Artificial intelligence in residential real estate: Reality or hype?

Every few years, the vast technology industry gets animated by one concept. This concept need not be brand new – often new packaging suffices to kick in the Silicon Valley hyperbole machine.

In our 25 years in technology, we’ve seen several of these epochs come and go as people train their sites on the shiniest new object or idea.

In 2019, we are indisputably in the era of artificial intelligence. From TV commercials to books, from marketing slicks to sales talk tracks, AI makes an appearance in almost every conversation connected to the technology industry.

Interestingly, it has also appeared as the chief guest in the real estate industry. More on that in a bit.

Artificial intelligence is not a new concept. In fact, it has been discussed for decades, long before the tech sector became the behemoth it is today. Recently, many books, articles and debates on the history of AI have made their way into the mainstream, provoking a series of important debates on issues of applicability, automation, ethics and innovation.

In the world of real estate, for the past five to seven years progressive entrepreneurs have been introducing a variety of technology and digital techniques into what was considered an analog industry. In some ways, this is ironic. After all. residential real estate itself is the world’s largest asset class, reaching $30 trillion in the U.S. alone. On the surface, one would think that it would have been the first to adopt technologies that help maximize profits, given real estate’s enormous size and importance as a bellwether for the entire economy. Still, perceptions can be faulty, as there has indeed been a lag in adoption of advanced technology as the industry shows sign of succumbing to inertia versus embracing innovation.

When asked about this lag, industry experts point to a variety of issues: the large real estate ecosystem, regulation, cost, business models and other factors. In the tech world, however, intent is not the only factor in innovation and adoption. Sometimes industries are pushed into change by fledgling players that very quickly become large as they attempt to disrupt industries that are ripe for disruption.

Some of these real estate disruptors have become household names. The most prominent of these is Zillow, whose exotic name has reached something like Kleenex or Xerox status amongst real estate consumers. There are countless others, as well, with billions of dollars being poured into technology startups that focus on real estate. A new industry has been christened – Proptech.

Real estate-focused tech, or Proptech, is not immune to the hyperbole and exaggeration that characterizes the rest of the sector. A host of ills are associated with this hyperbole including false predictions, premature notices of demise for categories like the real estate agent, false hopes of transparency and a false sense of omniscience. We’ve seen it all.

At the heart of the issue is not simply the usual marketing-speak, but also the lack of understanding of what AI really is. AI is not simply using data, predictive analytics or even number crunching. Yes, they are all part of the AI model, but they hardly describe the sophistication and esoteric nature of real AI.

In addition, AI is not an isolated field that is hermetically sealed from other computational areas. Building an AI model that absorbs advances in mathematics, physics, genetics and other areas enhances its power and applicability.

With real AI, problems that were thought to be fundamentally difficult can now be managed. Advances in the power of computation, in the cost-structure of processing and in the application of mathematical methods to computing have created not only unprecedented scale and speed but also qualitative change in areas with millions of units affected by thousands of variables in a dynamic environment. Unpacking seemingly simple questions like, “What’s a house worth?” is now possible with AI.  Scaling up such questions – asking “What are these 5 million houses worth?” is finally possible with AI.

But not only that, as the metaphor goes, “managing business through the rear view mirror” is not useful. AI offers the ability to create propensity models on the fly – models that allow us to predict who might list a house to sell, who might default, what extrinsic variables really affect large movements in the real estate market and countless other hypotheses. AI, on its own motive force, learns from testing these hypotheses and comparing them to fact. In the process, AI refines the hypotheses and offers information in a timely fashion, timely enough in fact to make to be able to make decisions.

Here’s the thing though. You can’t “AI-ize” a technology stack after the fact. Either you approach your data and technology with an AI-native perspective or you don’t. In addition, AI –ready organizations are held back more often by culture than by technology acumen; while organizations can change, the after-the-fact bolt-on schemes don’t do the job.

In the business world, most of us search for simplicity. While that is a laudable goal, it also forces us to cut corners. In complex systems with a seeming infinitude of variables and outcomes dictated by data and emotion, planning for the “unknown unknowns” is key to success. Artificial Intelligence offers that, mind you, not by itself but with the active partnership of real humans with real experience. There are no AI silver bullets or magic buttons. Wisdom is still important.

Industries, even ones that are complex and large, change over time. Advancements in technology provide an accelerant for this change, with both winners and casualties. Real estate is no different-and if recent history provides ay lessons at all, it is that ignoring the winds of change can only be to the detriment of the entire ecosystem. Fear is understandable, that is if it spurs preparation for the changes coming and coming fast.

Which lead us to our final thought: constant invocations of AI in real estate lend themselves to hype but the maturity of some existing AI solutions as they apply to real and significant problems and opportunities in real estate is very needed, very real and very now. We are in the era of real AI for real estate, but you need to kick the tires on all claims.