Working with AI is the same as before
Experts are telling us that managing people who use AI tools requires a shift in how you lead. While there is truth in that perspective, I think it’s more straightforward than the hype suggests.
Understand what the AI can and can’t do. You don’t need to know how to create an AI, but you should understand how the tool works at a basic level, how instructions become output, and when to trust but verify the output.
Know what good looks like. AI can do quantity, but can’t do quality. If you don’t know what good work looks like, you won’t be able to guide or correct the tool. Your expertise becomes more important, not less.
See the whole system, not just one task. AI tools are part of the process. Think about where people do the work, where this automation can fit, and what happens when something breaks. Accountability and hand-offs still matter.
Get comfortable with constant adjustment. AI is moving quickly. That means you’ll need to keep testing, learning, and tweaking. Plan to revisit and refine it, just like the software you already have.
Be precise with instructions. Clear, structured instructions lead to better results. Vague directions require the AI to make judgements with incomplete context and make a statistically significant guess.
Realign roles on purpose. AI is taking the mechanics out of the work. We still need people to make judgements. Remind people that you rely on them for the quality of the outcomes.
This sounds like common sense, probably because it is. The fundamentals haven’t changed. The technology has.