By skyforbes 
A few months ago, I asked ChatGPT how to remove a carrot stain.
That was it — one short sentence: “How do I clean a carrot stain?”
The answer was long, confident, a bit too cheerful, and absolutely useless.
It suggested lemon juice, salt, vinegar, baking soda, hot water, detergent… a full chemistry set.
But who on earth would put vinegar in their baby’s mouth?!
That’s when I realised what I’d done. I hadn’t said from what.
Plastic container? Baby bib? Cashmere sweater?
The AI wasn’t wrong.
It was doing exactly what I told it to do — which, in hindsight, was almost nothing.
And I sat there thinking: This is how I talk to my team sometimes.
The invisible context problem
When you manage people for a while, your brain becomes a suitcase full of half-folded context.
You know why a decision was made, who you’ve already told, what’s urgent, what’s political. You stop noticing how much of it lives only in your head.
As a reminder, you are alone in your head.
Then you write a Slack message like: