A version of this article first appeared in the CNBC Property Play newsletter with Diana Orrick. Property Play covers new and evolving opportunities for real estate investors, from individuals to venture capitalists, private equity funds, family offices, institutional investors and large publicly traded companies. Sign up to receive future editions directly to your inbox. A home has enough value that someone is willing to pay for it. This is probably the only truth you can rely on when putting a price tag on real estate. Enter artificial intelligence. Like everything else on earth, AI is disrupting real estate. In March, Ryan Serhant, CEO of the celebrity real estate company of the same name, posted a video on Instagram titled “ChatGPT ruined my $50 million deal.” In the post, he explained how he brokered a real estate transaction, but said, “At the last minute, the seller asked via ChatGPT, ‘Should I sell at this price?'” Perhaps because of the way he asked the question, ChatGPT basically told him, “No, you shouldn’t sell at that price, it’s worth more.” ”The buyer then did the same and asked if they were paying too much for OpenAI’s AI tools, and ChatGPT said yes, they were, he said. “We provided him with comparisons to show why, but there was no context or understanding of the actual property,” Serhant said. Serhant’s post has been viewed more than 3 million times. In a subsequent post, he said he was able to win back deals by explaining to both buyers and sellers the following about AI: “AI doesn’t know the future, it can’t predict the future. It doesn’t know intent, it doesn’t know emotions, it doesn’t know what buyers are cycling through, it doesn’t know out-of-market comparables, it doesn’t fully understand exchange costs, and it doesn’t really optimize for trades,” he said. “AI can model markets, but it can’t model transactions,” said Serhant, who believes AI is an important tool for real estate agents and has even launched its own AI-powered workflow automation platform and operating system called S.MPLE, which he recently spoke on the Property Play podcast. And he’s not alone. For most real estate professionals, the data aggregation capabilities of AI will definitely enhance their expertise, according to Kamini Lane, CEO of Coldwell Banker Realty. “Market analysis, comparative analysis, these are important tools in a real estate agent’s toolbox. But the important thing is that they are the starting point for real estate agents to apply their judgment, expertise, and nuanced understanding of the real estate market to validate or strengthen the recommendations provided by data tools,” she said. Lane said agents, both buyers and sellers, are increasingly turning to sources like Anthropic’s Claude and OpenAI’s ChatGPT when pricing a home and calculating offers. Like Serhant, she warned about how these generalized, large-scale language models miss the nuances of home, neighborhood, and clientele. “One of the most important things that agents can see, and that ChatGPT and other AI tools don’t, is what’s going to happen in the future: in emerging regions, promising capabilities will be designed,” she said. “Anecdotal data that agents collect through conversations, something that could never be achieved with AI tools that can be collected in the same way as real estate professionals.” Some would argue that Zillow is the original AI pricing model for residential real estate. The company launched its so-called Zestimate feature in 2006, along with the launch of its website. The company recently launched an “AI mode” designed to guide homebuyers by learning their specific needs. This allows homebuyers to have a more personalized conversation with Zestimate. “AI guidance for consumers needs to be tied to real-world situations, real data, and real-world ability to act,” said Nicholas Stevens, vice president of products and AI at Zillow. “In that case, the AI guidance has to be deeply connected to what the real estate agent is trying to do. That’s the difference between what we do at Zillow and something like a typical third-party experience.” Agents must upload detailed floor plans and 3D visual captures of the entire home and surrounding grounds, along with all the information. Zillow then uses AI mode to advise buyers on what offers are good. “You’ll actually see remodeled kitchens. You’ll actually see upgrades to the home. This is helpful for both buyers and homeowners who are looking to sell or renovate,” Stevens said. Zillow’s AI capabilities are currently primarily aimed at buyers, but Stevens said the company plans to roll out tools for sellers as well. However, the accuracy of AI itself that attempts to understand human users remains questionable. Coldwell Bankers Lane said he is concerned that for both buyers and sellers, AI may not be able to recognize what they may need compared to what they claim they want. They also may not want to provide advice that is difficult to hear like a human agent. “Artificial intelligence is trained to be sycophantic, it’s trained to give you the answers you want, so you stay engaged. So the AI is more likely to give you the price you want, not the sale price of the home,” Lane said.
