A recent New York Times story about a homeowner who used ChatGPT to sell his house sparked a lot of discussion about whether AI can replace real estate agents.
The answer, surprisingly, was yes—at least for a big part of the job. AI helped write the listing, answer buyer questions, organize paperwork, and guide the seller through the process all the way to closing, where he finally hired a real estate attorney. For someone willing to do the legwork, AI proved to be a capable assistant and might have saved the seller a substantial commission.
But what it couldn’t do was the work that comes from years of experience. It couldn’t walk through the house and say, “Replace the carpet in this room. It’ll cost you five thousand dollars, but you can ask for fifteen thousand more.” It couldn’t identify which features buyers in that neighborhood care about most, recommend how to stage the house to create a stronger first impression, or use local market knowledge to position the property relative to competing listings.
As AI takes over more predictable, routine work, expertise becomes even more valuable because the hardest problems have never been about following a checklist. That’s the pattern we’re seeing across industries: AI is remarkably good at making expertise more accessible, but it’s still the human expert who knows when to break the rules, spot an opportunity, or make the judgment call that turns a good outcome into a great one.
Want to learn more? Check out our podcast: Episode #24: AI and Humans at Work
(art by Becka Rahn)
