When some people hear “artificial intelligence,” their mind jumps to robots so advanced they can eliminate humans and take over the world. Think Skynet in The Terminator or the machines in The Matrix.
Or they envision the loss of countless jobs as human workers are replaced by machines. Neither of those scenarios is likely to occur for real estate agents as a whole: there will be agents for decades to come and AI could make them more successful than ever. The real question is: Will you be one of them?
That’s the main takeaway from a panel hosted by the California Association of Realtors’ Center for California Real Estate and streamed live on Facebook Thursday.
“I define AI as the practice of making machines think more like humans and humans do two things really well: pattern recognition and learning,” said panelist John Berkowitz, co-founder and CEO of AI real estate tech firm OJO Labs.
AI is an umbrella term for things like machine learning, facial recognition, speech recognition, computer vision and natural language processing. But as fancy, and potentially scary, as those things sound, AI today is very narrow in that it is designed to accomplish a particular task and is dependent on humans to make processes more efficient, panelist Russ Cofano, an attorney and industry veteran, said.
Panelist Marie Hagman, director of product and machine learning at Zillow Group, agreed. “Right now, if you have a robot that can play Go, it can only play Go,” she said. Lifelike robots like those in the movie Her are the “holy grail” of AI, she added. Those in the AI industry are in awe of the human brain and know how far away AI is from that, Berkowitz said.
Panelist Prem Natarajan, Michael Keston Director at the USC Information Sciences Institute, said a key aspect of AI is that “it should be able to deal with at least reasonably novel situations.” Rather than encountering a problem and determining what to do based on a predetermined look-up table, AI has to generate answers to situations it has not seen before, he said. The idea is that, by feeding it lots of data, AI will be able to make connections no human could.
“Machines can look at a million documents and find shallow stuff across those different things that we would never be able to do in our lifetimes,” Natarajan said.
AI is a part of everyone’s lives now, according to panelist Peter Jonas, president of western regions for high-profile brokerage Compass. “AI is happening incrementally every day,” he said, noting that tech tools like Google and Uber use massive datasets compiled over a long period of time to provide AI-powered driving directions. In real estate, companies like Zillow and OJO are using AI to recommend listings or home features that consumers might like based on previous search history.
Is AI going to replace agents?
“My answer is an emphatic ‘No,'” Jonas said.
“What you can’t replace is the emotional side of this, the EQ [emotional intelligence]. A lot of that IQ can be replaced with artificial intelligence. Me having to look up the price of a house – that you can replace.”
Artificial intelligence is not going to help a buyer deal with the experience of losing a bid on a house or help a mom going through a divorce. What AI can do is give an agent the opportunity to let technology do more mundane, less nuanced tasks and ramp up from 50 transactions to 500 transactions a year, he said.
But that doesn’t mean agents should think “my job is not in danger,” according to Cofano. The real estate industry does not create transactions — it services demand, he noted. That means that there are only so many transactions to go around.
“The role of the agent is not safe. From a pure cost standpoint, AI won’t be deployed anytime soon across the industry, so the people that have access to it will get the lion’s share of the business,” he said.
Hagman took a middle ground. “Certainly the job of the agents will change. Does anybody remember the MLS books? … Now with the internet and search engines buyers are searching online. Is anyone here less busy? I don’t think so.”
“That personal touch really, really matters, but to the extent that we can use technology to make that process better, faster and provide new opportunities, it’s an exciting change that can move the industry forward,” she added.
Natarajan said he couldn’t see how “with all the growth and efficiency that’s being promised, there isn’t a shrinking of the [agent] labor pool.”
OJO is betting on agents being around for the next decade, Berkowitz said. But “do we think some underperforming agents will be removed? Yes.”
The panelists had some recommendations for agents that want to be among the thriving in an AI-driven world:
- Every agent should be an A/B tester, Cofano said. Try out different AI companies like OJO, he suggested. “Be a participant in this process as opposed to being a recipient of the process.” Hagman agreed. “Evaluate what tools make sense for you…what’s going to make your life easier. That’s going to be the virtuous cycle that supports the industry and does the right thing for the consumer,” she said.
- Make sure wherever you’re sending your data that they’re looking to accelerate, not displace agents, Jonas said. “Technology should accelerate someone — it should be you,” he said. That means asking yourself: Is this company a partner of yours, or are you a means to an end? “Are they making money off of you? That’s fine. But are they making more money when you make more money, or is their performance not correlated to your success?” he said.
- Understand what is happening to your data. Someday, there will be a technological solution to trust — such as private blockchains or coding that lets brokers and agents know what happens to data once they let go of it, Natarajan said. But in the meantime, Jonas suggested agents talk to their smartest friends, “the people that you think are really pushing the envelope,” and ask them which companies they trust. He also advised understanding how your data is being used. Panel moderator and CAR CEO Joel Singer suggested brokerages parse user agreements for their agents and educate them on the differences between companies.
- Don’t approach AI with fear and hoarding, Berkowitz said. “Figuring out how to use the data is going to be more important than keeping it,” he said. Jonas agreed. “If you hoard your data, if you’re not part of this, then it’s going to disadvantage you in the future.” A listing by itself is not very valuable, but aggregated listings are, Cofano and Hagman said. That means choosing tech partners that can use your data and combine it with a broader set of data to provide a benefit to you, Hagman said. It’s part of business to make mistakes, but are you going to make mistakes because of fear or sloth or because of ambition? she asked. “Hopefully they’ll be mistakes of ambition,” she said.
- Never stop learning, Hagman said. There are plenty of examples of companies that failed to use technology in creative ways to compete, Blockbuster and Kodak among them, she said. “I think that it’s just really important for people to never stop learning. [Think] ‘How can I use this to create an advantage for myself rather than staying entrenched and doing things the way we’ve always done them?'” It takes leadership to make the choice to pivot, according to Hagman, but doing so can be the difference between ending up like Blockbuster or soaring like Netflix.
- The future is unpredictable, so focus on providing better service. “I think the argument can easily be made that agents will not be disintermediated [by AI], but the services can be devalued,” Singer said. That may lead to lower commissions. To that Natarajan said, “It’s very difficult to predict the shape of disruption. … There may be a certain fraction of the market that is entirely automatable, but how that disruption spreads from there is near-impossible to predict.” Singer advised agents to couple their interpersonal skills with the best technology for a situation. “You put people into houses. You change their lives. Using technology to make you better is going to be the answer to a lot of things,” he said.