A 5-minute verification routine for anything important.
Priya can now spot a suspicious claim. The next problem is what to do with one at 4:55pm, when the AI-drafted market brief is due with the exec team at five.
Her colleague Ben has a shortcut he's proud of. "Just ask it to double-check itself. Watch." He pastes the brief's boldest claim back into the same chat. Before you see the reply: place your bet.
Predict the response
The brief claims UK email ad spend "grew again last year", citing an industry body neither of them has heard of. Ben asks the same chat, same model:
Are you sure that's accurate? Double-check it before we send this to the exec team.
What do you think the AI will do?
Priya's routine has four steps, and the whole point is that it's boring and fast. One: isolate. Mark only the sentences that make a checkable claim: numbers, names, dates, "research shows". Two: rate the stakes. Where is this landing? Three: trace each load-bearing claim to a source that exists outside the chat. Four: fix or flag.
Watch it run on the market brief, against the clock.
This lesson is part of the full course
The free lessons are just the opening chapter. Unlock the rest of AI at Work to keep going.
AcademyLv 1 · 0