73% of American adults do not have a financial advisor, while those trading to gen purpose AI chatbots risk hallucinated numbers and privacy exposure with sensitive personal data. Now, fintech fintech veteran Bill Harris, former CEO of PayPal and Intuit, is launching evergreen AI. The platform combines Gen AI with a patent pending deterministic engine to bring institutional grade wealth strategies to everyday investors.
Joining us live this morning to discuss AI's role in consumer wealth management is Bill Harris, co-founder and CEO of evergreen AI. Bill, good morning. Thank you so much for joining us. So it may be surprising to many, but surveys do show that about 70% of Americans lack a financial advisor. And yet you've stated that AI can eventually serve households with up to $20 million in assets.
So what was the impetus behind evergreen AI, and how does it democratize high net worth wealth management?
Yeah, well, we see a very important opportunity to fill and more importantly, a consumer need. Um, most Americans are living unexamined financial lives. They're just not paying attention. And it's not their fault. They don't have the time and they don't have the resources. We're trying to use AI to make it ubiquitously available to families of all types.
It predominantly makes sense for people with sophisticated and complex finances, because if someone is just, um, is essentially a spender and a budget, or there are plenty of software solutions and including AI solutions for that.
And bill studies, do you show that general purpose chatbots hallucinate or miscalculate more than half the time on financial queries? But why are traditional language models ill equipped for personal finance? And how does your engine eliminate some of these errors?
Well, a number of ways. The most important is that Llms are large language models. They do words and they use words to do concepts and reasoning, but they don't do math. And so it's really important because personal financial issues are a elegant mix of concepts and calculations. We spend as much time building the calculation engines as we do integrating and utilizing AI.
So that's the first thing repeatable quantitative results, which is essentially the core of personal financial advice. Um, the second piece is um citations and research and resources. We have a vetted knowledge base that we use on a regular basis. And in order to avoid running out to the web and getting whatever, whatever information you can get or the AI just making things up.
Hallucinations.
Yeah. And I do want to expand on this and talk about what's happening when it comes to your architecture. I understand that your CTO has filed seven pending patents for a system that separates the communication from the calculation. So how does pairing a conversation on a mostly deterministic tax, as well as investment engine actually help when it comes to consistency and also verifiable advice?
Well, yes, he's got seven patents. He's been working now. We have been working for more than a year to develop what we call a fiduciary level, um, ability to deliver advice. We are an Ria, a registered investment advisor. And so we have a fiduciary duty to our clients and users to make sure that what we're doing is correct.
Um, none of the, um, llms, none of the AI companies have that obligation. And so it's the combination of the calculation engines, the financial resources put together in a way where we can audit and log every single recommendation and every single piece of information that we provide that allows us to make sure not only to ourselves, but also to the regulators, that what we're providing is, um, is appropriate.
Yeah. And, Bill, on a regular basis, we hear about some of the risks when it comes to artificial intelligence. And we all know that personal finance requires transparency. But some investors out there may be hesitant to feed sensitive balance sheets into AI or even data. So how does your encrypted planet integration actually allow evergreen to analyze some of those real time portfolios without exposing these sensitive data to public models?
Well, that's a really good point. And it's essentially in three ways. First of all, um, we never train on your data. We never expose your data to to an AI that's going to use it in order to get better and prove itself, as opposed to improve your finances. Um, that's number one. Number two, we never expose your information over the open internet.
And so everything you're doing, as it is securely tunneled between our app and our back end. And then third, we are we exist in a protected environment where we're using many of the most sophisticated tools to make sure that everything that we are holding for you, all of your information is encrypted and protected from hacking or or anything else.
And Bill, finally, before I let you go, you are a fintech veteran. So tell us what you believe. The rise of fiduciary level AI advisors mean for traditional human advisors, as well as the wealth management industry as a whole.
Yeah, well, I think that human advisors will always have a place predominantly because money is as much emotional as it is rational. Um, but human advisors are remarkably expensive to expend many thousands of dollars a year, too expensive for most American households. And so the other thing about it is it's inconvenient.
When you need an answer. You have to call, make an appointment, and then days later, perhaps get in touch with someone. What we're providing is free and available 24 hours a day.
Well, Bill, we will have to leave it there for today. But always great talking to you. Thank you so much for joining us today. And thank you so much for sharing all of your insights as well as your perspective.
My pleasure.
Thank you.
