How do I use AI for sourcing without getting burned again?

You asked the AI for a number or a name, and it sounded right, so you moved on it. Then it wasn't right, and the cost landed on you, not the machine. State that plainly, without dressing it up. You would never let a new hire make a sourcing call without checking their work first. But you let this tool skip that step, because it talks like it knows. It doesn't know. It predicts. The gap between knowing and predicting is exactly where you got burned. This is not a reason to stop using it. It is a reason to put back the check you dropped. The tool drafts. You verify. That division of labor was always yours to keep, and it still is.

◆ How this problem reads on the two dials

GuidanceKnowledge

More coaching

A little to learn

1:1 with AureliusWith others (a Pod)

Mostly you & the coach

A little with peers

The core failure is behavioral — a dropped habit, not a knowledge gap — so the page should coach the verification discipline back into place more than it teaches new concepts.

How the two dials adapt to you →

What’s really going on

The AI did not burn you. You burned yourself by skipping the check you'd run on any new analyst. Fix that habit, not your trust in the tool. Choose a rule now: no recommendation moves until you verify it yourself. That is the whole answer.

🔒 What you’ll build togetherUnlock by starting

A moveBefore acting on any AI recommendation, name one outside source that could confirm or kill it — a contract, a report, a second system — and check it.

A moveWrite one sentence stating what would make this recommendation wrong. If you can't name it, you don't understand it well enough to use it.

A moveSet a dollar threshold. Nothing above it moves without your own sign-off, in writing, after a check.

A moveTreat every AI output as a draft from a new analyst — fast and useful, but unverified until you verify it.

A moveAfter any mistake, write down exactly what you skipped checking. Not what the AI got wrong — what you skipped.

What changes unlock by starting

You catch a bad recommendation before it costs you, not after.
You keep the speed of using AI without carrying its risk alone.
You build a verification habit simple enough to explain to anyone who asks how you work.
You stop blaming the tool and start owning the process, which is the only part you actually control.

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.