But when it comes to the implementation of AI technology tools in other areas due to the reverse mortgage industry, there are many options. Packer explains some of these in this second part of his conversation with Housing‘s Reverse MortGage Daily (RMD).
Reviews
In addition to checking for potential bias that can result from the AI training process, Packer explained that AI tools can be useful in assessing real estate assessments within the reverse mortgage activities.
“One of these places where we use the AI tool is to watch reviews,” he said. “We can take determination about whether or not the assessment represents the value of the property. I know that some of our competitors use people to do that process, but we largely trust some models that have been built.”
This comes down to a question what more can be familiar to achieve the goal, he explained.
“Do I trust a model that looks emotionally at appreciation of the house prize, or at the history of real estate, the compositions used with regard to other compositions?” Early Packer. “Do I trust that if less is biased than a person who brings their own thoughts, expectations and experience to the table?
“I would claim that I think the Gedempt AI model will be less agistic, especially if it has been correctly developed over time and unfortunately has been tested than people.”
Right service
Packer said earlier that he was not entirely sure of the idea that older customers prefer ‘function above flash’. This feeds the need to see older customers as individual individuals instead of having them knew in a group.
“We are not a monolith,” he said. “I even look over my group of friends. Some of them have refinanced many times and are very comfortable in an online environment. Some of them have not refinanced or less able to make that journey unguided.
“And I think that’s one of the enormous forces of Longbridge: if you want to communicate with us and go through that platform digitally, we have a solution for you.”
Customers who want personal contact can also be served by this system if they prefer. If that means a telephone or video call or a visit to the office, the company is structured in a way to meet one of these preferences, Packer said.
“We are happy to get in touch with you where, when and how that customer wants a very personalized basis, because one of these generalities falls apart quickly as soon as you come to the individual,” he said.
“I would even say that in an individual transaction, if you have more than one borrower, they have other parties that help with the progress of the process.”
Limits of customer -oriented AI
Trusted advisers may prefer different engagement options than potential borrowers, and Longbridge wants to accommodate these preferences. Ultimately, options will include a consumer-oriented AI-Chatbot that is still developing, Packer said.
“That will be a big leap for us, and we first have to become super comfortable with our partners and see some evolution before we will do that,” he said. “But I think that day will certainly come.”
Packer shared a personal story about staying in a large hotel chain and wanted to find the tag “do not storm” on his door. He couldn’t find it in his room, so he tried to call the reception, but soon realized that he spoke with an AI chatbot.
It was unable to understand the nature of his question and he immediately longed for direct human interaction.
“It gave me the option to report and talk to someone else,” he said. “I think we all at such a moment, we want to reach that real person. Longbridge will always support that, because I think that is always logical.”
