Any relationship starts with a conversation. And for wealth managers getting comfortable with rapidly developing AI technology, most are getting to know the new space by dipping their toes into language programs, specifically ChatGPT.
According to a new Arizent survey of wealth managers, 63% said they know and use AI technology as ChatGPT. The next most familiar AI tools to wealth managers were Bing AI, Bard and Wolfram Alpha — all language and conversational learning-based tools. And there's a reason for that.
OpenAI's ChatGPT has gained widespread popularity in being an easy AI entry point at little to no cost to the user. It has 100 million users weekly and as of April 1, OpenAI said people can use the tool instantly without needing to create accounts.
"If I'm thinking about product positioning, for example, if we're launching a new product, I had ChatGPT up to pretend that it was Steve Jobs and then (asked) it, 'how would you, Steve Jobs, actually present this product to me as a customer?'," said Rajat Deva, head of marketing at Savvy Wealth, a tech-driven platform for financial advisors based in New York, NY.
That general interaction with ChatGPT has bled into many work-case uses for wealth managers. For Savvy Wealth's tech-based platform, Deva said they use AI language technologies to create educational content for wealth managers to engage with clients about unique financial situations, like medical professionals or athletes who come into greater wealth earlier in life.
"We did one for medical professionals . . . on how to think about first-time mortgages that you might be able to get a competitive rate as a doctor; how to think about medical school loan refinancing; where to go for your first car," he said. "All of those different pieces speak to the advisor's core audience."
READ MORE:
At the same time, even earlier AI adopters recognize there are major trust issues with AI, partly because it's a new technology, but also because language applications like ChatGPT have been widely known to so-called hallucinate.
"In one test scenario, I got a very well-written, almost five-page answer that was structured correctly, but entirely a lie," said Nick Graham, executive vice president and chief technology officer of broker-dealer Cambridge Investment Research based in Fairfield, Iowa. "It chose correct numbers and percentages" but "it basically lied like my 18-year-old son does to me."
Graham added that AI can do much more than the popular language models once it's been trained on a particular topic such as regulation in finance.
"These are tools that are out there today, have been around for more than a period of time" but "learning to leverage them effectively has often been a low-percentage activity," he said. "But I think the new AI buzz has got everyone revisiting those thoughts and reinvesting in those activities of adoption."
READ MORE:
When Arizent, Financial Planning's parent company, asked wealth managers the areas of their personal lives that they trust AI to be mostly responsible for, only 25% said in making financial recommendations, while 50% trusted AI to predict their car or house maintenance needs. This correlates to the challenge wealth managers face in implementing AI — they might feel comfortable in using it to help craft a sentence, but they don't trust it to make accurate, predictive financial decisions.
READ MORE:
"Generative AI on its own, if it's not fine-tuned and trained and paired with analytical tools as well, it goes back to that fluent bullshitter," said Brooke Juniper, CEO of TIFIN's Sage, an AI-powered investment platform for advisors based in Boulder, Colo. "There's mistrust there. And what's important for these platforms to build — and it's something that we're very focused on at Sage as well — is explained ability."
Those providing AI-based platforms, like Juniper at Sage, say there's a certain space for AI in wealth management, particularly where the technology can quickly scan and interpret complex financial or legal documents, or accurately predict client behaviors to help the advisor better communicate.
"The way that we're leveraging AI is taking these tedious processes and using technology for what it's intended to do: reduce cost, reduce the time it takes to do something, drive efficiency," said Danny Lohrfink, co-founder and chief product officer at Wealth.com, a tech-based platform for estate planning for financial advisors based in Phoenix, Ariz.
Advisors could use AI as a meeting-recording tool that interprets the information to produce summaries and even the client's sentiment during the call.
"So, was the client frustrated? Were they happy? Are they primed right now for you to be able to ask them for a referral," Lohrfink said.
If regulation didn't apply, more than 50% of advisors surveyed by Arizent said they would use AI in their profession for specific needs, including: research and fact checking; assisting employees with routine inquiries; proactively defending against hackers; and assisting customers with routine inquiries.
But these more advanced capabilities for wealth managers are also in the early stages, Graham said.
"We're starting to see the use cases prove themselves out in the form of savings in efficiency of workflow function," Graham said. "That's the five-year vision right now. You're seeing pieces and glimmers of that — that haven't been fully integrated with other technologies — that are going to make that a reality."
Both Graham and Juniper are on Financial Planning's advisory council for its new
