A TOPHEADS career assessment
20 questions, about 7 minutes. You will see how exposed your role is: whether AI could take over the work, whether the people around you could absorb the role, and what protects it. Answer by what has actually happened, not by what you plan to do.
It looks at your role the way a company does when it decides what to cut with AI. At the end you get one of five levels and see what exposes your role and what protects it.
Six things: how much of your week is work AI can’t do, whether the people around you could absorb your role, whether it produces a result the company counts, what trust rests on you, whether you lead the change with AI, and what is already moving in your company.
Mike Torchinsky’s 25 years in executive search, plus published research: AI helps most with information work such as gathering, writing and passing on (Microsoft Research, 2025); among young workers, jobs fell where AI replaces tasks, not where it helps people do them (Brynjolfsson, Chandar & Chen, Stanford).
Almost everyone believes their role is safe. So the questions ask what has actually happened: last week, the last ten meetings, the last budget review.
About 7 minutes, 20 questions. Your level is free; the full breakdown opens with your email.
One of five levels, from “In the Line of Fire” to “AI Works for You”, a score for each of the six areas, and up to three things to watch.
It has started. In a Duke University survey of about 750 finance chiefs, published in 2026, 44% expected AI to replace some jobs at their company, and Cloudflare cut a fifth of its staff, mostly middle management, finance, legal and internal audit. Yet a large study in Denmark found no effect larger than 2% on pay or hours two years after ChatGPT (Humlum & Vestergaard). So the quiz shows how exposed a role is, not whether it will be cut.
Start with the weakest of the three things a cut decision looks at first. Move your week toward decisions and people, own a figure the company counts, and lead one change with AI yourself, with a result in numbers.