AI and ChatGPT Skills Every Finance Expert Needs to Automate Data Analysis and Reporting |Tom Hinkle
In this episode of Future Finance, host Glenn Hopper talks about the intersection of AI, data analytics, and finance with guest Tom Hinkle. Tom is a Microsoft MVP and analytics expert who shares his unique knowledge about AI’s impact on Excel, data science, and workforce analytics. He also discusses his Minesweeper project in Excel and the changing role of finance professionals in an AI-driven world.
Tom Hinkle is a Microsoft MVP with 20+ years of experience in analytics, data science, and business intelligence. Based in Charlotte, North Carolina, Tom has worked with major financial institutions, including Bank of America, Wells Fargo, and TIAA. He is known for his problem-solving mindset, technical expertise, and ability to bridge the gap between business and technology.
In this episode, you will discover:
How AI is revolutionizing Excel and what that means for finance professionals.
The role of AI in data cleansing and automation, making finance and analytics work more efficient.
Why the best coders may not be the most valuable team members in the future of finance.
Insights on Copilot in Microsoft 365 and how it’s changing financial modeling and reporting.
The story behind Tom’s Excel Minesweeper project—how he built a game in VBA using AI.
As AI continues to automate repetitive tasks, finance professionals must adapt, refine their analytical skills, and focus on storytelling to translate data into meaningful information. Tom’s journey highlights how embracing new technologies like Microsoft Copilot can enhance productivity rather than replace human expertise.
Follow Tom:
LinkedIn - https://www.linkedin.com/in/tomhinkleclt/
Website - https://excel-cafe.teachable.com/
Minesweeper Game - https://www.thefpandaguy.com/new-page-3
Join hosts Glenn and Paul as they unravel the complexities of AI in finance:
Follow Glenn:
LinkedIn: https://www.linkedin.com/in/gbhopperiii
Follow Paul:
LinkedIn: https://www.linkedin.com/in/thefpandaguy
Follow QFlow.AI:
Website - https://bit.ly/4fYK9vY
Future Finance is sponsored by QFlow.ai, the strategic finance platform solving the toughest part of planning and analysis: B2B revenue. Align sales, marketing, and finance, speed up decision-making, and lock in accountability with QFlow.ai.
Stay tuned for a deeper understanding of how AI is shaping the future of finance and what it means for businesses and individuals alike.
In Today’s Episode:
[01:46] - Introduction to the Episode
[03:36] - Meet Tom Hinkle
[06:07] - The Future of AI in Finance
[11:53] - AI’s Impact on Excel and Financial Modeling
[15:31] - The Skills Professionals Need in an AI-Driven World
[24:02] - Building Minesweeper in Excel with AI
[29:51] - Exploring Microsoft Copilot & AI Coding Assistants
[33:07] - The Importance of Storytelling in Data Analytics
[36:54] - Fun Segment & Tom’s Personal Quirk
[40:53] - Closing Thoughts
Full Show Transcript
[00:01:46] Host 1: Paul Barnhurst: Welcome to the future finance show. This week we are joined by Tom Hinkle. Tom comes to us from Charlotte, North Carolina. He is a Microsoft MVP with a strong analytics and data based background. He is a versatile problem solver with a passion for tackling complex challenges across industries and disciplines. With 20 plus years of experience, he brings a logical mindset, a deep understanding of business processes and an adaptable approach to every project. Tom thrives on identifying gaps that hinder growth and efficiency, leveraging any tool at his disposal to drive meaningful solutions. Beyond his technical expertise, Tom's true superpower lies in his ability to build strong relationships and foster collaboration. His knack for clear communication, bridging the gap between technical and non-technical audiences, has made him a trusted mentor and leader in cross-functional projects. Always ready for the next challenge, Tom is committed to making an impact wherever he goes.
[00:02:55] Host: Glenn Hopper: Welcome to future finance. Paul is a little bit under the weather today, so the dynamics of the show are a little bit different. You've got the Glenn and Tom show today, so we'll be diving in solo here. So I think we are going to use Paul's intro though. So that was before when Paul was feeling better. But now we've got Glenn and Tom. So Tom, welcome to the show.
[00:03:15] Guest: Tom Hinkle: Hey thanks. Great to be here Glenn. I'm looking forward to having this conversation. A lot of interesting things happening in the tech space these days.
[00:03:23] Host: Glenn Hopper: Yeah, absolutely. And it's funny, this conversation, this is our second time we tried to get it going this week. I know you and Paul had been trying to get something going for a while before that. So at least half of the future finance team is here and we're excited to dive in. So I guess to get started, tell us a little bit about kind of your background, what you're doing now and how you got here.
[00:03:44] Guest: Tom Hinkle: Well, I'm just going to tell our viewing audience, I'm over 30. If you couldn't tell that already. I know I look young, but many moons ago, I got a computer science degree from NC state. And you know, these degrees change. Now you have data science degrees, data engineering data things. A lot of things I think was similar. We focused a lot on algorithms, and I focused on database technology and how to work with data and use that to drive applications.
[00:04:10] Guest: Tom Hinkle: So I spent, you know, first ten years or so in tech space. And then things changed around nine over 11, and I moved into the business side in banking and finance, where I really became focused on reporting and analytics. So for the last 15 or 20 years I've been in different banks. I'm in Charlotte, North Carolina, there's Bank of America, Wells Fargo, TIAA, Truist, you know, they're all here to some degree. So that's kind of where I am now. I'm still doing data and analytics, trying to learn the new tools. But it's funny. It's like many sports, even the baseball pros, when they go to warm up on their practice, they just start by throwing the ball back and forth to each other and then maybe throwing a couple ground balls to pick it up. The basics are always going to be in whatever tool you choose. So that's kind of where that's where I ended up. And I enjoy the work I love. It's like playing a games like solving problems. So that's really fun to whatever tool you use.
[00:05:01] Host: Glenn Hopper: Yeah. And it's I know, you know, one of the first things I think before Paul even told me who you were, he talked about this, the minesweeper game that you built in Excel. I was just ready to, you know, have one of Paul's Excel geeks come on and primarily talk about that. But it's funny that you say, you know, the tools don't matter because I think, you know, was it Satya Nadella has been saying tools are about to get a whole lot less important if you believe what he says, which basically his comments were around SAS.
[00:05:28] Host: Glenn Hopper: And the idea is if all this software is just a UI layer on top of a database, and really what we need is that database. If agents come about or when agents come about, depending on how optimistic you are, when agents come about, that UX layer isn't going to matter. It's just going to be here's our data schema and tell the agents to go do this for us. So I think that agnostic approach to tools is about to like really come front and center and it's going to be so with your background in analytics and database design being tool agnostic and just knowing that you've got to get those key, you know, data science principles and how you're going through and working with the data. What do you think is going to happen in areas like workforce analytics and compliance reporting? I mean, where do you see these trends going? And maybe you have to answer it two ways. One is while the human is still the main driver of this, and two, what if these agents take over? And if we get closer to AGI and generative AI and or not just generative AI, but AI in general, you know, are taking over a lot of what we do. What do you, where are you kind of reading the tea leaves right now on that?
[00:06:32] Guest: Tom Hinkle: So I look at it, AI is an incredible tool. And I think it's one of the paradigm changing tools that are out there. You know, there's a couple different things I've seen in my career. One of them was I'm old enough to be around before what everybody knows is the modern day World Wide Web, the internet. I was around when it started and people were just writing simple HTML. There was no interaction in pages. That was a game changing technology. A couple of years later, the first smartphone came out, the iPhone. Now, it took a little while for people to really realize the potential of that, but that became a huge paradigm shift in our technology world as well, because people now could. You didn't have to go into a meeting or do a share to get somebody's information or report you have on your phone wherever you want. And somebody could design a mobile report and send it to your phone. And I think AI is the same sort of thing. Now, for me, I look at AI differently than a lot of other people do. I think I think a lot of people are really wowed by, oh, it can do all this magical stuff and everything here for me. What I see it doing is all the rudimentary kind of grunt work, if you will, that we all know how to do And they had to do and we've got proficiency with it.
[00:07:48] Guest: Tom Hinkle: There's one example I'll share. When Excel 2007 came out, they added a button called Remove Duplicates which and I'm just going to talk Excel a lot today because whatever tools we work with with Python and Power BI and Tableau. Excel is always a part of every solution. It still is. And I'll get in religious discussions with people on this. But but it is. But this button that came out in Excel 2007 was removed duplicates from a table. You could define what a duplicate was. Is it one field or multiple fields? And you click a button and it removed all the duplicates. That seems really trivial today because we've all lived with it for 1520 years. But before that, I had to have a Visual Basic macro to run on a table to remove duplicates or do it manually by sorting, looking through them and deleting the rows. So that button became a tool to just take a bunch of busy work out in one of the data cleansing things is removing duplicates. That's what I see. AI is going to do across the board. You asked about a couple industries and a couple things, but I think that's really where you're going to see. The most benefit is that, you know, setting up a FP&A model or actually forgetting a couple of the examples you asked. But think about these models.
[00:09:02] Guest: Tom Hinkle: There's, you know, if there's ten steps, there's like six or 7 or 8 that are pretty standard that you do on every single model before you get to the specifics of this one. Well, AI is going to be able to do all that stuff for us with very, very minimal interaction to the AI tool where you say, hey, I want you to go clean my data. I want you to remove the duplicates. I want you to identify any categories that seem similar and call those out to me, and present them separately so I can make a decision. And I want you to make sure all the numbers are formatted to two decimal places. That's going to be the extent of your cleansing, of your cleansing coding. And it'll be done. And that leaves this can free the analyst up to spend more time on insights and finding patterns and doing some of the more exciting stuff in the AI space. But for me, the game changer is just taking off all that busy work. That's just it's trivial to do.
[00:09:56] Host: Glenn Hopper: Yeah. And I think already like cleaning data sets outside of Excel. I frequently one of my use cases for ChatGPT all the time is I will take a data set in a CSV and have, you know, because whether it's date formats or missing values and just all the all the data cleaning that you have to do, generative AI is pretty good at data cleaning. And that's and that's before it's fully integrated via Copilot into Excel and everything. And it's just, you know, Microsoft didn't invest whatever they're up to now $15 billion between across the different their different investments. You know they're going to get this right. Copilot is going to get there. And I don't know if you're having any luck with copilot right now, I'd like to hear it. But I guess kind of on that vein, you as a Microsoft MVP and a big advocate for Excel, and I'm right there with you when you know, I love to do stuff in R or Python, not R as much anymore, maybe more Python, but the place that the place that I start is always Excel. I mean, everybody does visualizations and all that and yeah, and I don't think Excel is going anywhere, but how do you see like and I'm honestly I have not. It's kind of like dipping my chocolate in the peanut butter. Like I'm not a great coder, but I'll do some stuff in Python. I have not messed around at all though. With Python in Excel I can't. I can't join those in my brain. But I see with Python and Excel and with copilot, you know, getting there and I know it'll get there soon, but what do you see? How do you see AI changing Excel in the coming years? And maybe it is just like I see like all the formulas that we that coming up in my FP&A background, you know, I was very proud of your complex formulas.
[00:11:42] Host: Glenn Hopper: You did and the nested if statements and all that. And I feel like maybe that's going to go away, like, yeah, we're going to you're just going to in natural language, just tell it what you want to do and the AI will go do it. But what are you what are you seeing? What do you think is going to happen to excel with AI?
[00:11:55] Guest: Tom Hinkle: Sadly, I agree with you, Glenn. And coming from a coding background, I'm also really proud because you know what? Any room I'm in, I'm the best coder in the room. I'll tell you that right now. No, I think that's that goes along with coding. It's not arrogance, but there is an ego to many good coders where they really, you know, they found some ways they everybody's got a different style and they've tweaked it and, and done some things. But yeah, we're going to have to let that go a little bit. You know, back to my example about the Remove Duplicates button. I made a macro that went through and removed the duplicates at a touch of a button. It wasn't necessary at all anymore because it was a button. You press on the ribbon. So I had to let that go. Now, I still think you're going to have to have the technical knowledge to do it, because AI is not magic.
[00:12:45] Guest: Tom Hinkle: It's, you know, based on an LLM, which is, you know, large language models, which goes out and looks at everything that's been done in the internet, everything I say, but, you know, looks at a large set of information and based on that comes up with some recommendations. So it's not always going to be perfect. And you're going to have to look under the hood, or sometimes you're going to have to test it and see what was wrong. And for that you still need the knowledge. But to your point, we're not going to be sitting there in front of our computer for, you know, half a day or two days coming up with this really, really robust and elegant formula that solves our problems. We're going to have AI do it, and it's maybe not going to be the way we do it. But if you focus on the outcome, hey, it's not the way I do it, but I got it done in a 10th of the time and it works. Now you have a different metric. Now you can do ten times as many things in a work environment as you did before. Other than solving a really difficult problem that took you ten hours to do, now you can solve ten problems in the same amount of time, but it's not going to have your personal stamp on it.
[00:13:47] Host: Glenn Hopper: The interesting thing to me, so the day that we're recording this, either this morning or yesterday, I saw it this morning, the OpenAI just rolled out their first commercially available agent operator, they're calling it. So that's, you know, operator was level three of their, of their roadmap. And with agents going back to Satya Nadella again, I don't know why I'm just quoting him exclusively in this episode, but his thinking around agents is when you hire someone, you're hiring the person. And how did he put it? If you're hiring a data analyst or a financial analyst, you're kind of hiring the analyst plus plus his or her spreadsheets. You know what I mean? Like, you're coming with the skillset or whatever work experience, too. Yeah. So. And what he was saying is, in the future, you're going to be hiring a person and the agents that they've created. So if you've got, you know, just like building a really cool macro, if you've got a great workflow that you've, you know, you've built this special agent that does month end close and the, you know, budget, actual report or whatever. And but what's interesting to me to think about is what for the human in the loop and assuming, you know, we're not all just checked out at the on the holodeck and doing our thing while the robots do all the work for the human in the loop, I think, like you said, for domain expertise one if you're a finance person, you know the right questions to ask.
[00:15:13] Host: Glenn Hopper: You know when it's appropriate to use. I think we were talking about it the other day when it's appropriate to use EBITDA versus net income. And, you know, you know, sort of the specifics and, you know, the difference between, you know, kind of your, your gross margin and your net margin and all the things that come with being a finance person. But what I think is going to be more important is and this is with your analytics background, is people having data science skills. You don't just intuitively know that correlation doesn't mean causation and you don't know you know how to you know all the stuff that goes with that and how to sort of look at data and what it what to think about a histogram and standard deviations and all that. So I guess all that is to say, if being the best coder in the room isn't your claim to fame anymore, what do you think? What do people need to be thinking about? People that are not as, you know, young people like Paul, not old guys like me and you, you know, say.
[00:16:08] Guest: Tom Hinkle: No, I think I think with young people. And I was this way too. Like, they have no fear. They have no barrier for fear. They think they're going to change the world. In many cases they can. But so I don't think we have yeah, we have to worry about guys like you and I and what the next step is because they just fearless.
[00:16:23] Guest: Tom Hinkle: One of the big benefits. But I think what it's going to do You hit the nail on the head there. Like, I know some statistical stuff. I can go in, I can hit. Well, now it's ChatGPT. But, you know, you and I used to live in Stack Overflow, I'm sure. And, I can go and I can get some some code to write something that'll do some cluster analysis or some correlation code or things like that, because the AI is going to do all that work for us that we were normally doing 80, you know, 80, 85% of our day was just doing this type of work. It's going to free us up to have more innovation and to be able to learn more of those statistical models, learn for, you know, learn more of those advanced things and then also learn how we plug those into the business. That's the innovation part. Ai is fantastic. It gives you a lot of answers. And we might get pushback from this, but it's still not smart. And that's kind of counterintuitive. By saying artificial intelligence, if you think about what it's doing, it's going out to, like I said, everything on the internet grabbing everything and analyzing it, and based on what you put in and saying, based on everything I see out there, this is what I think the answer is going to be. So it's still deterministic.
[00:17:39] Guest: Tom Hinkle: It's going off of past things that have happened. It can't invent something brand new out of thin air. Now people are probably going to come at me and be like, what about the AI generated pictures? Well, I don't know how that algorithm works, but I guarantee you it's something similar. You know. Hey, draw. I want you to draw a picture of me, you know, standing on the moon with the sun behind me and a dog walking around me. It's going to pick up all those keywords and go look at all the pictures it's pulled in with those keywords and synthesize something because of those. But there's not really innovation there. It's analyzing data that's already out there, but we get to really flex our muscles and become more innovative with what we can do. Sorry if I ramble a little bit there, but that's kind of where I see that the opportunity lies when we get I, it's like, oh no, I'm not out of a job. No, I get to do cooler stuff like, I can, I can pass off this stuff to the AI bot or agent and, and then I can come up with some neat new things.
[00:18:47] Host: Glenn Hopper: Yeah. I was talking to someone the other day about, you know, AI is great, but human in the loop is very important. And it's, you know, Yann LeCun said the current, you know, where we are with LLMs right now, they're as smart as a cat, basically.
[00:18:59] Host: Glenn Hopper: You know, it's like they're not, it's not like you said, it's not true. You know, it's weird because if you listen to this sound bites, you know, oh, a PhD level reasoner. Well, the only reason it's PhD level is because it's referring back to papers that were written by PhDs. And it, you know, and it has access to all this data. So it's, you know, it's the world's deepest generalist, you know. So if it's read all, all, you know, collective digital data that's out there. Of all history, you know, anything that we've digitized is what it's trained on. I mean, that's kind of inconceivable. But to your point, we're not seeing these breakthrough new ideas. And the best way for me to an example of that is if you get these things to write, they can write an email, they can write a meeting agenda. But they're so by nature going to be cliche because it's the statistical probability of what's most likely to come next. So there's phrases that are yeah, but if you get it to write anything like it just catches these phrases that and it just sounds so anodyne and watered down because it's just taking.
[00:20:06] Guest: Tom Hinkle: I'm going to share with you my favorite prompt, because I use ChatGPT just about every day. And let's talk about copilot in a second, too. But, I use it a lot for communications. And here's the other thing that's helped me with when I'm making a PowerPoint deck or an email that needs to go to leadership or something.
[00:20:20] Guest: Tom Hinkle: You know, the older you get, the more cognizant you are that the communication has to be tight, robust, has to grab people, use the right words. I've kind of stopped. I've kind of stopped really trying to craft those exotic emails. And I'll write an email that has the points I'm trying to get across, like, and it'll be an email, but it'll be, you know, something that I'd have to go back, you know, over for an hour or two before I send it out. I'll just take that email. That's not perfect. Go to ChatGPT. And my prompt is make this better. And then just paste my email in. Now, like you said, a lot of it could sound cliche. So again here's that aspect of I'm going to go back and read what it said. And I will tell you just this is just a guess, but I'd say 85 to 90% of the time I send the email. ChatGPT gets back to me. And here's the rub. Like some people might say, oh, that sounds like AI in there, but it truly sounds way better than what I wrote. But it's the same information. And here's the kicker. I wrote the substance of the email to begin with. So I'm I don't feel like that's cheating. I feel that's using a tool to make my product better. No different than using a spell checker.
[00:21:33] Host 1: Paul Barnhurst: Ever feel like your go to market teams and finance speak different languages? This misalignment is a breeding ground for failure in pairing the predictive power of forecasts and delaying decisions that drive efficient growth. It's not for lack of trying, but getting all the data in one place doesn't mean you've gotten everyone on the same page. Meet qflow.ai, the strategic finance platform purpose built to solve the toughest part of planning and analysis B2B revenue. qflow.ai quickly integrates key data from your go to market stack and accounting platform, then handles all the data prep and normalization under the hood. It automatically assembles your go to market stats, make segmented scenario planning a breeze, and closes the planning loop. Create air tight alignment, improve decision latency, and ensure accountability across the teams.
[00:22:40] Host: Glenn Hopper: Yeah. And you know and that is one thing it is very good at is synthesizing and organizing data. Because whether you take meeting notes or I know several people who will, you know, open up a recording app on their phone, walk around talking to themselves like sort of free flowing ideas, and then take the transcript of that, dump it into an LLM and have it synthesize it and come up with their action items to do list and, and all that. And it's very good at that. So it's about how you use it. And if you're writing a marketing copy, you can probably get a rough idea of where you want to go. But then, you know, take out, just keep that human in the loop and refine and go from there. But if you're sending a company wide or you know, an email to a management team or to, you know, other coworkers, you're not trying. You're not trying for a Pulitzer. You're trying to be effectively communicating. And I think that to your point, these tools are great at organizing the data and putting it out. So, yeah, it's that exoskeleton thing. Well, I do actually. So we alluded to this earlier and I'd love to include this in the show notes. And before we get too far into the robots in space, I do want to go back to Microsoft Excel because you built that a minesweeper game in Excel using VBA.
[00:24:02] Host: Glenn Hopper: And I love just what a unique challenge that must have been. And like I said, I want to put it in the show notes so that our users can check it out. But tell us about building the game. What led you to do it? Yeah, just give us some background on that.
[00:24:17] Guest: Tom Hinkle: Yeah. You know, I'm older, but I'm a constant learner. And, you know, with all the tools we've had for analysis, analysis stuff. I used to be a fantastic coder and I did a lot of C, and then I really got into VBA and like, really deep, like, you know, one of the eight people on the planet understands class modules and what they're for and why you don't need them very often and things like that. So this was a project for me to just kind of like, let me see if I can rebuild minesweeper within Excel, because Excel kind of has an interface that lends itself to be like a grid. And, you know, if you're not familiar with minesweeper, it's similar to the battleship game we had as kids where you find things. So this project started off as something I was going to do just on my own to just have something to work on. Throw in a portfolio. That's the new buzzword for looking for work these days.
[00:25:06] Guest: Tom Hinkle: You got to have your portfolio. And then before I got started. Many times. You have a project like that, you know, it's a little bit of time going through it in your head how you're going to, you know, flow out the program, what you're going to have to do. And when I got it done, I said, hey, why don't I see if I can do this in ChatGPT? Let ChatGPT write all the code for me. And so I did it, and it was done in three hours and there's nothing in there I couldn't have done myself, but it would have taken me a few days to do. And it was really cool because I got to just. And I'll tell you too, it was done in six prompts. That's it. To write this whole thing. And I'm talking not just the engine of the gameplay, but the presentation too. I mean, my first prompt was totally based on setting up the game board, and I basically went in there and just like, I'm talking to you, Glenn, here's I'm going to tell you exactly how I would design this in the beginning, this first step. And that's exactly what I typed in ChatGPT. Okay, Glenn, what I want to do in building this minesweeper game is I want to go out and check my workbook and see if there's two tabs named Minesweeper Game and Minesweeper Hidden. If they're not there, create two tabs that have those names from there.
[00:26:23] Guest: Tom Hinkle: Give me a ring. Give me a grid of cells in each of those sheets. Say it exactly like that in each of those sheets. I didn't go back and put the sheet names in. I just set up a conversationally in ChatGPT and I said, give me a 15 by 15 grid starting at cell D10, and I don't remember the exact reason for that. So cell d10 and what I want you to do is when you have that grid, make these range names. And I gave it a range name for each one of those. And then color each cell gray and make each cell look like a square. Because you know in Excel the cells are rectangular. And most people would probably go in there and say, well, I'll just resize them myself. No, I just told ChatGPT to make the cells look like squares, so it resizes the cells for me in those ranges. And that was my first prompt. And it built the structure for kind of build the wireframe, if you will, for the gameplay. And in the next one was because the way I designed it is, you have this gameplay tab and you have this hidden one that has all the bombs are. So the next prompt was simply place 25 bombs within the range on minesweeper hidden. Make sure they don't overlap each other so it randomly places bombs out there.
[00:27:40] Guest: Tom Hinkle: That's prompt number two. So that's how it came about. And I had to go in and tweak some things. The very first thing I did wrong in the first prompt, when it came back, it came back in Python code. So I had to go back and say, hey, write this in VBA. But that's the cool part. You could have said, write it in R, write it in C, write it in Java. Like you can just pick the language. And go write it. So I'll pause for a second. If you have any questions on where I should go for.
[00:28:08] Host: Glenn Hopper: Yeah. So I mean that use case is so good. And this speaks to why that domain expertise is important because no one would ever call me a good coder. I was a half assed coder because the only kind of coding I ever did was citizen development. And the reason I, the only reason I ever had to code is I would ask it for something and they would be like, yeah, you're back of the house, you can get the back of the line. And I'd never get anything, so I'd end up doing stupid stuff on my own laptop with, you know, setting up my own tables and running SQL queries and or, you know, doing things with Python and, you know, to the point where, like, back in the day, the developers would they called it a Hooper loop.
[00:28:51] Host: Glenn Hopper: Even my name is Hopper, but Hooper loop rhyme. But because I was, I would never close my loops properly in Python. And so they would just, you know, the loops that would just run, run forever. And but when ChatGPT came along, my ability to code went way up because things that would take me forever, just because I'm not sitting in front of that blinking cursor. And if you. I don't think you could go and get it to code for you in a reliable way. And maybe this is getting better every day. So maybe by the end of 2025 this won't be the case. I don't think if you didn't, if you didn't know anything about coding, you couldn't ask it to go create a Python script for you and run it. I don't know, maybe you could, but you wouldn't even know what environment to do it in or need to know.
[00:29:34] Guest: Tom Hinkle: Like the basics of what a loop is or what you know. Yeah, you know what an assignment is or something like that. You need to know some of the basics because it's again, it's looking at all this information. And in some cases it's coming up with a new idea based on these five things that seem similar. So you're going to have to tweak it a little bit.
[00:29:51] Host: Glenn Hopper: But the amazing thing, again, it goes back to that exoskeleton. You can then because you've got that skill set, you just become that much faster at coding. And you can you're basically move more to a QC role and like making sure that, you know, you've got it architected the way, the way you want to, but the actual writing of the code and I don't know, do you use like a GitHub copilot or any of any of the coding assistant tools these days?
[00:30:16] Guest: Tom Hinkle: I have like I said, my my job right now, I don't have to do a ton of coding unless it's like SQL where it's, you know, that's more joins and, you know, creating data frames and data sets. But I use those a lot now. I don't use those yet.
[00:30:29] Host: Glenn Hopper: What about Copilot in Microsoft? Have you seen a lot of use cases for that or are you using it today or is it.
[00:30:36] Guest: Tom Hinkle: Let's talk about that in the MVP thing a little bit. So yeah I'm an MVP. Woohoo. And that means it's just nice that Microsoft saw a couple posts I made and thought that I'm a good ambassador for their tool set, and I'm always championing Excel and the Excel versus Python battles and things like that. But we get to see things early. Copilot has had a rough start, but I think Microsoft is like in the last week, even has done some things to make it better. And I want to communicate this.
[00:31:06] Guest: Tom Hinkle: Copilot is now included in a personal office 365 license. So before it was going to be an extra charge, like $30. And I actually didn't have access to it because we're supposed to get it for free as MVPs, but their mechanism for getting it to us, it just wasn't working. The tenants weren't working out right. And because because the models, they're supposed to charge for it. Anyway, long story short, it's now 365. It came right up. And now part of it's integrated. So I'm just starting to get my feet wet working with copilot. And I'm going to do the same sort of things I did in ChatGPT. Just before I write that formula, I'm going to go ask copilot what it is. And that's how you get better just doing the reps. But I think it was a very smart move with them. I think they are increasing the O365 bundle by like $7 a year. But to me, I'm much more. I'm like, look, charge me whatever you want to charge me and put all the goodies in the bag. Like, I don't want to pay $100 for office 365, an extra 25 for copilot, an extra 15 for Python, you know, an extra ten for PowerPivot. Like, just put it all together in a bundle, pay it once a year and it's good. So for those of you with office 365, you got it now.
[00:32:22] Host: Glenn Hopper: It's good. Yeah. And I think it's going to, I mean it is only going to get better. And there's I do think Microsoft has a very smart strategy around Copilot and it's the writing's on the wall. You can see where it's going. It's just how quickly they can get there.
[00:32:37] Guest: Tom Hinkle: And people have to get really comfortable with this because like we talked earlier, these are skills you built up over your life that you're proud of and that you're good at. And that's true. But unfortunately, you know, especially in the space we're in with technology, progress just keeps going. You got to keep learning. You know, you have to keep up with things and move to the next thing and be able to let some stuff go. At times it's hard to do.
[00:33:00] Host: Glenn Hopper: And maybe one way to kind of roll it all together would be in one of the prior conversations we had. You've talked about or you've talked about this a lot, the importance of communicating complex technical concepts to non-technical audiences and that, you know, it's interesting. You know, you were talking about aggregating your thoughts into an email, synthesizing them, going out using chat, using the technology for that. But I think as we get further removed from the grunt work of, you know, of building these reports and doing all that, that communication becomes even more important that you know that the real human thing of being able to genuinely communicate a concept. So what do you think about that ability? How critical is that skill in the finance team of the future, and what advice do you have for finance professionals now who are kind of want to improve their storytelling through data? And that's, you know, either with AI or without, but just in general, that's going to be more important.
[00:33:55] Guest: Tom Hinkle: Well, that is a tough skill and it takes a long time to learn. I think I'm decent at it, but there's people that go to school for communications. They're really good at it. But this is another area of AI that I'm really going to start looking into. I just recently upgraded to the paid version of ChatGPT and you know, a lot of things that help with these rollouts and things is making short videos on things to explain to people. Well, now, some of these AI are these agents. They actually have video creation agents or ChatGPT or GPT tools to do that. So I'm going to start playing with those and seeing how those work again, still writing the concepts. Hey, I need to explain, I need to well, let's just take what we've been talking about. I need to explain to somebody who's a non-technical person what an AI agent is. Can you help me write a video to do this? Now, I could do that.
[00:34:48] Guest: Tom Hinkle: That'd be exciting. I could spend all afternoon. I'll put my lights up. I'll get my lighting right. I'll record it. My dog will start barking. The kids will come home from school. But I'm going to see what the videos do. I mean, we're still going to see that video presence look pretty much like Max headroom from, you know, 30 years ago. But if it can communicate eloquently and give the message that's something I think can help too. I hope I answered your question there.
[00:35:14] Host: Glenn Hopper: Yeah. No, absolutely. And it's funny when you say, you know, looking like Max headroom. So I don't know if you've seen some of the early video creation stuff. There's that Will Smith eating spaghetti. Have you seen this video? There's a whole meme around it, but how creepy. And like Nightmare fuel those first videos were to how good the videos are now and that, you know, just a little over a year. And now imagine another year. Imagine five years. I mean, it's going to be deepfakes. And, you know, maybe you and I will just phone in future podcasts where we have our, you know, our digital avatars interviewing each other.
[00:35:47] Guest: Tom Hinkle: Ask some questions about AI. And then everybody, everybody out there in Lincoln will just have bots that will watch the video for them and summarize the main points.
[00:35:57] Host: Glenn Hopper: Yeah, it's a wild world right now, but it's well.
[00:36:01] Guest: Tom Hinkle: I think we're going to getting ready to wrap up. But along those lines, I think one of the big things is it's so easy to access so much information and extra and so much extra information. I think it's going to be up to us as AI users to be able to cull that down, because, yeah, you can give somebody every definition they want in the world, but most people don't want to hear all that. They want to hear relevant things like, hey, tell me the three most important things out of ten, and then I'm good. If I want more, I'll come back for the rest. But that might be something that we have to be cognizant of. Not just giving everybody everything because then it's overwhelming.
[00:36:36] Host: Glenn Hopper: Yeah, well, that goes back to the storytelling bit. It is finding the signal in the noise. When all this information is out there, that you have to be able to tell that story and to be able to grab what the meaningful data is in this, you know, fire hose of data that's coming at us every day. Well, we are getting towards the end of the show. And so normally at the end of the show, I think we've gotten better at creating these questions. We have a grab bag of random personal questions. And normally when Paul is here, he has an approach that he uses. And then I have an approach that I use. So what we've started doing initially we were just asked and we would try a different models from Claude to Gemini to ChatGPT to Llama, and we would say, come up with 25 fun personal questions that we can ask our guests. But what we started doing is I built a GPT and Paul, I don't know if you built a GPT or if he's just doing it, but we will actually take like your LinkedIn profile and whatever.
[00:37:37] Host: Glenn Hopper: If you've got a website, we'll take that and we'll dump that into the GPT and say, come up with a list of 25 questions. And then the delivery would normally. Paul, he gives you two options. He can either do a random number generator or you pick a number and we ask that question. And then for me, because I've already done it in the GPT, I'll just say give me a random question and do that. But since Paul's not here, I think I'm going to give you his approach. You can either pick a number one through 25 and I'll, I'll read you the question from that or we'll just let ChatGPT pick for us.
[00:38:11] Guest: Tom Hinkle: I would love to, I'm going to go do this later. Give me a choice. 1 to 25. I want to say that seven is probably the most commonly picked number when you ask people that. So I'm going to pick 18 okay. Let's see if that one.
[00:38:25] Host: Glenn Hopper: Okay. This is pretty good I think if LinkedIn had a feature where you could list personal quirks, what would yours say?
[00:38:36] Guest: Tom Hinkle: Mine would say that I talk to strangers all the time and that got worse with Covid. I'm more of an extrovert, but going into work really helped that. And I say that because, oh my God, my family hates it. When I do that, I'll talk to anybody walking along the street or strike up a conversation. And I think they hate it because a lot of times when you start conversations with people, everybody wants to talk.
[00:39:00] Guest: Tom Hinkle: But I'm not coming with something 100% original every time. Like, if I see somebody walking to the grocery store, it's like, hey, you know, did you leave your cart out there? And then the next time I go to another grocery store tomorrow, so same thing. And my kids hear the same conversation starters over and over again. So they really get annoyed with that. But it's just a way to make connection with people, I think.
[00:39:22] Host: Glenn Hopper: And I think it's funny, as you say that, and as a father and thinking about all the times that I annoy or embarrass my kids. It's one of those things that like, you start doing it and then you realize, wow, I've become that dad or whatever. And then, yeah, so you crank it up when you're around the kids, maybe. Or at least I do. But it's, a I don't know. Maybe it comes with just where we are in life. But I hear you, and especially after Covid and being, you know, any time in lockdown where you weren't interacting with people, just getting that one on one contact is great.
[00:39:53] Guest: Tom Hinkle: So yeah. Well, let me let me add one little thing to that because this one's I think is funny again, to keep the day light and entertaining when we do, when we go put our name in somewhere for food or, you know, in a drive through or whatever, and they ask your name, I always say my name is Groot. And that way when my order's up and they say, hey, order for Groot, I could say.
[00:40:14] Guest: Tom Hinkle: I am Groot.
[00:40:18] Host: Glenn Hopper: I love it, love it. And your kids go running. They scramble whenever that happens, right?
[00:40:23] Guest: Tom Hinkle: Yeah.
[00:40:24] Host: Glenn Hopper: They're like, I'll be in the car, dad. So my wife, I need to tell my wife to do that because. So her name is Kerith k e r I t h. And no one, no barista has ever gotten it right. It's always Karen Kerri, who knows what they're going to come up with. So she just tells them a random name. And I think, how are you? How do you even remember what you told them. When you call it? But it is now. I need to tell her. She needs to just tell him her names Groot.
[00:40:50] Guest: Tom Hinkle: It's so awesome. It's really fun.
[00:40:53] Host: Glenn Hopper: Well, Tom, I really appreciate you coming on the show. I hate that Paul couldn't be here because, like I said, I know you guys have been trying to get this conversation going for a while, so I do appreciate your time. And thanks again for coming on.
[00:41:05] Guest: Tom Hinkle: This is great, great to talk to you today, Glenn. We'll talk soon.
[00:41:09] Host 1: Paul Barnhurst: Thanks for listening to the Future Finance Show. And thanks to our sponsor, qflow.ai. If you enjoyed this episode, please leave a rating and review on your podcast platform of choice and may your robot overlords be with you.