AI will not solve all of our real estate problems

This winter, American real estate agents have a plethora of issues to think about. Rising interest rates have created a frozen housing market, and shifting work practices will push office vacancy rates to 55 percent above their pre-pandemic peaks by 2030, according to a new study by Cushman & Wakefield. The prospect of stranded assets is looming.

But if you’re listening to real estate agent chatter, there’s another topic that’s fueling even more discussion: ChatGPT, the generative artificial intelligence-enabled chatbot launched last November by Microsoft-backed OpenAI.

The explosion of this tool has already sparked a vociferous debate about AI’s impact on fields like education, science, and media. However, the real estate world has attracted less investor focus.

This is a mistake. A wave of AI experimentation in this sector is already underway: realtors are using ChatGPT to write listings and social media posts, calculate mortgage payments, search real estate databases, and gather expertise in new areas like agriculture. As these experiments accelerate, they raise a bigger question: Will an AI-equipped broker really need that many human brokers—and brokerage fees—in the future?

Or to put it another way: Is rent-seeking in the real estate sector about to be attacked by robots? Not in the literal sense of participation in leases, but as described by economists: when powerful incumbents manipulate a market by making it opaque or complex in order to rake in high fees.

Brokers themselves do not believe that they will lose their jobs. “A common feeling [in the property sector] is that it will help, rather than replace, real estate professionals,” the Cushman & Wakefield report points out. “Some tasks will be automated [but] Rather, it’s about helping people become really good at what they do.”

Maybe like this. After all, robots cannot perceive the intangible qualities of a building the way humans can. Nor can they read the body language of buyers and sellers and provide the necessary encouragement or reassurance. Given the extreme anxiety associated with many real estate transactions, this is a serious deficiency.

Additionally, the AI ​​tools that are getting so much buzz like ChatGPT don’t always work well without human oversight. To understand the risks, consider an experiment recently conducted by Sarah Bell, an AI specialist specializing in real estate. When she asked ChatGPT to rank the attractiveness of different clients based on nationality, she found that the Python computing code in the bot was deeply biased towards Australians.

Even more alarmingly, it was impossible to pinpoint exactly why Australians didn’t like it because the system was a black box. A real estate agent unwittingly deploying anti-Australian Python code would be “responsible” for violating anti-discrimination laws, Bell notes.

What the industry really needs is not so much AI in the sense of artificial intelligence, but a different kind of AI: augmented intelligence – ways to help people think smarter. The only way to achieve this is to add a third AI, anthropological intelligence, which would infuse the process with a vital cultural context.

While humans are needed to understand AI, it’s not clear that the industry needs so many of them today. After all, over the past decade, the Internet has already introduced new levels of transparency into business: non-professionals can now use real estate websites to value properties, view listings, arrange mortgages, and match sellers and buyers.

Despite this, the number of residential real estate brokers and dealers in America (about 562,000 at last count) has not been declining, but appears to be increasing: the number is expected to increase by another 5 percent over the next decade. Employment in the commercial real estate sector is almost 4 million – and the trend is rising.

The new era of digital transparency hasn’t pushed up brokerage fees either. In 2020, the average commission rate for residential real estate sales was 5.66 percent. That was less than the 6.02 percent level in 1992 – but higher than the level recorded between 2011 and 2018. This is absolutely perverted. Or, as an economist would say, a strong sign of rent-seeking.

Could that change now? Realtors obviously don’t hope. But entrepreneurs see an opportunity for disruption, especially given the other macroeconomic stresses now hitting the market. “Tech entrepreneurs today have a unique advantage when starting a real estate tech company in the AI ​​space [since] Most incumbents have had a challenge in the past two years,” argues Kunal Lunawat, one of those would-be challengers.

Investors — and economists — would do well to watch what happens next, not least as a test case for whether tech innovations can challenge the quest for retirement. Because if robots can make the industry more efficient and lower commission rates, many property owners will be excited. Rightly so: with or without AI, the industry is overdue for reform.

gillian.tett@ft.com