Fresh off the heels of the AI Symposium and ahead of the coming Equity Symposium and Exchange conference, Evan Harp sat down with Wealth.com’s Co-Founder and Chief Product Officer Danny Lohrfink to discuss artificial intelligence and how advisors can implement it in their practices today.
Evan Harp: Artificial intelligence is getting a lot of discussion as being the next big driver for massive change across many sectors of the economy. How will AI impact the job of the financial advisors specifically?
Danny Lohrfink: AI is going to impact every single sector. The job of the financial advisor will obviously take an impact from AI as well.
I take this specific interview right now as an example. There is a transcript being recorded of this interview, and it’s probably by an AI. It might even be called otter.ai, which is a recording tool that I use. That’s a speech recognition artificial intelligence that’s able to recognize the different characters and sounds coming out of my voice, and then transcribe it so that you, on the other end of this interview, are able to go back through the notes and accurately reflect my quotes. Well, same thing for a financial advisor. Financial Advisors can leverage a tool like a transcripting AI tool to be able to keep track of the conversations that they have with the client so that they can better serve the client on an ongoing basis.
Now, more broadly, how will AI impact the job of the financial advisor? I think it’s really important to note here that, much like the rest of the American population, financial advisors are both consciously and unconsciously already integrating artificial intelligence into their advisory practices. They probably just don’t even realize it yet. But they are leveraging AI to handle many of the mundane, tedious tasks that soak up a lot of time and resources, such as account opening processes, fraud detection. This is all embedded into the bowels of their practices. But it’s already there. Now, where is it moving forward towards? Well, it’s only getting more and more advanced. And the utility in the application of artificial intelligence is becoming more broad.
Lohrfink on How AI Will Impact Society
Harp: To broaden the picture beyond how financial advisors will be using AI, how do you see AI changing society at large in the medium to long term, instead of just right now?
Lohrfink: Yeah, it’s a good question. The purpose of technology is to drive efficiency. That means faster, less expensive, more scalable output without compromising on the quality of the output. That is the purpose of technology and that is what artificial intelligence can deliver.
Now, the unique nature of artificial intelligence is that it’s taking all the existing and documented knowledge in the world and combining it into one massive, large language model to be able to pull upon. Rather than taking the expertise of one person’s brain, or three people’s brain, or a team of people’s brains, is combining everybody’s brains into one to be able to create output. The power of that cannot be quantified right now. That is going to have a massive influence on society at large because you’re unlocking a communal knowledge base. That’s incredibly powerful.
Using AI In Your Practice
Harp: We spoke a little bit about how advisors and really everyone is already having their lives unconsciously affected by AI. But is there a conscious decision advisors can make right now that can help them leverage AI to make their jobs easier or more efficient?
Lohrfink: Absolutely. I’ll give you one example from our company, Wealth.com. Wealth.com just introduced Ester, which is the AI legal assistant. And what Ester is, is a proprietary machine learning model specifically trained on estate planning documents. Ester is able to empower financial advisors to understand the nuances and complexities of existing estate plans in just a matter of minutes. Otherwise, it would take financial advisors hours to read through estate planning documents and extract the key information.
Our model is trained and conditioned to understand the different linguistic tendencies of trust and estate attorneys, and then be able to read documents and extract key information in just minutes. It’s that type of application that advisors can leverage today. It, obviously in our case, exists in the estate planning realm. But it also exists in the realm of CRM systems, where advisors can leverage AI within CRM systems to scan through emails, extract patterns, and surface unique insights for conversations that the advisor can have with their client, either at present date or later date down the line.
Lohrfink on AI Misconceptions
Harp: That’s terrific. What’s something that most people get wrong about artificial intelligence?
Lohrfink: I’d say there are really three things. The first is that, as it relates to financial advisors, AI is not going to replace financial advisors. What it will do is create a material advantage for advisors who are using AI versus those who are not. Those using AI will see their businesses scale more rapidly as their employee’s time will be spent on higher-value tasks.
Think about life before the calculator or before the internet. And now, pitting an advisor from that era against an advisor today. The advisor that is not using the latest technology, that’s not using the internet, that is not using a calculator, that is not on Bloomberg — they are at a material disadvantage to the advisor that is.
That’s misconception number one — that you can be an advisor that doesn’t leverage AI and still compete. That is not going to be the case in a few years when there’s greater adoption.
Number two, AI should not just be a buzzword, there actually needs to be utility. You can’t just embed ChatGPT into your website so that clients can ask the question to and say, “Yeah, I’m leveraging AI.” There needs to be real utility. What do you need to do, as a financial advisor? You need to ask yourself, where are the pain points? Where are those mundane, tedious tasks that make me say, “Gosh, I wish that this could be automated.” Find that pain point. Odds are, somebody has developed an AI solution for it. Now go find it.
But don’t think that just using AI, by implementing an off-the-shelf large language model, like ChatGPT — don’t think that is going to be your magic bullet. You need to have utility behind it.
The last thing that I’ll mention around misconceptions here is that AI is inaccurate. A large language model that is not trained and is not conditioned is inaccurate. But if you take Ester, Wealth.com’s proprietary estate planning technology, for instance, we have trained Ester on estate planning to understand the nitty-gritty nuances of “what is an executor?” Or, “What is a beneficiary? What is a guardian? What do specific provisions look like in estate planning documents? Is there a marital trust in this document? Is there a spendthrift clause in this document?”
By training the model on estate planning specifically, we’re able to massively increase the accuracy of the output. You can contrast that to an open model like ChatGPT and the accuracy is completely different. Because ChatGPT isn’t – it’s impossible, at this point in time, to comb through all that massive amount of data and condition it to give you the accurate output that you’re looking for. This is why ChatGPT has an F1 score that tests around 49%. An F1 score is a measurement of the number of elements that you’re looking to extract and the accuracy of the returned output. Whereas Wealth.com’s Ester is able to boast an F1 score north of 90%. That’s the difference when you have a trained model versus an untrained model.
How to Think About AI When You Are Wary
Harp: That makes a ton of sense. Finally, what would you say to advisors or investors who might be wary about using AI. Perhaps they got burned by NFTs, which were similarly pitched as this broadly applicable tool that was going to revolutionize a bunch of different fields, and then ended up being a bit of a bubble as an investment. For people who are skeptical about buzzy tech, what would you say to them specifically about AI?
Lohrfink: You’re already using it. Stop being skeptical. Look at your iPhone today, and go to your photos. Type in “dog.” You’ll be surprised you’re going to have 159 hits of dogs, because you’ve been taking a lot of pictures of your dog and all of those pictures will instantly be filtered for you.
Okay. Also, open up your iPhone or your iPad or whatever, and say, “Hey, Siri,” and then Siri is going to say, “Yes, can I help you?” Okay, well, that is an AI model. You are online, and you’re either on social media or you’re shopping. Why do you think you continue to get recommendations for those denim jeans? Well, because you had a Google search for denim jeans and now the AI has analyzed your behavior, your interest, and is now delivering a tailored experience to you.
Same thing in social media, you seem to be liking the same themed accounts over and over, you’re really into health these days and so we’re going to continue to feed you the Athletic Greens social media account, or the Pure Barre social media accounts at the top of your feed, because that’s what gets us the most time out of your day. Because, again, it is social media. The attention economy. Your time and your attention is their profit. They’re leveraging AI to do that.
You’re already using AI. To think that this is something that’s a fad that will pass by is a fallacy. Then you think about all of these amazing innovations, like self-driving cars, and you say, “Well, I never trust that.” But then you look over and you see the person on the road in the Tesla with their hands off the wheel, because Tesla is using AI to be able to recognize the patterns and the landscape and the topography around to drive the car. That’s a real-life application that, in five years, you might be in that same Tesla.
So I guess I’m saying don’t knock the technology because you’ve been burned on other types of technologies in the past, think about the true applications and use cases in today’s society, and then extrapolate those into the future. And you’ll see the true possibilities with AI.
For more news, information, and analysis, visit VettaFi | ETFDB.