There is a particular kind of anxiety that comes from feeling you are supposed to understand something before you have been given time to learn it. AI has brought that feeling into a great many workplaces. The announcements arrive quickly. Confidence takes longer.
I think the most useful response is to begin at a manageable scale. Find one piece of work you understand well. Learn where an approved AI tool can help, notice where it struggles and keep enough of your own attention in the process to tell the difference.
That is a modest beginning. It is also a much stronger position than either avoiding the technology or assuming it can take responsibility for you.
Your experience still has somewhere to go
McKinsey’s discussion of worker adaptation makes a useful distinction between automating tasks and replacing an entire job. Its research suggests that many skills remain relevant even as the mix of work changes. These are findings and projections about the labour market, not a guarantee that every role will stay the same.
For an individual worker, the practical question becomes: which parts of my experience help me direct and evaluate this tool? A person who understands the customer, the process or the consequences of an error brings context that a fluent answer can easily lack.
Consider preparing a routine report. AI may help organise a first draft. Someone still has to decide whether the question is sensible, whether the figures belong together and whether the conclusion deserves to be acted on. Experience becomes more valuable when it is used to make those judgments explicit.
Build a small practice you can repeat
I would set aside a regular period for learning rather than waiting for an urgent assignment. Choose a low-risk task. Use information your organisation permits you to use. Compare the assisted result with the standard you would normally apply. Keep a short note of what worked and what needed correction.
Over time, those notes become something more useful than a collection of clever prompts. They become an understanding of the work itself. You begin to see which steps are repetitive, where important context gets lost and where a human conversation would resolve the issue faster.
Bernard Marr’s report on Microsoft’s workforce transformation describes attention to workflows and operating models, rather than treating AI as a simple software installation. That seems like the right level of ambition: improve the way the work is done, then judge the result.
Give newer people a way to grow
The opportunity is not only for experienced workers. Employers can use this transition to make learning more deliberate. If an introductory task becomes automated, a junior colleague still needs a way to develop the judgment that the task once helped teach.
The World Economic Forum’s entry-level work discussion argues for redesigning career pathways as AI changes the work. My own preference would be to give newer people supervised responsibility: let them question an output, explain a decision and receive feedback from someone who understands the stakes.
A manager can make the change less intimidating by admitting that learning is part of the job. Clear expectations, patient coaching and permission to ask an ordinary question will usually do more than a confident speech about transformation.
From Toronto, I see this as a constructive challenge for workers and employers alike. We have useful tools to learn and work worth improving. We do not need perfect certainty before taking the first step. We need enough care to learn from that step and enough responsibility to remain answerable for where it leads.