Talk it through with Aurelius
Library›Aurelius›The problem
Aurelius · Work & Leadership
Knowledge + Guidance

We use AI to model decisions—but who checks its hidden assumptions?

You did not fear this tool when you adopted it. You feared slowness, and the tool cured that. Now it decides things quietly, and you call your unease 'worry about assumptions.' Name it plainly: your team accepted a judgment it never inspected. This is not a technology problem shared by everyone who uses AI. It is your team's problem, made by your team's choices. Someone chose speed over inspection. Someone chose to trust output because it arrived formatted and confident. No one voted on this, but everyone agreed to it by staying silent. A model hides nothing on purpose. It simply does not know what it has left out, and neither do you, until you ask together. The gap is not in the machine. The gap is in your pod's habit of asking.

◆ How this problem reads on the two dials
GuidanceKnowledge
More coaching
Some to learn
1:1 with AureliusWith others (a Pod)
Some one-to-one
Practise with peers
The team needs a working method for surfacing hidden assumptions, but the deeper shift is owning authority they quietly handed away — that requires coaching, not just instruction.
How the two dials adapt to you →
What’s really going on

The model is not the danger. Your silence is. You have let a tool make choices for your team that no one has named out loud. Gather your people. List the assumptions. Assign one person to challenge each. That is the whole remedy.

🔒 What you’ll build togetherUnlock by starting
A moveBefore the next model run, have each person write down one assumption they think it's making — compare lists out loud.
A moveAssign a rotating role: one person's only job each week is to argue against the model's output.
A movePick the last decision your team made using the model's output. Trace it back and name the three assumptions buried inside it.
A moveStop saying 'the model suggests.' Say 'we chose to follow the model's suggestion.' Make the choosing visible.
A moveSet a standing five minutes at the start of every model review: 'What did we not check this time?'
PractiseThe Hidden Assumption Audit · a Pod of 4 · 30 min

What changes unlock by starting

  • Your team can name, in plain language, the assumptions behind its last three model-based decisions.
  • No one on the pod defers to the model's output without first stating what it might be missing.
  • You have a repeatable habit — not a one-time audit — for surfacing hidden reasoning before it becomes a decision.
  • Disagreement with the model becomes normal team behavior, not a sign of distrust.
One object, two jobs: a public answer to a real problem, and — the moment you start the chat — Aurelius’s live plan for your version of it.