It’s a fact, AI is integrating into our daily lives more than ever. There are many proponents of artificial intelligence, as AI helps augment the work we do and improves the efficiency of our personal life. Opening your phone with face ID, using Grammarly to spell check emails, smart home devices and yes, Netflix, are all powered by AI to deliver a tailored, custom experience. When used correctly, AI frees up our time to focus more on things that computers don’t do well, like creativity and empathy-related exercises. So all pluses, right?
It’s no wonder then that AI has made its way into the real estate sector. By means of algorithms, machine learning and AI technologies, artificial intelligence can help to ensure better investment decisions based on hard data. At least that’s the dream. In comes Zillow (https://www.zillow.com/), an American online real estate marketplace company founded in 2006, that launched a service that uses artificial intelligence to estimate the value of homes. In February 2021, Zillow announced a new “Zestimate” option that was essentially a cash offer from the company to purchase property, providing clients the convenience of selling their property in just a few clicks, which would ultimately minimize interactions with others especially during the height of the pandemic. To accomplish this, Zillow compiled algorithms with machine learning, plus lots of data to look at current/past/future housing prices in an effort to accurately predict the value of a home. Fast-forward 8 months later and Zillow is now shutting down that part of the business.
In a statement from Zillow, CEO Rich Barton says: “We’ve determined the unpredictability in forecasting home prices far exceeds what we anticipated and continuing to scale Zillow Offers would result in too much earnings and balance-sheet volatility.” In other words, getting the algorithm right proved more difficult than they had anticipated.
Of course, this doesn’t mean the end for AI in the real estate sector. There are still countless ways that AI is being used and no doubt this will continue to become more common place moving forward. But predicting house prices now or in the future isn’t something that AI can easily tackle. There are countless unquantifiable aspects of putting a price tag on a home, and these factors vary from person to person, neighborhood to neighborhood, which make it immensely challenging for AI to get right.