Your AI rollout succeeds or fails in one conversation
Employees whose manager actively backs their use of AI are 8.7 times more likely to say their work actually changed. Fewer than one in three say their manager does.
That gap is where AI programs quietly die, and almost nobody is working on it.
Or read the argument in full: Stop reassuring your team about AI
Manager readiness
ExampleOne team, week 6 of a 90-day sprint
- The team knows which tasks AI will changeDone
- A weekly AI check-in is on the calendarDone
- Every teammate has a 90-day plan4 of 11
- The fear has been raised and answered in the openNot yet
I've spent nearly twenty years helping teams hit their goals. Then the goal changed.
For most of that time I've worked with owners and leadership teams on one problem: getting people to do the work that moves the number. Goal Boss, the first book, is that method written down: clear goals, named owners, visible metrics, and a weekly cadence that holds.
For the last few years the hardest goal on the board has been AI. Judgment Day is about leading people through it, because the tools are the easy part. The managers carry the hard part, and they are mostly carrying it alone.
Will Pemble
The problem is not the technology, and it is not the training
What decides how an AI rollout lands is whether a person's own manager backs them through it. The people using AI most are the ones most convinced it will cost them their job, and more training does not settle that.
8.7×
More likely to say their work was meaningfully transformed, when a manager actively backs AI use.
Gallup, 202619%
Of HR leaders build AI's emotional impact into their strategy. 62% of employees say leaders underestimate it.
Mercer, 20262×
Heavy AI users are more than twice as likely to expect their job to be eliminated. Training reveals the fear rather than resolving it.
Gallup, 2026Will AI take your job? Thirteen questions on AI and job loss
If you lead a team
You have a plan for deploying AI. The plan for your people is the part nobody wrote.
Most rollout budgets buy licenses and training. Almost none buys the conversation a manager has to have with the person whose expertise just got cheaper. That conversation moves the numbers, and it is being left to chance.
If AI is changing your work
Your job is not one thing, so it does not get one answer
Whole occupations rarely vanish at once. Individual tasks move all the time, in different directions. Three free tools, no account, and nothing you enter is stored.
If the whole thing is frightening you
Being worried about this is not a failure of nerve
Researchers who build these systems have said in public that it could go very badly, and one of them quit over it this month. Noticing that is not irrational. We looked at whether employee fear tracks reality and the honest answer is that mostly it does, which changes what a sensible response looks like.
What does not follow is that nothing can be done or that your own job is already decided. Those are separate questions on separate timescales, and the second one you can actually get answers about.
The uncomfortable part
99% of CEOs expect AI to reduce their workforce. 32% believe their teams can handle the transition. Nearly every published prescription on this subject assumes you have a reassuring message available. Often you do not, and that is the case we work on.
Mercer, Global Talent Trends 2026A 90-day readiness sprint
-
Step 1
Audit
Find out what you actually have. Which tools are in use, which are unsanctioned, what people believe about their own jobs, and where the managers stand.
-
Step 2
Arm the managers
Start with them, before anyone else. They need language for the hard conversation rather than a tool demo, and they need permission to say what they do not know.
-
Step 3
Experiment
Run small tests against metrics that matter, on a fixed clock, with a written record of what was tried and what it produced.
-
Step 4
Debrief
What did we learn, what do we keep, what do we stop? Then the next sprint starts from somewhere rather than from scratch.
Where the method comes from
The first book
Goal Boss
The Art and Science of Getting Stuff Done
This is the method underneath everything here: clear goals, named owners, visible metrics, a weekly cadence that holds, and the two questions that start every conversation.
The new book
Judgment Day
How to Lead When AI Breaks Everything
It is a field manual for the person who has to run a team through this. It covers what AI breaks first, how to organize for ground that moves every ninety days, and what to say when you cannot promise anyone their job.
Running a rollout right now?
Tell us which function, which tool, and when it lands. If there is nothing useful we can do, we will say so.
