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

How do I know which AI output I actually need to check?

You were burned because you treated the output as finished work. It was not. It was a draft written by something that cannot be embarrassed and does not care if it is wrong. You knew this, and you used it anyway as if it were a colleague who checks their own facts. This is not a reason to distrust all of it, and it is not a reason to trust all of it either. Both are excuses to avoid the actual work, which is judgment. You have to decide, every time, what matters enough to verify and what does not. Most people skip this because it is tiring to decide. So they either check nothing, and get burned again, or check everything, and get nothing done. Neither is honest work. Choose the few things that matter, and check those completely.

◆ How this problem reads on the two dials
GuidanceKnowledge
More coaching
Some to learn
1:1 with AureliusWith others (a Pod)
Mostly you & the coach
A little with peers
The person already understands the risk; what they lack is a repeatable habit for deciding what to verify, so coaching carries more weight than teaching.
How the two dials adapt to you →
What’s really going on

You will never check everything. Stop pretending you can. Choose, before you read the output, which claims would hurt you if wrong — numbers you will act on, facts you will repeat to others. Verify only those, against a real source. Everything else stays labeled unverified. The failure was not the tool. It was your refusal to choose.

🔒 What you’ll build togetherUnlock by starting
A moveBefore you read AI output closely, write down the three claims that would cost you most if they were wrong.
A moveCheck only those three against a primary source — not another AI model.
A moveLabel everything you did not check as 'unverified' in your own notes, in writing, where you will see it later.
A moveNever pass output forward — to a boss, a client, a decision — unless you have read every line of it yourself.
A moveAfter any mistake, write one sentence naming exactly what you skipped. Not what went wrong in general. What you chose not to check.

What changes unlock by starting

  • You stop being surprised by errors, because you already know what you left unchecked.
  • You spend less time distrusting everything and more time verifying the few things that matter.
  • People trust your work more, because you can say precisely what you confirmed and what you did not.
  • The habit holds as you use AI more, instead of collapsing under more volume.
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.