Building a financial model. I give it the business logic, the assumptions, and the structure I want. It builds the model, explains every formula, flags where my assumptions are weak. I had a full working P&L model with sensitivities in under an hour. Something that used to take a team two days.
Deep research. Say you need to understand regulation across 5 markets. Instead of spending 3 days reading PDFs, structure the research as a series of hypotheses and let AI validate, challenge, and source each one.
Learning something new. I’m not an expert in everything. When I need to go deep on a topic fast I set up a dialogue: “Explain X as if I’m a partner at McKinsey who needs to advise a client on this tomorrow. What do I need to know, what are the traps, and what would a specialist challenge me on?”
And my absolute FAVORITE: the expert panel. I ask AI to simulate a panel of 3-4 specialists debating the problem. A CFO, an operations lead, a pricing expert, and a skeptic. Each brings a different lens. The quality of thinking that emerges is something I used to get only in the best partner meetings at McKinsey