The fear was specific. The rollout was not.
Meta description: Accurate fear deserves accurate information: what the work is, how it will change, and where human expertise stays necessary.
Miklos Roth
The fear arrives in leadership’s office as a soft problem: morale, adoption resistance, a culture issue for the CHRO to work on. I have sat with the teams themselves, and the fear I find there is not soft. It is specific, well-informed, and often accurate: this task is being automated, that one is next, the career path that started with those tasks now starts nowhere. Leadership responds to a generic fear with a generic reassurance — augmentation, new opportunities, AI as a tool to make jobs better. The team hears the reassurance precisely. It does not address what they are afraid of, and they know it.
[AWAITING MIKLOS — scene: a town hall about “augmentation” and a team that has already seen the task list shrink without a new task that matters]
What do leadership messages assume the fear is?
The messages I read assume the fear is ignorance: people fear AI because they misunderstand it, because the media has alarmed them, because they have not yet experienced how helpful the tools are. The remedy follows from the assumption — demos, success stories, training, exposure. I have watched these remedies deployed on teams whose fear was not ignorance but information. They had read the vendor’s case studies too. They had noticed which tasks the case studies automated first. The assumption of ignorance is comforting to leadership, because ignorance is fixable with communication. Accuracy is not.
What do teams actually protect?
The fear I treat as accurate is the one grounded in visible facts: individual tasks are changing, and people can see it. Executives hear promises of scale and lower costs; employees see their tasks shifting and worry that their skills will no longer be needed. Neither perspective, on its own, describes the entire operating system. The response is to make the actual conditions and boundaries visible — what the work is, how it will change, where human expertise remains necessary, and what evidence will determine whether the new process is helping. That information is more useful than either extravagant promises or vague reassurance.
What I observe is that resistance, examined closely, is rarely about the technology. Teams protect things the rollout plan never enumerated: the judgment call that makes the work theirs, the relationship with the client that the task was an excuse to maintain, the expertise that took a decade to build and that the tool now performs adequately in seconds. The rollout plan counted tasks. The team counts meaning. When leadership reads protection as resistance — an obstacle to adoption — it mislabels the most informative signal the organization is sending about what the work actually consists of.
Why does reassurance widen the gap?
Read more: more about Miklos Roth.
The false explanation is messaging quality: the reassurance needs to be more empathetic, more specific, delivered by more trusted voices. I have watched reassurance improve in craft and widen the gap anyway, because reassurance makes a promise leadership has not verified. “Your job will change, not disappear” is a forecast, and the team knows who is making it and on what evidence. Each reassurance cycle raises the stakes: when the promised augmentation does not arrive — when the task list shrinks and the meaningful work does not grow — the next reassurance is discounted further. Trust is being spent on a forecast.
What pattern marks fear that is rational to the workflow?
The marker I look for is precision. Irrational fear is diffuse — a general unease, a resistance to everything. Rational fear is surgical: the team can name which task, which quarter, which client’s work is affected, and their account matches the vendor’s roadmap better than leadership’s does. When I find a team whose fear is precise and verified, I stop reading resistance and start reading intelligence. The team has done the analysis leadership has not. The rollout plan would be better if it had access to what the team knows — and the reassurance campaign is precisely what prevents that knowledge from traveling upward.
One of the strongest barriers to effective adoption is the belief that mistakes are unacceptable. Organizations should distinguish between an error discovered during supervised learning and an uncontrolled error released into production. The former can be valuable evidence; the latter can cause real harm. A company that treats every early mistake as proof that AI is dangerous will train people to hide their learning instead of share it.
[AWAITING MIKLOS — scene: second scene]
What question has to be answerable before adoption talk?
Adoption conversations ask how to bring people along. The question I find missing is prior and harder: what, honestly, is the plan for the work these people do — not the message about the plan, the plan. In companies where that question has a real answer, the fear conversation changes character. In companies where it does not, adoption talk is a request for trust backed by nothing, and the teams — being intelligent — respond accordingly.
The question a town hall should be able to answer honestly: what will actually change in this team’s work, what will not, and where is the supervised space in which people are allowed to learn?
PR excerpt: [PR — awaiting Miklos material; cut 80–120 words from his paragraphs once available.]