If you’re like a lot of people, at some point in 2023, you created a ChatGPT account and started asking it questions. At first, it’s extraordinarily impressive. You can speak to it as you would any person, and it will respond in a very human-like manner. For many, ChatGPT passes the Turing Test with flying colors.
But it didn’t take people long to start finding cracks in ChatGPT’s armor. Its tendency to “hallucinate,” or invent facts that have no bearing in reality, is well-documented. Two New York lawyers who used ChatGPT to write a legal brief even discovered, to their professional detriment, that ChatGPT will invent fabricated citations to justify its conclusions. In short, if ChatGPT doesn’t know the answer, it will often make one up.
With that tendency in mind, it’s not hard to arrive at a point in which you view ChatGPT’s output with a skeptical eye. That makes the increased use of AI chatbots in financial matters particularly concerning!
How is AI Being Used by Non-Professionals?
Non-professionals are using chatbots mainly as a resource to learn information. If you ask ChatGPT, “what is a Roth IRA?” it will probably return an answer that’s generally correct. If you don’t understand the answer, you can even refine the results with prompts such as “Explain it to me as though I’m a 5th-grader,” and the chatbot will give you an answer that’s easier to comprehend.
This can be useful, to a point. As noted earlier, chatbots often answer questions with incorrect or outright false information. In order to determine if the answer you’re given is correct, you need a sufficient understanding of the subject to evaluate it properly. You need to have a good idea of what the answer should be before you can decide if the chatbot’s answer is sound.
Some non-professionals are beginning to use AI for more dangerous pursuits, such as choosing stocks in which to invest their money. This is a practice I find incredibly unwise for a number of reasons. First, as I’ve mentioned before, it’s generally a bad idea to rely on only one source for investing advice. At Asset Preservation Wealth & Tax, we get input from multiple viewpoints because we know that no one source can be right 100% of the time.
A second troubling aspect of getting financial advice from an AI is that it can be difficult to determine how it reached its conclusion. If I go to one of my investment sources and ask, “What led you to conclude I should move my clients’ money into that investment?” that source will be able to explain themselves in great detail without making anything up. AI either won’t be able to do so or, for the hallucination reasons discussed above, won’t be trustworthy when it does.
As I’ve said before, one of the most important investing questions to ask is, “What if I’m wrong?” AIs won’t ask that question, and they will present factually incorrect, or just plain bad, advice with the same confidence they have when presenting sound guidance. In other words, if you don’t already know what you’re doing, learning from AI is fraught with peril. Why would you want to risk your life’s savings on an algorithm that may not fully understand your question?
How is AI Being Used by Professionals?
Financial professionals are using AI in a number of different ways. As with just about any computing system, AI is capable of absorbing massive amounts of data — far more than any human can process. As such, there’s a use case for AI as a gatekeeper. A professional could direct, “Find the 20 investment strategies with the greatest potential for superior returns,” and then analyze those strategies to find one to recommend to a client. Because an AI can look at quite literally every potential investment strategy, it will base its conclusion on more raw data than a human would.
As with amateurs asking AI questions, that strategy is also dicey. How can an advisor know if the AI has actually picked the 20 best options? Blind trust in the output of anything, even if you have the expertise to vet the results, is dangerous.
That’s why at Asset Preservation, we aren’t using AI as part of our process to advise clients. Doing so would increase our efficiency markedly; we could advise more clients with a smaller staff. But at what cost? We recognize that declining to use AI in our advisory process might put us at a competitive disadvantage from a volume perspective, but from a quality perspective, I’ll bet on our humans every time.
There’s another problem we haven’t yet explored, which affects both professionals and laypeople: privacy. Generative AI systems train themselves in part via their interactions with people asking them questions. In other words, anything you type into an AI system can, and probably will, be stored. When dealing with financial information, it’s important to keep your information private, but AI is potentially able to regurgitate the data you give it.
Think about that for a moment: AI knows who you are and is capable of relaying your data to someone else. Everyone, professional or otherwise, should assume the public will eventually see anything you input to an AI, and they should make decisions as to what they tell it based on that assumption.
The final point I want to make about artificial intelligence is its inherent weakness that will persist regardless of how sophisticated it becomes. Even if AI eventually proves it can be trusted not to invent facts or misunderstand what’s being asked, it will still miss a vital component of human intelligence: It won’t care. At Asset Preservation, we entered this career field because we have a genuine interest in finance. We advise clients because we care about helping them grow their money and secure dream retirements.
AI doesn’t have the capacity to care. If it gives you bad advice and you lose everything, that won’t even register as a problem for the AI. When seeking financial advice tailored to your unique situation, it’s important to get it from sources that care about that advice’s impact on you. Because you can only get that from humans, it’s important to have a decision-making partner that’s a real, living breathing financial advisor.
Stewart Willis is the founder and president of Asset Preservation Wealth & Tax, a financial planning firm in Phoenix, Arizona. Investment advisory services offered through Foundations Investment Advisors, LLC, an SEC registered investment adviser.
The commentary on this blog reflects the personal opinions, viewpoints and analyses of the author, Stewart Willis, providing such comments, and should not be regarded as a description of advisory services provided by Foundations Investment Advisors, LLC (“Foundations”), an SEC registered investment adviser or performance returns of any Foundations client. The views reflected in the commentary are subject to change at any time without notice. Nothing on this website constitutes investment, legal or tax advice, performance data or any recommendation that any particular security, portfolio of securities, transaction or investment strategy is suitable for any specific person. Personal investment advice can only be rendered after the engagement of Foundations for services, execution of required documentation, including receipt of required disclosures. Any mention of a particular security and related performance data is not a recommendation to buy or sell that security. Foundations manages its clients’ accounts using a variety of investment techniques and strategies, which are not necessarily discussed in the commentary. Any statistical data or information obtained from or prepared by third party sources that Foundations deems reliable but in no way does Foundations guarantee the accuracy or completeness. Investments in securities involve the risk of loss. Any past performance is no guarantee of future results. Advisory services are only offered to clients or prospective clients where Foundations and its advisors are properly licensed or exempted. For more information, please go to https://adviserinfo.sec.gov and search by our firm name or by our CRD # 175083.