Closing Time: Learning Lessons From AI

CLOSING TIME: In the classic sci-fi film “2001: A Space Odyssey,” two astronauts find their lives in jeopardy when an onboard ship computer thinks it is smarter than the humans that built it and takes over their space mission. A central concept to the story was the danger of artificial intelligence (AI), capable of out-thinking and possibly overtaking its human creators.

Today, AI is not science fiction. There are machines that can perceive, learn, reason, plan, problem-solve, move, manipulate objects and even develop basic social skills. Yet to date, AI has not been able to perfectly simulate human intelligence, nor can it simulate the kind of multifaceted skill set required to help someone buy a home.

For the foreseeable future, the likelihood of an AI-driven mortgage remains dim, and we have at least two decades of evidence to prove it.

The “myth” of AI

AI is a hot topic in today’s mortgage industry – with good reason. Automation designed to “think” for us has been introduced and implemented in practically every aspect of mortgage sales and production. Loan origination systems can assess a series of variables and create underwriting decisions. Other types of software can evaluate data about a consumer and determine whether that consumer is ready, willing and able to buy a home.

Although these are very basic examples of AI, they save an incredible amount of time and effort. Consider, too, how technology has dramatically increased the amount and the speed with which data is being shared, including data about mortgage loans. All of this information would be impossible to manage or process without machines capable of learning and making decisions about data on a large scale.

However, AI cannot do it all. So far, humans have only been able to create machines that can grasp information, make decisions and act as the machines are told. For example, a driverless car can “see” an obstacle in the road and, based on other factors such as speed and the proximity of roadside barriers and nearby vehicles, swerve to avoid it. But it only does so because it was programmed that way.

The types of automation we see in the mortgage process today are essentially the same. For example, more and more lenders are launching online portals that allow consumers to get a mortgage without the involvement of a loan officer. These tools mimic many of the tasks loan officers do. Yet they have only been programmed to do fairly rudimentary tasks. They aren’t capable of independent thinking.

The other question is whether most borrowers are ready for a completely automated, human-less mortgage. Consider the fact that the IRS has been letting consumers file taxes electronically for nearly 30 years. But today, only 40% of Americans file electronically. Because tax filing takes much less time and effort than buying a house – for most of us, anyway – I find it unlikely that most borrowers will trust their home purchase to a website any time in the near future.

Right now, the majority of people still want to work with a human being for one of the biggest – if not the biggest – transactions of their lives. That means a mortgage professional is in the best position to guide borrowers.

The lesson here? Automation and digital mediums are revolutionizing the way loans are originated and the way consumers think about the process. However, you should temper your enthusiasm for fully digital mortgages. Lenders that can strike a balance between technology and human touch will prosper in this rapidly evolving market.

Chris Backe is the director of financial services at Velocify and a sales automation expert with more than 20 years of experience offering technology solutions to multiple industries. He has spent the last 10 years in the financial services industry, holding various positions at industry-leading technology companies, including Ellie Mae and Salesforce.