Science and technology amaze me. In a unique way, I choose to embrace science and technology for the good that it can do. I am bullish on humanity making the right decisions, but most of all, I am willing to give up certain freedoms in exchange for allowing technology to blossom as a contributor to humanity.
Perhaps the most significant advancement we have seen over the last decade is deep machine learning. Deep machine learning is a process that allows algorithms to run against massive amounts of data in a mostly unsupervised way. It allows computers to think and be creative. Computers can ask, answer, and critique questions. Through this process of data interrogation, computers can learn. Live product examples are the ability for computers to detect attributes in a photograph. But there are many more applications in the pipeline.
Another example is the Zestimate®, often thought of as an algorithm. The Zestimate is not likely one algorithm, but a group of algorithms – hundreds or perhaps even thousands of them. Using deep machine learning, they put values on homes. When the value fails (an actual transaction happens at a different price from the Zestimate) the machines interrogate themselves to learn how it failed. Every failure or group of failures drive improvements. There are only about 5 million examples of transactions in real estate each year, unfortunately. If there were more, the Zestimate would become more accurate much more quickly.
Emerging AI Technology
Two TED Talks that I watched this weekend give insight into the future of technology and technology in real estate– a future that is likely to happen in 10 years. Remember, technology develops slowly, and then very quickly. The biggest area of phenomenal growth will be in artificial intelligence – AI, an area that is in the slow phase now. We’re seeing OJO Labs and Restb.ai, among other companies in the marketplace, commercialize AI products today; it’s the first leg of the race.
Machines are watching humans and learning. Google analytics is a great example of a machine watching how humans interact with data. Siri and Alexa are great data gathering tools for understanding how humans speak – what questions humans ask and how they ask them – accents, word choice, even with mood inflection. To some extent, the machines even listen to background noise. Now, imagine if a machine could silently listen directly to the voice in your brain. Because of a recent technology breakthrough, a small piece of tape attached to your neck can now allow a computer to listen to the voice in your head without you uttering a word. Not record the audio coming out of your mouth, but the signal of language that comes from your brain. A direct link between your brain and a computer is now possible.
Watch this short TED talk to see this amazing technology in action.
Mind blown, right? A device connected to your neck would allow you to silently communicate to a machine. But there is other technology that can will make AI even more interesting.
If you agree with the notion that speech is the most complex activity of the human body– it requires highly coordinated thought, perception, shaping of the mouth, tongue, and air control. You will agree that what really makes speech profound is the coordination with facial expression and other body movements. Speech is more than the sound: it is acting. A new technology can match your exact physical expression to speech. Watch this next video demonstrating that a machine can not only listen to our brains, but also create a facsimile of a human that is an exact replica.
Now that your mind is tying these two remarkable technology solutions together, you may be a little scared. Or you are thinking about the Holodeck from Star Trek becoming a reality.
I am not afraid of the future because I trust in humanity. As dangerous as these technologies can be, I am more bullish on these technologies being used for good. Although 2016’s Word of the Year was “post-truth,” followed in 2017 by “fake news” and 2018 “mis-information,” I believe that this type of technology can be just as precise at detecting bullshit as it is at delivering good experiences. Photoshop has been used for disguising the truth, but we can detect it. If only Facebook would shut down fake images and fake videos by using technology to detect and block it like YouTube blocks pirated music.
As for the future of the real estate industry, data is an asset that will enable brokerage companies to leverage AI in ways beyond our comprehension. Imagine tying digital humans with Matterport tours for virtual showings. It is going to take a lot of consolidation of data owners to make these things happen in real estate. More than that, it is going to take a lot of investment. SoftBank Group knows that you need to start with the data originators – the brokerage company – ergo Compass and Opendoor. What we see regarding data from SoftBank is just the beginning of their investment in building the real estate economy of tomorrow.
To prepare for this future as an MLS, it is important to adopt RESO® data standards to structure data in ways that allow AI and machine learning to work across massive data sets that will serve your subscribers. Small MLSs will not have enough data to make this effort worthwhile. They will need to merge their data into larger and larger databases.
Right now, Realogy, RE/MAX, and Keller Williams Realty are in a great position with big data repositories. But I think that the industry needs to think more along the path of UpstreamRE, where every broker’s data has a chance to be leveraged as an asset of the broker regardless of size. If Upstream does not make it, small brokerages will perish in a data economy and only the large will survive.
Vista Equity Partners has positioned itself well for the future through the acquisitions of transaction management solutions Instanet Solutions and zipLogix™. Even though they do not own the transaction data, it is on their servers and they can dip their straw into the data to interrogate it. Similarly, I think that Realtor Property Resource® (RPR) has this opportunity. I am mad that RPR® has not yet made any profound discoveries from the data pulsing through their systems.
CoreLogic and Black Knight Financial Services have both created over a billion dollars in annual revenue from their non-MLSs data aggregation. It’s impressive. ATTOM Data Solutions has done well too – but not quite as well. I believe that the data held by these firms and and others like Zillow will yield value when combined with AI and deep machine learning. Matterport has a huge head start on virtualization data. You can understand why Zillow® has started their investment in this type of data capture. 3D scans of buildings are fundamental to virtual reality.
If brokers want to have a hand in the future with data, they better pump up the efforts on Upstream adoption for data collection, storage, and access. Data is the table stakes in the future of technology. I think that brokers participating in Upstream will create a future asset in their business by being able to license data independently or as a collective.
Brokers should take data licensing seriously right now. The broker needs to perfect and copyright the ownership of the data they are collecting, and be very careful when allowing others to license that data. A broker would be wise to never sign a data license agreement giving away the rights to their data in perpetuity to a third party, and indemnifying the licensee forever. To sign away these rights disqualifies the data as an asset for the broker and collects liability for whatever that company does with the data. We cannot possibly imagine the future applications or value of real estate data, but it is easy to imagine the possibilities if you have a lot of data.
The future is about 10 years away. Will you be ready?
WAV Group specializes in Data as an Asset strategy development. If you want help, give Victor Lund a call.