPMG, FOBO tracking
· Public studies. Full citations on request.
AI Culture Readiness
Most companies bring in change management when AI is already failing. By then the speak-up culture is already gone. We come in earlier.
Applied at the moment AI is being introduced, not after the rollout is losing momentum.
Leaders practice fallibility about AI before they ask anyone else to. Mistakes get sorted into the right kind (intelligent: testing AI in low-stakes domains) and the wrong kind (basic: skipping verification where limits are known). The team installs concrete rituals. AI After-Action Reviews. A direct channel for "I don't trust this output." Speak-up moments where doubt is welcomed instead of suppressed.
The technical implementation is the easy part. Whether the team trusts each other enough to actually adopt it is what determines whether the investment lands.
Who it's for
Companies mid-rollout where adoption is losing momentum, leaders are picking up sabotage signals, or the team has gone quiet on what they're actually using.
Format
Diagnostic workshop (2-3 days) plus a 6-month advisory cadence. Or as a module inside the existing Year-Long Culture Process.
Pricing band
€50K-150K diagnostic. Year-long extension priced separately. [placeholder pending founder confirmation]
What it includes
- Pre-rollout diagnostic of team trust state and existing speak-up culture
- Workshop with leadership team naming the AI-specific fears and patterns
- Installation of concrete rituals: AI After-Action Reviews, leader fallibility moments, intelligent-failure framing
- Optional 6-month follow-up cadence as the rollout continues
What it is not
- AI tool selection or vendor evaluation
- Technical implementation
- Training on specific AI products
- Change management theatre. No slides, frameworks, or OKRs as deliverables.
If you sign the budget
Anything you can describe as analysis, synthesis, documentation, or framework will be cheaper to run with a model than to commission with a partner.
What does not move to AI is the work that requires a human in the room. Senior team conflict. Stuck decisions. Trust that has broken without a name on it. Cultural patterns nobody in the org will confront.
Less spent on documents that AI can generate, more on the live work that AI cannot replace. The savings on the first side easily fund the second.
Where the market is going
As AI absorbs the analysis, executives are paying premium for the work that requires presence.
Signal · 01
Live coaching demand: up year on year.
Where executives put their hours has shifted. Less reading, more conversation.
Signal · 02
Executive retreat enrolment: rising.
YPO, INSEAD, peer networks. The format senior leadership trusts is in-person, closed-room.
The room AI cannot enter
One call.
Human, on both sides.
The first conversation is 30 minutes, on a call. No deck.