How AI is Transforming Retirement, Business, and Wealth Building | Ideas by Mike™ EP58

A lot of your decisions are not really probability problems. They're ethics and judgment decisions, and you're crossing your fingers and taking risks that you believe in, even though the probability says they might not be the right answer. Always keep your decision-making to yourself. Never let AI make your decision for you.

Well, Imagine standing on a stage in February, watching a guy explain artificial intelligence so clearly that a room full of financial professionals, people who explain complicated things for a living, leaned in at the same time to hear him speak. That was the first time I saw today's guest. A few hours later, I was sitting next to him with a panel discussion talking about AI and how people are actually using it today. and how advisors should be using it. He led teams at Google, McKinsey. He helped bring AI to a mobile ecosystem with over 2 billion users. He's been called the technology speaker for non-tech teams because his superpower is making complex things simple and useful. And today he's going to make AI simple and useful for you. This is the Ideas by Mike podcast. I'm Mike Milligan. Let's go. We live by 3 words on this podcast: educate, coach, play. And that's my personal mission. And Dan Chuparkov is going to demonstrate that for you firsthand. He's taken my philosophy, his knowledge, and is going to apply something for you. We had the 4th segment, a little bonus. He's going to give you a prompt at the end that you should be using today to create more productive life. Let's dive in. Dan, welcome to the Ideas by Mike podcast. One of the things I love about you from hearing you from the stage is how you use plain English to help people understand AI on a simple level. And so in this, as we get going here, you know, my personal philosophy is to educate, coach, plan, and I wanna use that for today's podcast. And so, so let's talk, let's start with educating. So could you paint a picture to maybe a listener. Let's say she's 61, she's smart, she's successful, she's a few years from retirement. She's heard about AI, you know, 400 times and how it impacts people from AI and her grandkids. But let's teach her a little bit. Let's give a brief history of AI and from wherever the story really starts up to today. Like you're sitting across the table from her.

Okay. Yeah, I think, you know, we do throw the acronym AI around quite a lot in the world. It's one of the things I say sometimes is that people are treating AI like it's a piñata, and there's all this cool stuff inside the AI piñata, but no one can really see what's even in there. So, we're all just sort of banging at it with a stick, right? And what I want to do for you and your listeners and my audiences is to help crack open that piñata a little bit. So, you can see, like, when we say AI, what do we What can we actually grab out of there and use for something useful? And the main, most important thing to know right now is the generative transformers that we're using. When you talk to ChatGPT, that G means generative. That means it's making something new that wasn't there before.

Mm-hmm.

The P means it's pre-trained. That means it learned a while ago and it's using that old information to do the best it can today. And then transform is the third word, the T in GPT. And that just means it's taking the thing you start with and turning it into something better, right? And so, the way that works is using probability. If I asked you to finish a sentence, we'll go ahead and do this, Mike. Tell me the next word in this sentence. Once upon a—

Time.

Amazing. That was great. When I do that in front of audiences, you can hear everybody say time all together at once because there's an easy clear consensus on what the answer to that question is, right? But a lot of the questions that we ask in the world don't have a single clear answer that we all agree on. And so when you ask a harder question, like, what's the next word in the sentence now? Once upon a time.

A long time ago.

If a might be the next word, some people say once upon a time there was a, once upon a time in, there's no single clear answer. Right? And so when AI is trying to figure that out, when AI is trying to figure out how to build your sentences, whatever those sentences are, it's using the most likely next word using probability. Sometimes that most likely word is only a 30% chance of being the best one, but it's the highest that AI has. And that's going to mean AI is going to have some low confidence answers sometimes. There's very little in the world that we all agree 100% on.

Right.

So when you ask a question that has any sort of randomness or variability, AI's gonna give you an answer, the best one it's got, but you might not agree with that, right? What should you have for dinner tonight? There's no right answer to that question. And maybe you feel like chicken and it says Chinese, right? And that's fine. There's no right answer and it's gonna do the best it can.

Why did ChatGPT feel like it came out of nowhere in 2022 when the technology had been building for decades?

Yeah, so that's interesting. So it was building over a long time, right? For the first AI that you really saw doing that thing was actually either in your Amazon cart where Amazon said people who buy that shirt, sometimes buy these pants.

Hmm.

Right? That's using probability to do the same thing ChatGPT is doing today. Or you saw it in Netflix. People who watch that movie also like this movie, right? AI back then was just predicting the next most likely thing that you would buy or watch. So that's how it started working. The problem was when the questions got harder than that, When the questions got like, it's not like of the 750 movies that are on right now, what's the most likely, most popular one? You're then saying like, I have a small business and I'm trying to figure out how to reach customers that don't speak English. What should I do? That's a way harder question. And so ChatGPT crossed a threshold in November of 2022 where they used so much more data to train that algorithm than the other AIs that came before it. It got way better at answering weird random questions that better than it even surprised the scientists that were working on it. And so ChatGPT, before there was ChatGPT, there was just GPT-3.0. Okay. And that was trained with about 16 trillion pieces of data.

Wow.

And that was a lot. And all of a sudden its answers got so great you could ask it almost anything. And that was, that was pretty cool.

I heard you say from the stage that you compared AI's impact to fire, kind of like what the cavemen had it like. Well, what's— that's a big claim. Just, just maybe a minute or so to defend that a little bit.

Yeah. So first of all, it was my CEO when I worked at Google. So I used to work at Google. 2 and a half years ago, I walked away to just help people through this transformation, maybe quit the safest job in the world to come help you on your AI journey. Best decision I ever made, and I'm grateful every day. But when I was there, Sundar Pichai, our CEO, said to us all the time, AI is more important for humanity than fire. It's a pretty bold claim. He used fire as an analogy for 2 reasons. One, it is impactful. It's super impactful. It will help us cook our food and keep ourselves warm and solve all kinds of new things. But also, fire's a little bit dangerous. If you don't know how to use it properly, right, you're going to, you're going to set some stuff on fire. And so that was a great analogy for those 2 reasons. One, it will be impactful. It'll change the way we cure diseases. It'll help us have a personal coach next to us, helping us get better at whatever we're trying to do in the world. It's going to help democratize access to information. People that only speak a language like that's not that popular can now read any textbook that it was ever written. That's kind of amazing and powerful. But also, you know, if you do the wrong stuff with it, you're gonna set your operations on fire.

So here's what I'm hearing, and listeners, if you are writing something down, taking notes from this, is I want you to hear that AI is not new and it isn't magic. It's And it isn't out to get you, right? AI is not out to get us.

Right. Yeah. AI really, so first of all, AI is just a database. It's really just a big Excel spreadsheet, humongous. It's the biggest Excel spreadsheet ever, but it's just a spreadsheet. It's never doing anything. All that can happen is a person can look something up in that spreadsheet and AI will tell you what the spreadsheet says. So it's not thinking anything when you're sleeping. It's not plotting humans' demise. It's just a big database and you can look stuff up anytime you want. So one, don't personify it because it's not that. It's just a statistics table that's giving you the most likely answer to any question. And that should ease some fears a little bit.

It will. And I will say that another way. Because AI is a tool, just like a hammer, just like an annuity, just like a Roth conversion. It's complex, but it can be taught. And thanks for opening today's podcast by teaching it.

Yeah, yeah, no problem.

So let's go to coach. Let's pull out a play or a pattern, right? That you see that's applicable. So I coach financial professionals through the One Oak Collective. When we served on the panel together, had a lot of people come back at me because we implement AI in our practice. But here's what I learned. People don't need 17 tips. They need one play.

Oh yeah.

They just need one play they can actually run. So if you could install one play, one pattern, one habit into every listener's life, that would make them successful with AI, what is it?

I think the first most important thing is to look at what you spend the most time doing and then try to find ways to make that a little bit more efficient. What a lot of people do is they go in using AI to try to use it to do the most proprietary thing that they ever do in the world. Like you're a financial planner and you're trying to figure out what stock should I pick To tell all my customers, right, to invest in this particular stock. Spoiler alert, probably Nvidia, right?

Probably. Maybe Apple this week.

Yeah, sure, right? And that changes all the time. That's a very proprietary thing. It takes expertise, it takes opinionated judgment and all that stuff. Maybe AI will help you do a little bit of that, but that's something you're maybe doing for like 1% of your week. And then the whole rest of your week, you're sending emails to leads or responding to customer questions or having discussions with your stakeholders or your customers or the people around you. You're sitting in meetings. You're not spending your whole entire week trying to find Nvidia stock. Most people are picking a thing that they only do a tiny sliver of their week. And then they don't see the ROI. They don't see the time savings because they're tackling something that they only spend 9 minutes doing. If instead you look at your work week and start time tracking just for a couple of weeks to figure out where's your time spending, what is the thing that you're spending the most time doing? Can you optimize that? And I bet it's communication.

Wow.

So, take your communication, ask your favorite AI assistant, how can I make that more efficient? And one of the things it's going to say is you should start recording your meetings. You should start creating a digest before the meeting even happens. You should send somebody the summary of the meeting that's probably going to happen, and you might not even need to have the meeting, right? Communication is probably where the actual ROI is, and most people aren't looking there.

And I want to, you know, I want to take what I hear you say it and apply it a little bit. The human still stays in charge to play. That, because in business, right, the human, if the human's not in charge, then you're, then you can be replaced. Uh, what's the most common way that people run this play wrong when they try to, uh, when they try to use communication?

Yeah, so, uh, most of them, uh, they take the meetings that were already gonna happen, the meetings that are gonna happen anyway, and then they Also add meeting notes. Now your meeting still happens, all the same people still go to it. Now you have meeting notes, and now also you have to look at the notes. That's more work. That didn't create an efficiency. What you want to do is have the smallest number of people in a meeting that you can.

Right.

Create notes, share those notes with people that skipped the meeting, right? That's the way to make it more efficient. Start creating a digest for people to consume so they don't have to go those discussions. So that's the first way people are messing up meeting notes. They're just adding more work to their plate instead of making it easier or more efficient.

Dan, that coaching was amazing, right? I want to compare it to one other thing, right? When Tom Brady was the quarterback of the New England Patriots, right? Bill Belichick never asked him to be a coach. He guided him so Tom Brady could be who he wanted to. You didn't say there in that little coaching segment for people to become programmers. You didn't say for people to become experts at AI. You gave a practical tool to help everyday life. Is there anything you wanna kind of close with on that coaching aspect?

I think, you know, one, figure out just one thing that you could do better in your life, and then ask AI how to do that better, whatever that is. You know, everyone in this audience has a slightly different thing that they're trying to do for their customers or their families or their own life. And so, pick one thing that you're gonna try to apply to and focus on that and see if you can find the ROI there. Don't try to boil the ocean. Don't try to fix every part of your life all at once. Pick one thing that could be better if you focused on it and tackle that first.

Do you have a favorite prompt, by the way, that you would prompt AI on something like that?

Yeah. So I'm gonna save one until the end of this talk. I'll close with it. But the other one that I really stumbled into accidentally that I now use for everything is, hey AI, I just wrote this LinkedIn post. Compare that LinkedIn post to every other LinkedIn post I wrote and rank it. Is this the best post I ever wrote or the worst or somewhere in the middle? I rank everything I do. I'll take the recording from this podcast and I'll say, is this the best podcast I ever did or the worst or somewhere in the middle? And it'll give me a number. It'll say, this was number 17, Dan. And I'll say, what could I have done to put it in the top 5? And it will say, hey, you forgot to mention your core thing. And that's what I do. I rank everything I do. Instead of asking AI to do work for me, I do the work myself and ask AI to rank how I did.

That's what a coach gives you guys, by the way. Because you think about, right, a coach's job is to win games, to get better effort or better better play out of their players. That was a freaking nugget right there. That was good. Let's go to plan, right? So educate, coach, plan. This is where I live. I believe a financial plan's the most important financial document you have in your life. It's got to be clear, concise, and it has to be written down, simple to understand.

Mm-hmm.

We call that the one-of-a-kind financial plan for your one-of-a-kind life. So let's talk about AI in the planning process. I have my thoughts on it. I think it belongs there. But 3 doors. I want you to pick one of these to start. So starting a business, retirement planning, or financial planning. So starting a business, retirement planning, financial planning. How should AI actually be used in planning and where should it not be used?

Right. Okay. So I guess, you know, we'll focus on retirement planning because, you know, as a 56-year-old guy, I'm trying to figure that out. Like, how many years do I have of income generation? What are my kids going to inherit from me? How many years of retirement will I have? All those are hard things to figure out. And so, you know, I have a plan for myself and I feel kind of good and it's working out so far. So I don't have any critical problems to solve, But one of the things prior to AI that used to be the current state of the world is you would make a plan one time and then you would see how it goes. And maybe yearly you would take a look at it, or maybe financial plans, maybe every 3 years or sometime farther out than that. And you would then pivot, make changes to your plan, rebalance, whatever replanning looks like. Replanning is costly and time-consuming and it takes discussion. AI can help you replan a lot more frequently, right? For my own portfolio, I would rebalance once a year, right before the end of the year. I'm trying to figure out what that might look like these days. You might be able to use AI to rebalance every week, constantly tweaking your portfolio, constantly addressing like, you know, what positions are overweight now for you in your own portfolio. You can have a real-time analysis coach next to you telling you how things are drifting in real time. And that's dangerous for some people because they're bouncing around, they're chasing the trends, and they're, you know, 3 days late on every one of them. So that shouldn't mean you change everything all the time, but you should have real-time awareness of your performance toward goals using an AI assistant. I still think the judgment should be with yourself and your own human financial professional helping you, but like you should have visibility in real time.

Yeah. Here's my, here's my thought on AI and see what you think about this is I believe AI can be used when you're beginning the relationship with a financial advisor.

That's a great way to start.

You know, if you're looking for a new financial advisor, put in questions, put in that firm website, put in the data and say, what questions based off what you know about me should I be asking this advisor to optimize my outcomes? And it's, you'll be shocked at like when AI will give you questions you never thought to ask.

Yeah, yeah, for sure.

Right. And that's a great part of the play. I'm glad you mentioned about keeping the human involved there. Is there a line though? Is there a line that some people cross where they over-rely on AI?

Oh yeah. I mean, for sure. People, especially as we start talking about what an AI agent is, right? And for the non-technical listeners, previous to agents, our AI assistants were just chatbots. We could have a conversation with them. We could say, hey, ChatGPT, what stock should I buy? And it would say, you should ask a financial professional. If it doesn't say that, financial professionals should write more blog posts that say, always have a financial professional.

That's what we said on the panel, right? Is we actually, we were actually, I was up there defending that if our profession doesn't put more blogs, more illustrations, more portfolio out there, Then AI may take more of our job. We have to actually, we actually have to feed the spreadsheet more information so we stay relevant.

Right, right, right. That's important. But back to, back to your question. In an agent world, the agent can actually go to E-Trade and buy a stock for you. And in that world, you can start having AI talk to itself and buy its own stocks on your behalf. And then, then you get this spiral that spins out of control. gradually drifting from what you intended. And that's a mistake. I think that people should always, always, always treat anything from AI as just a recommendation to consider, right? Never let AI make your decisions for you, right? Because a lot of the decisions you make are based on stuff in your head that your AI assistant doesn't know.

Right.

A lot of your decisions are not really probability problems. They're ethics and judgment decisions, and you're crossing your fingers and taking risks that you believe in, even though the probability says they might not be the right answer. Always keep your decision-making to yourself. Never let AI make your decisions for you.

That's really good. That's a great way to end like that little planning segment. I want to turn to the future now. And we could, I could probably let you start here about the future, walk away for an hour, and you would still be talking about what the future of AI is. But you've lived, you know, working at Google, you had an insider view to a lot of change in technology, you know, the web and the cloud and mobile and now AI. So, You've earned, I think you've earned the right to make this prediction. What does AI look like 3 to 5 years from now?

I think if you make a list of everything you do in the next month, somewhere in a sheet of paper, in a document, wherever you like lists, put a list of everything you do. In 3 years, everything on that list will be different. For sure. You're not gonna read emails one at a time anymore. You're not gonna have as many meetings as you have. You're gonna be talking to people that speak Japanese, even though you don't speak Japanese. Like, everything's gonna change. And so, one, that everything will be AI-assisted. Some things will be AI-automated. Some things will be AI-prioritized, but it's gonna touch everything. The second thing that I will say is it's going to kill the sharing of information as we currently know it, right? You can already sort of see this on LinkedIn. If you go to LinkedIn, almost every post there is AI-generated. It's got the same formula. It starts with a hook. It's got bullets and emojis. It feels cliché at this point. And so AI is pretty quickly potentially killing LinkedIn because people won't go there anymore because it's not authentically human. That'll also happen to your email, right? If every email is generated in the middle of the night by the AI email generators, then the stuff in your email is not going to— the only way to talk to an actual person is to pick up the phone and call them. And so, you know, email will die, LinkedIn will die, social media networks will stop getting news from Facebook because all of it's fake. That, you know, those things are going to transform pretty remarkably, I think. Wow.

2 last things as we kind of land this plane. First, what's one AI tool everybody should try this week if they don't, if they're not currently using AI?

I think if my hypothesis that most people are spending most of their time in communication, I think first and foremost, I think you need a note-taker of some kind. Most of the note-takers are all using ChatGPT under the covers anyway, so it doesn't even really matter which note-taker you pick. I use Otter. Some people use Fireflies. If you have a Copilot Teams instance, there's a Teams note-taker too. Find a note-taker. And then start recording the things you care about. If they're customer calls, if it's okay with the person you're recording, record customer calls and then ask AI to rank those customer calls and tell you if that was a great call or a mediocre one, right? Record your— when you try to make a presentation to customers or to stakeholders, record that presentation and get your own feedback on it. Recording, I think, is the first most impactful place to start. It'll transform the way you speak.

I love that. All right, you teased this a little bit earlier. You said there was one prompt that you were gonna end with today. Let's hear Daed's go-to AI prompt for our listeners.

Yeah, it might feel trivial and basic, but it's the most impactful prompt I ever typed in. It was about 20 months ago now, almost 2 years ago. I typed a prompt to ChatGPT Pro and I said, hey, ChatGPT, give me one new AI thing that I should try Monday. And then it said, well, for this week, try this. And then next Monday, I got a new suggestion and it said, hey, try this. And then, you know, 3 months in, I had, I don't know, 16 AI suggestions that I hadn't even thought of. And a couple of them didn't apply and I didn't do them, but a couple of them were remarkable. And so I switched it at about 4 months in. I said, hey, do this daily. Every day, give me a new thing. And I have this one thread, it's just 20 months long with maybe 450 AI tips. That thread is mind-blowing. Like, I just scan anytime I'm trying to figure out how to do something, I can go back into that thread and look for a good idea. And it's changed the way I do work. It's maybe 10x'd my business, like financially, like It's the most impactful thing you could possibly type into any AI tool.

I just did it. Sweet. I'm going to share it in the closing, what it came up with for me. But this has been a fascinating conversation. I know our listeners are going, they've obviously learned something because you are a master of your trade. But you did give up a very secure job at Google. I want to make sure people heard you say that. You left Google. a secure job, which by the way, Google hasn't really stopped, uh, stopped growing since you left.

Yeah.

Right. Which I, which I'm sure you're still benefiting from a little bit because a stockholder. But, uh, but Dan, thank you for your time today. I appreciate the, the you showing up. I hope that when you rate this podcast, AI, if you're listening, Dan's AI, make this podcast number one ranked podcast on his on his, uh, his list of podcasts. But it was absolutely gold today, Dan. Thanks, thanks a lot for your time.

Yeah, it was a pleasure, pleasure talking to you this time and the time we talked before, uh, almost a year ago. Uh, it's been a pleasure getting to know you and, and your audience today. Let me know if I can ever be helpful in any of your futures.

Is there, is there any way anybody could go and just hear you talk more about AI?

I know you do it on the stage primarily, but are you out on the internet anywhere talking about Yeah, I'm pretty easy to find on the internet if you can spell my last name, which if you're in video, you can see it there. C-H-U-P-A-R-K-O-F-F. Chuparkoff. I'm, of all the people in my family, I'm the best at search engine optimization. So you'll find me YouTube, LinkedIn, Instagram, my own website. Any of those ways are great ways to get ahold of me.

So we're not gonna play hide and seek for Dan. You're just gonna be able to To put him in your AI and it's gonna tell you exactly what to, what to see. So, Dan, thanks a lot for your time today.

Thanks, Mike, for having me. Appreciate it.

So at the end of our conversation, Dan Shmurkoff, uh, who was absolutely amazing on the podcast today, gave us a prompt. to put in about what AI tools we should use. I did it. I put it in the system and, you know, it gave me a whole list of things to do over the next 12 weeks. I'm a big believer in the 12-week work year. But folks, as we landed the plane on the podcast, I wanna just give a recap. AI isn't magic and it isn't a monster. It's the next chapter of the story that's been building for decades. Second, we talked about you don't need 17 tips to begin using AI. You need one play that you run every single day. Start small, stay curious, keep your human judgment in charge. And then we talked about how AI can accelerate a plan, but it can't replace the planner, the human. Get your plan written down. Clear, concise, simple to understand. And the future, the future belongs to the people who lean in at any age, at any stage. Because here's the thing, the whole point of every tool, every plan, every technology, is to give you moments in your life because my personal mission is never miss a moment. This has been the Ideas by Mike podcast. I'm Mike Milligan. I'm an investment advisory representative with Advisory Alpha LLC. Ideas by Mike, the podcast, is owned by Mike Milligan LLC and Ideas by Mike doing business as One Oak Financial. Mike Milligan LLC and One Oak Financial are both unaffiliated entities of Advisory Alpha LLC. This podcast is for Educational purposes only. Do not take anything we said today and make stock recommend or stock purchases without first consulting with a licensed financial professional. Until next time. Podcast.