Is ChatGPT Better Than You (Me)?

If you’re a business professional—accountant, lawyer, engineer, or even (gulp) an office broker—you’ve probably wondered: Can AI do my job better than me?
 
As a broker who advises office tenants, I tested ChatGPT in a domain I know well. I asked it for average Class A office rents in downtown San Francisco over the last 30 years, including supply and demand dynamics. In less than a minute, it delivered a surprisingly accurate answer.
 
That’s the power of large language models (LLMs). When trained on enough data—and prompted well—they generate insights quickly. Much of the market data I use comes from private sources, but a lot of it ends up in public research reports. That means LLMs are catching up fast.
 
But raw insight is only part of the equation.
 
I then gave ChatGPT a real-world scenario: a high-credit tenant, recent anchor vacancies, a distressed landlord with a maturing loan. I asked for a negotiation strategy—and it offered a basic playbook that aligned with the one I’d pursue. That was eye-opening.
 
LLMs are closing the gap on the data and insight side of brokerage. Where they still fall short is in execution. Strategy isn’t just about having data—it’s knowing how to use it to apply pressure, to read the room, to manage nuance. That’s the art of negotiation. And for now, it’s still human.