As more companies embrace the incorporation of artificial intelligence (AI) into their business- and customer-facing operations, a concern has emerged among firms focused on delivering services to older people
AI training typically occurs with younger staffers, who could then incorporate unconscious institutional biases that might leave seniors feeling alienated from the technology.
Representatives of Longbridge Financial have spoken to HousingWire’s Reverse Mortgage Daily (RMD) in the past about the company’s forays into the AI space. In 2023, chief operating officer Bill Packer sat down with RMD to discuss potential use cases — including enhanced product education and streamlined loan manufacturing.
Now, with more AI experience under the company’s proverbial belt, Packer gave another interview to talk about the need for curation as Longbridge further develops its in-house tools.
AI and ageism
Packer described the need for clarity on what the true dynamics are for senior-friendly AI usage.
“There was a 2024 study that was titled, ‘Is AI ageist?,’ and really what they were looking at is … comparing different large language models and saying, ‘Compared to each other, which one gives more ageist-like answers?’”
The authors of that study may have brought some of their own biases into the equation, which was conducted qualitatively, Packer said.
“I don’t dispute at all that we need to be careful in how we approach this topic, and I think we’ve been very thoughtful in this area,” he explained. “But I would also say we have to unpack it a little bit because the headlines can be different than the underlying components.”
There may also be an idea out there that older potential users of AI, whether engaged with the reverse mortgage industry or not, may prioritize function over form. But Packer isn’t so sure.
“I would like to live in an environment that’s exciting and innovative, and I think most of our clients would like that, which also preserves a human being in the loop,” he said. “I think Longbridge excels at that, and we’ve been very careful as we have developed our AI ecosystem.”
Careful hand in training
One way the company is careful is by not allowing its data to be used in the training of large, publicly available AI tools, he explained. Longbridge does not see such use as appropriate.
“All of our data is always on shore, and we’re explicit about which tools we do and don’t use,” Packer said. “And we do a lot of internal testing. Bridget, which we just released last month in the broker and correspondent space, was in testing for well over a year and in development for 18 months.”
Packer has previously discussed the development of Longbridge’s “Bridget” AI tool. In April, he reported that Bridget was being trained and had become “highly knowledgeable about both the [Home Equity Conversion Mortgage (HECM)] and our [proprietary] Platinum products, and is currently learning our servicing guidelines.”
But getting to that point took a lot of time, attention and caution, he said.
“We wanted to ensure that what we’re putting out in the marketplace is a tool that will really resonate with our customers,” he explained. “If you look at our own employees, they’re typically on the older end of the spectrum. We’re making sure what we’re making doesn’t suffer from some of these issues.”
Digital natives
Packer also shared a personal anecdote. He’s been using technology tools for decades, including the first electronic spreadsheet program, VisiCalc, which was released for the Apple II computer in 1979. He’s also been doing spreadsheets in Microsoft Excel for the better part of 30 years, saying he can do “pivot tables in [his] sleep.”
But as we get older, information recall for these once regularly used skills might become harder to access.
“I knew that I knew how to do it, but I couldn’t remember how to do it,” he said. “What I did is I went to Microsoft Copilot, and asked it how to do what I was trying to do. It didn’t give me the entire answer, but it gave me enough of the steps that, all of a sudden, that brain cell re-fired and everything clicked again.”
Since he is not a “digital native” compared to younger generations, he sees this as an example of a tool for those who may not be digital natives to compete with those who are.
“I think these tools give us wonderful capabilities to sort of level that playing field,” he said.