AI is Making You Go Broke (Top 5 Risks of AI Investing)

AI is everywhere right now. It's writing emails, picking stocks, and even claiming it can manage your money better than a human. The question everyone is asking is, should you trust it? It sounds efficient, logical, [music] even smart. But when it comes to your future, there's more at stake than just math.

Here at [music] our firm, we've seen what AI can do and what it can't. Today, I'm sitting down with Phil Clark, our CEO, who's got five [music] big concerns about using AI to replace your financial adviser.

AI has no accountability. It can crunch the numbers, but you better make sure your inputs are right.

I'm not saying there's a bubble, but if there is, I do believe that AI is supporting that.

I think there's a lot of mutual fund companies out there that are using AI to help them build mutual funds. That's already happening. We just don't know it.

We'll also talk about when AI actually makes sense and how to use it wisely to [music] build lasting wealth. Welcome to We're Talking Money.

Phil, let's start with the big question. Will AI replace financial advisors?

Can I say it in one word?

Sure.

No. [laughter] Just [snorts] no.

All right. Simple as that. Show's over. Thank you all for tuning in. We'll see you next time on We're Talking Money. All right.

Sorry, but it won't. Okay. Will it do a lot of things? Will it add value in some ways? Of course. I think it will. And I think advisers who use it well will be someone that can benefit their client. But I think for people to simply say, "Oh, we can replace the advisor and all we need is AI." I think that is a mistake and breeds tremendous amounts of mistakes.

Yeah. And sometimes people won't know that those mistakes are happening until it's too late.

It's true. It's true.

Phil, you've said that there are five major concerns when it comes to using AI or using AI for financial advice. What are those?

For me, Spell, AI, the first one, and I've said this many times around the firm, AI can never know anyone. It doesn't know us. And it's not a person. I know people think it acts like one. In fact, I've run into people who use it as a counselor.

Yep.

Well, it's not a counselor. I think it knows data very well. It can go out and find the data, but it's limited even in that. But AI simply, for me the number one reason is it just doesn't know us.

So, you think what you're maybe saying is it doesn't understand empathy. It doesn't understand human emotion. Is that what you mean by it doesn't really know us?

That's actually one of my other reasons.

Oh, okay.

When you talk about the five and one of them is it has no emotion.

Yeah.

Right. That it can't think at that level. And everything, so we're skipping around here a little bit, but everything's driven by emotion. Fear and greed are two of the greatest emotions. And the most prominent emotions when it comes to financial planning and investing. So AI can never understand your emotions or mine.

Yeah.

It's just not. It can't do that.

Only humans can interpret each other's emotions.

That's right. A machine can make it sound like it understands you, but it can't. Period.

You and I are similarly aligned on our view of AI. I don't love AI. I use it for certain things. I very rarely use it for anything work-related. In fact, I don't ever use it for anything work-related. Would you use AI to manage your money?

No. away.

Yeah.

No, I think AI can give you certain parts to a money management program, but there needs to be human intervention. If you don't have the human bringing certain parts to it, then I don't believe AI can get it. People need to understand right away, Spell, that AI is limited to data that is open on the internet because that's what it's deriving all of its answers from. And it's also deriving it from lots of speak that people are putting out there. But there's plenty of walls that it can't get through.

Yep.

Subscriptions that it can't get through. For example, it can't go drag in data from Bloomberg, right? Because it's a paywall, right? And so those securities keep it from doing that. So it's limited.

It's a paywall. And I did not mean to interrupt you.

It's a paywall and it's a very expensive wall.

Yeah. Yeah.

And so it'd be, if AI could do all of that, right, then Bloomberg wouldn't need to exist.

That's true.

Or they certainly wouldn't have a competitive advantage.

That's true.

So no, I don't think AI is just, it's never going to be the end-all be-all, but it will be a great tool if used properly. It is a great tool if used properly, but no, it wouldn't manage my money.

So you mentioned that it can bring some things to the table. What are those? What does AI bring to the table for an investor, for anybody? What does it bring to the table?

I think that if we talk about these five reasons that I think of AI not being great. So I said one is AI can't know us, right? I think two, what can it bring to the table? Well, it can bring to the table a lot of things. But my second reason why I wouldn't use it or why I don't think it can replace an advisor is the old saying, garbage in, garbage out.

Yeah.

And so if you don't know how to put the request or the questions or the conversation into AI, it's never going to bring you back what you need.

Yeah.

So if you give it some basics, it'll give you back some basics, but it can't understand exactly what you're trying to accomplish. It'll never understand. And I said my third point was emotion. It'll never understand if your wife has a strong desire for a particular kind of investment or if she has a strong desire to avoid a particular. It can't know that. You can tell it some of these things, but it can't really understand what that wife's emotion is going to do to, say, the marriage, right? For example. So if you're pushed forward with what AI says, could cause your wife to have a meltdown, that's no good.

That's a good point.

So it just can't know those things. The other thing that AI can bring in that's really, really good. It can crunch the numbers as good as anybody and as good as any calculator. But you better make sure your inputs are right, Spell. For example, if you want to talk to you about taxes, you better be sure you've got the tax brackets right. You better be sure your tax information is being given to it correctly. All of those things matter, or if you're off just a little bit on any of the things that you want to calculate, it's going to give you a flawed plan.

That's true. Let me add something to that because I was watching a video that someone else did on can AI replace financial advisors. This advisor put a text message from a client in that he received, and AI gave the client the wrong answer because in the input, to your point, you got to get the inputs right. They didn't specify where they lived. And because they didn't specify where they lived, they got a tax calculation wrong. So to your point, there's so many variables, and if you don't get everything in and the system or AI doesn't recognize all of the variables, you could find yourself in the face of bad, bad information.

Well, so garbage in, garbage out, right? But here's the other element that I think so many people miss. Someone needs to be able to properly discern what it's giving back to whoever the user is. And if you don't know what's right or wrong and you just accept the AI's answer at face value, you stand a really, really strong chance of getting it wrong.

Right. Right.

So that's going to put you at a disadvantage right away. If you don't understand what the right answer is,

Mhm.

you run a huge risk. So I think AI is valuable, but only a human who has the knowledge to know the difference or do deeper research to be sure. Only that human can intervene there and say, "Wait a minute. This is bad output. We've got to do more here. We got to start over." Or it could be right, but it's in the wrong context. Got it? So those things have to be considered.

Yeah.

The other thing I wanted to mention when you ask about the five points is AI has no accountability. This is one that everybody seems to miss.

Yeah.

But in my opinion, you can't blame AI. If it gives you bad output and it causes you to screw up your financial life, who do you blame?

You could sling the computer out the window.

You can sling the computer out the window. You can yell at AI. You can send it a nasty note. And I hear that it will feel bad for you. Except it doesn't have emotion. So it only says that it feels bad because it's programmed to say that.

So does it know how to lie?

I think AI, that's my fifth point, is AI tends to be, in my opinion, and I can't prove this. I don't know. My opinion of AI is that it is a herd-thinking tool. It follows whatever's out there.

Yeah.

It's not, I don't believe that AI is objectively thinking about every situation. I think AI is giving you what is probably the majority of the information. So if it finds lots of stuff that tends to agree with whatever it's looking for, then it brings that back, and I believe that that herd way of thinking can get a lot of people in a lot of trouble.

Yeah.

Now, I can't prove that, but that is how it seems to me when I listen to what AI is doing. I'm not an AI hater. I believe that AI has a long way to go. I think it's going to get better, but I don't think anything can replace, when it comes to finance, when it comes to money, when it comes to making decisions that affects the person for the rest of their life. I don't think you replace the human mind that's well trained, seasoned, understands the laws and the regulations, how to do these. I don't think you replace that.

Yeah.

With an AI tool. Not saying it's all bad, but

Right.

So I would tell you it doesn't know you. It's never going to know you. It doesn't have emotion. I think garbage in, garbage out. And I think there's no accountability. You can't sue it.

Yep.

If you could sue AI, then I might say, "Okay, well, I'll take that one out." And I think it just brings back whatever it can find out there, which is all it has. And you better hope that what it finds is what you really need. And if it doesn't, then you better know how to discern. And if you don't, then you're probably going to get in trouble.

Phil, if everyone is using AI-generated advice and they're getting the same advice, my mind goes to the times when, for example, the early 2000s in the dot-com bubble or 2007208 with the housing market crash. If everyone's getting the same advice from AI, particularly for those who are using it to pick stocks or tweak their portfolio, could that make the market riskier for people?

Well, I think so, because if I, I haven't done this, but I bet if we ask AI how to invest and we said I'm 40 years old and I have 20 years before I retire and you give it the basic information, Spell, I would imagine that the bulk of the investment, it's going to recommend that you be in the S&P 500.

Yeah.

Or some index. It's going to be something like that. Well, if everyone's getting that, and right now if you look at the S&P 500, everyone knows, well, not everyone, but a lot of people know that the S&P is largely being driven by se the Mag Seven.

Yeah.

A handful of stocks. So maybe there's five to 10 stocks that's driving the S&P. In fact, I was talking to our chief investment officer and I believe he told me that the S&P 500 roughly today is 45%, somewhere in the 40s, technology.

Wow.

It's never been there. We've never seen that much exposure. Well, if everyone's pushing into that,

Yeah.

then that gets bigger and bigger and those tech stocks get larger and larger. I'm not saying there's a bubble, but if there is, then what does that mean for all these people piling into it? And I do believe that AI is supporting that. The other thing that supports it is 401k plans. Everyone is told just get it in the S&P or in the NASDAQ or whatever.

Mhm.

And so you have money just flowing in at a faster rate than we've ever seen. I think that causes problems. And a lot of that's AI driven. A lot of employees that don't know how to do this inside of a 401k plan, they go to AI.

Yeah.

And so I'm not saying AI is bad because it does help generate some reasonable things. And frankly, I think we all use it today whether we know it or not.

Yeah.

A lot of our systems use it. But I do think that's dangerous if it's pushing everybody in the same direction.

Yeah, I agree with you. So, if everyone's getting pushed in the same direction, how would you instruct those individuals to appropriately use AI in their wealth management journey?

Well, I probably wouldn't tell them to use it, okay? Because I don't think AI is a good tool for that. Let's go through some of those reasons again.

We'll get right back to today's show. But first, I'd like to ask you a quick favor. If you're enjoying this episode, hit that like button and make sure that you're subscribed to our channel so you don't miss what's coming next. And hey, if something in today's episode has you wondering about your own financial plan, why not get a second opinion? Visit omnistarfinancial.com to schedule a consultation with one of our team members. It's quick. It's easy. It could make a huge difference in your financial future. Now, let's get back to the show.

It can't know you. So, it doesn't really understand what you're trying to accomplish. It might give you an allocation for your investments.

Yeah.

That's built on a framework that you really don't need AI for, but it's just going to go out, find the data, and come back and give you those averages, and it's going to say, "Well, put 40% here and 60% there." And maybe that's okay.

Yeah.

But it can't know all of the emotions that you're dealing with. It can't know what's keeping you awake at night and it can't address those things. So that's point one. Point two, if you don't get your inputs right, it might give you a bad allocation that's not fitting for you or it might tell you to do things that really are not appropriate at your stage of life.

Yeah.

It has no emotion. So it doesn't get upset when the market falls. So if it doesn't know what upsets you or can't feel that, how can it help you with that? It can't help you work through that and talk you off of a cliff. Because we know that most investors because of emotion, they lose something around 3 to 6% a year in investment return because of their emotions.

Mhm.

What we also know is that Vanguard, one of the least expensive platforms out there, in fact, they're known for teaching people how to just simply use index funds.

Yep.

They did a study recently and found that a wealth adviser or financial adviser actually helps the client generate an additional I believe it was around three or 4% additional return.

I remember reading that.

So that to me says it all. If Vanguard, the king of index, if they're saying that an adviser brings value, yeah, you probably ought to listen probably. I think their study says something. And then AI, let's go back to accountability. There's no accountability.

Yeah.

So AI can give you all this good information, so to speak, but it has nothing to lose.

Yeah.

It doesn't know if it's giving you something that's really good for you or bad for you, nor do you. You just hope and you accept it for truth and you go with it. I think that's a problem versus you sitting in a room, Spell, with a client, reading their emotions, understanding what's keeping them awake at night, knowing that they've got a family member in the nursing home that's causing them anxiety, knowing that they've got bills coming that AI can't know about or they've got challenges in their life that AI can't know about, nor is it going to ask. You might help them manage through all of those things.

Yeah.

Without allowing their investments to create additional anxiety. AI can't fix that.

That's a really good point, Phil. Let me ask you a different question, and I've been thinking about this for a while. Do you think that AI is the next mutual fund?

Well, I think that's already happening. We just don't know it. I think there's a lot of mutual fund companies out there that are using AI and AI algorithms to help them build mutual funds, right? And Spell, in some ways, we use Bloomberg. That's a big part of our portfolio management at OmniStar. And we use those algorithms and certainly Bloomberg uses AI technology. There's no question. So, it's certainly helping us go retrieve data.

Yeah.

But believe me, and you know this, there's human eyes on all of that, on every bit of it. Whatever those algorithms bring in, we're still questioning. We're still making sure that it meets a certain test that we apply, right?

Not taking it at face value, right?

So, do I think we'll see that? Yes. And I think the robo advisor world is one version of it. That's more managing funds. That's putting together funds and allocation and so we already saw some of this. There's probably a few trillion in robo advisors now and everyone said years ago, Spell, so go back to one of your questions, you said will AI replace us. Well, everyone thought robo advisor would replace us and lots of advisers across the land got really worried about that and I leaned into the robo world. In fact, what we did at OmniStar is we said, okay, what part of our book of clientele could benefit more from that and should we let them go if it made more sense for them, where our value wasn't worth what they, in other words, we couldn't give them enough value to justify the fee.

Yeah.

Let them go. And in some cases I think robo could be okay. That's a form of AI, but it didn't replace us, right? Back, I'd say that the registered investment advisor world, the fee only or fee-based advisory world, stronger today than

Yeah. Giving stronger. I agree. I'm going to throw a name around that's been on this channel before and that's Dave Ramsey.

Sure.

Dave Ramsey, I think, paralleled or coupled something that you just said maybe slightly differently, but he said that AI could be beneficial for someone who's just getting started in their wealth management journey, but that he would not subject, say, a million-dollar portfolio, which for our firm is maybe a little small, but he wouldn't subject even that amount of money to an AI managed robo advisor. Do you agree, disagree, have additions?

I would partially agree. I think for someone getting started, if they don't have anywhere to get advice, I think a simple question for AI such as, I'm 30 years old and I want to start investing. I plan to work until I'm 65. What allocation would you tell me to use? AI is gonna come back with the standard, right, probably 8020 or 7030. I don't have a problem with that because I think that feedback is pretty simple.

Yeah.

It's germane to what this person needs and I think it's reasonable and then that person needs to go find the funds. What I wouldn't do is wait on AI to tell me which funds to use.

Yeah.

But I wouldn't have a problem if I then ask AI, give me the allocation to say the S&P 500 so I could at least see what that looks like. If I was going to do this myself, I could see the allocation in the S&P and I might say, "Okay, well, I'm going to try to build something similar to that." But I don't think I would simply tell it select everything for me and me just accept that and hope that it's right.

Yeah.

But I don't think Ramsey's wrong on that. I just wouldn't leave everything to chance.

Yeah.

And hope that AI gets it right because as I said, you can't sue it. You can't blame it. If it gets it wrong, sorry. That's on you.

Well, it's on you. Well, I'm under 30 and I would not

Would you use it?

No, I would.

Would you let it handle your things?

No, I wouldn't.

Well, if you said yes, I'd be shocked because one, you're credentialed. You've got lots of experience and you understand the ability of the human mind to bring a different perspective to things and to help people work through things.

Yeah.

If you said yeah, you would let it, I would think, man, what the heck just happened? I might be in fear for my job. Just, I'm kidding. I'm kidding. But AI is not all bad and I don't want anyone watching this video to think that we feel that way.

Yeah.

AI has to be embraced and I think if we embrace it then we get better.

Yeah.

Right. We have to embrace it. What we'll never do at OmniStar is depend on it. It'll never become a crutch and it'll never become something that replaces, right, the human side

Yeah.

of what we do.

Well, and when you embrace something, just to clarify, because I know how Phil thinks, when you say you embrace something or we're going to embrace something, it means that we're going to study it. We're going to learn it, not just what it's capable of, but what it's not capable of. We're going to learn its limitations, and we're going to fill the void, so to speak. And we're not going to use it as a crutch or fill anything in our client meetings.

When I say embrace, Spell's right, he knows a lot of my tendencies and the way I think. When I say embrace, I'm not saying that we accept it, right? But if we don't embrace it, it means we ignore it and that means we don't understand the strengths or weaknesses behind it, right? So I think you're exactly right. We have to embrace it so that we can learn from it and understand strength and weakness, and if we get to that point, then we know how to use it. And I think it can make us better in a lot of ways, but we have to know its weaknesses for sure.

You know what AI won't ever do, Phil? AI won't ever invite you to a wine tasting. AI won't ever call you when you've lost a loved one. AI will not call you when it's your birthday. AI will not invite you to an oyster roast.

It's true.

AI will not put you in a room with similar people, like-minded people, maybe clients of our firm. AI is not going to be able to do that.

Here's something else that AI is not going to do. You're exactly right. Those human touch points, they can't be replaced by a machine.

No.

Ever. But here's something else. I was a little bit late to our show today to get started because one of our Platinum Plus clients reached out with questions on a major transaction that they're dealing with. They didn't have to wonder what AI would say to do or try to push all kinds of data to AI and hope it gives them a good answer.

Yeah.

We got on a phone call or actually on a Teams call. So, we're looking at one another through a camera and we're talking and I can see the things that are bothering me. I can hear the emotion and I understand how to deal with that.

It's true.

So part psychology here, part science, being able to put that together, you can't replace that with AI and that client will tell you all day that they wouldn't trade that level of relationship to help them think through things. Don't jump off the cliff or make an impetuous decision. Rather, let's talk through it and think through all the moving parts. Sorry, you can't get that with AI. It just doesn't happen.

You know what? You're exactly right. To say it simply, AI can't hear the things that aren't said like a human can, but we can certainly see them.

That's right.

When you're looking at someone face to face, they say that a picture says a thousand words or it's worth a thousand words. We'll believe that.

Yeah.

Because when you look at a client, you see on their face what's bothering them. You see when they're happy. And unfortunately, AI has no idea.

No clue. Now, one day will they have facial recognition or rather emotion recognition? Maybe. But it'll never replace what a human can do. They're angry with the exception of basic task. Now, I think when that robot comes out that'll vacuum the house and do the dishes, I might be in. I might be in. But when it comes to major decisions, my money, my health, no.

Yeah.

No way. Not going to happen and AI doesn't threaten me at all when it comes to those things. In fact, I welcome it and again, we embrace it 100%. I think people out there that are depending on it in such a way that they're ignoring the human element, they are probably missing a lot of opportunities. Now, if they're young enough, Spell, they can probably get through it even with lots of mistakes.

Mhm.

But when you're a little older and you don't have all the catchup or makeup time, yeah, you better be careful what you allow to be fed into your mind. You better get good advice. And usually to get that it requires a