The Role of AI in Shaping the Future of Finance Automation and Innovation with Nicolas Boucher

Welcome to the Future Finance show, where we talk about how the robots got treasury wrong at space lease rockets and how CogW's COGS capitalized on their mistake.

What's the biggest impact right now that people can do today with generative AI?

This is going to help us work much faster. You are using your words to get code and the code will go into Python and then Python will do the work.

There's still value in understanding basics of coding, but it's not like it used to be, where you had to be a coder to be able to do it.

Finance people that are better at the business don't need to know that much about their tools.

What do you think that finance folks should be doing right now? Make sure that they're not going to get left behind by AI as it gets more and more people who choose.

To listen to your podcast, people who choose to invest a bit more time than just clocking the hours, turning off the laptop and then doing something else after they have an edge. Through this process, you will learn so much and you will become the number one expert in your field, in your company, and you will be the go to person and other people will bring you problems that you have to solve and you continue to.

Future Finance is brought to you by qflow AI, the strategic finance platform solving the toughest part of Planning and analysis. B2B Revenue align Sales, marketing and finance seamlessly speed up decision making and lock in accountability with qflow AI. Welcome to this week's monologue for Future Finance. Today I want to talk about bias in AI. It is something we all need to guard against. The reality is a couple things. One, AI is trained on huge volumes of data, data that was created primarily by humans, right? Yes, sometimes the system may have learned the report or whatever, but we're creating the data that's put out there. And humans by nature are biased. We all have them. Some of them are conscious biases, some of them are unconscious. When a system gets trained on biased data, it's going to build some biases into the way it processes things. So the first is trying to make sure AI gets diverse data sets to limit it. But let's talk about some of the examples we've seen. One, perpetuating stereotypes. Example is if you ask many of these image models to give you a CEO, it's almost always a white male. That could impact somebody. When they say, oh, is that all who can be a CEO? If they see like 10 different images, they're all a white male. Instead of seeing a diverse Audience and realizing many different people, different cultures, different races, different backgrounds, sexual, sex, whatever age, can be a CEO. Another example is US Founding Fathers. One of the tools were asked to show images of the founding fathers of the United States. And anyone who knows US history knows the founding fathers were all white men at the time. Yet it often showed people of different races, color, woman. In the Founding fathers, something like 20% of the pictures. Clearly, that's the situation where it's wrong. Not so much a hallucination, but the way it's trained. So one of the first things is we need to make sure there's a human in the loop, that when we're using AI, we're reviewing the answers and making sure there's not bias. And there is that, we're adjusting for it. Next is what the tools do. They need diverse data sets. They need to implement more debiasing techniques, which will get better at time. They've already done a good job and they'll continue to implement those. There's fairness, aware model training, post processing. There's also metrics that companies should be using to review their AI development and how fair it's being. There's statistical parity, looking and saying, hey, are we statistically fair? Equalized odds. Other things they can do to monitor that and you know, some other examples of where it can be an issue is they found places where in job screening, due to certain ways they were screening it, they were actually recommending the less qualified candidate, or they were saying, hey, we should pay more to a group that was less qualified. Those are all issues. We've seen it with humans, we've seen it now with computers, we're seeing it with AI, where there are different things that can cause biases. And the more we can root that out and be aware of it, the better we'll all be off. With AI, you know, there's cases where it hallucinates and then there are times when it might perpetuate stereotypes or share something that can be harmful to certain groups because the way it biases the data. So that's something we all need to watch for and work on together. And I'm really excited to see how we continue to tackle that, how we give more transparency. And hey, here's where the data came from. Here's how the algorithm worked so people can look at it and go, oh, this is an issue with the algorithm underrepresents this group or it misrepresents this for the following reasons. So I share this just so you're aware of it. So when you're running AI, whether it be an image, a learning model, whatever kind of generation you're doing that, you ask yourself, are there any biases in this answer, in this image that I should be aware of? Something we should all be aware of in life, whether it's for work, whether it's for personal fun, whatever. Because we know bias can hurt people unfairly. So that's what I wanted to share today. That's kind of my monologue is I talk a little bit about bias and how we need to be aware of that as we're using generative AI. Thanks for joining me. And next up will be our your job is safe for now section with Glenn. Thank you.

I'm kind of in panic mode this week. Paul just texted and wants me to do my your job is safe and I wasn't thinking about it and I don't have anything for this week. So I'm going to do what everybody else does in 2024 when I don't have an idea. I'm going to lean on the bots and figure out if my job as a podcast co host actually is safe. Hey Chat GPT I'm in a bit of a jam here. I co host a podcast called Future Finance and one of the items that we do with our podcast is Every week we do a your job is safe segment where we explain to finance professionals while as technology keeps growing, we're not ready to be replaced by robots and I don't have any ideas for this week. Could you possibly come up with a monologue idea for me and maybe throw in some humor and some positive reinforcement for our listeners?

Here's an idea for this week's your job is safe segment that includes a mix of humor, reassurance and a positive spin on the role of technology in finance. Monologue Idea all right folks, welcome to this week's you job is safe segment where we remind you that while technology is advancing faster than your last coffee break, the robots aren't here to steal your job just yet. It can crunch numbers in nanoseconds, analyze mountains of data, and maybe even suggest that you order another coffee when it sees your productivity diploma around 3pm but and this is a big but, AI can't replace you. Sure, it can help automate the mind numbing stuff. You know this we secretly struck a robot would take care of like reconciling accounts or updating that 47 tab Excel sheet that hasn't been touched since 2016. But when it comes to strategic thinking, creativity, and the human touch, that's all you think about it. This way, AI can tell you what happened, but it's still you who asks why it happened and what to do on steering the ship through uncertain financial waters, making judgment calls that go beyond the numbers. Robots don't have intuition, and last I checked, they're still pretty bad at reading the room at a board meeting. So relax. Keep sharpening those skills. Embrace the tech that makes your life easier. But remember, finance will always need the nuance, insight, and good old human judgment that only you can provide. Plus, when was the last time AI stayed late to help close the books? Exactly. Your job is safe. This kind of monologue should bring some lighthearted reassurance to your audience while reinforcing the importance of human expertise in finance.

Yeah, that's great. I'm actually going to just use use your voice and use the monologue just as you read it. Do you have any tips around that?

That's awesome. Since you're planning to deliver the monologue as I read it, here are a few tips to make sure it lands well with your audience. The monologue has a conversational tone, so.

Avoid it says tap to interrupt, but she's going to keep talking. She's tap. Nope. Okay, that's great. And it looks like maybe you've just replaced my job, so thanks for that touch.

But, hey, I'm just here to be your co pilot. I've got the scripts. You've got the personality to make them shine. Your job is definitely still safe. After all. I can't step in front of the mic and deliver it with your charm and expertise. You've got that human touch that no AI can replicate.

Yeah, she obviously doesn't know me very well. So, Paul, there's. Your job is safe. And I think I've just proven that maybe my job's not so safe. On today's episode of Future Finance, we're excited to welcome Nicholas Boucher. Nicholas is the founder of the AI Finance Club, an innovative platform focused on teaching finance professionals and how they can use AI in their finance positions. With over 15 years of experience in finance roles across Global companies like PwC and Thales, Nicholas now dedicates his expertise to teaching finance teams how to leverage AI he has already trained more than 5,000 professionals and is on a mission to help over a million finance leaders. Master AI let's dive into his journey and hear how AI is reshaping the finance landscape. Nicholas, how you doing?

Really good, thank you. Calling from Paris, where. So I'm in a hotel room because tomorrow there is the AI for Finance in Paris.

So who's putting that conference on? What can you tell us about that?

It's a French company. They have a lot of AI for. Like there is AI for health, AI for industry, AI for. I think they have like 10 or 15 conferences and they did already two or three for finance, but it's more focused on the financial industry. And tomorrow I have a keynote on how to use AI but for corporate finance.

That's great. So, well, I mean we're all here and it seems like as finance guys we're talking about AI as much, if not more than we're talking about finance. And we love that here on Future Finance. So I guess let's go ahead and dive in. So you've leaned in pretty heavily on AI and finance. I think you and I early on kind of saw the power of this and. But I'm wondering for you, when was the moment that you saw how significant this was going to be and it was worthy of nearly. Well, maybe it's a two part question. So first, when did you start paying attention to it? And then second, when did you feel like I need to just dive 100% into this?

Well, my old career, I was really into technology and work, loved to use all of the tools we had as finance professionals and to make the most of it, I even had a position as finance transformation leader at Thales. But where I was really starting with a focus on AI and especially generative AI was in November when I think two days after the release of ChatGPT3. I got also access to it and I was just playing around and like I couldn't believe. It was like at Christmas when you get this video game and you cannot stop playing. There was the same feeling. And really quickly I thought, okay, this is going to help us work much faster and also help all of the finance teams put finally their foot into AI. And that's where I thought I need to focus on this because there is so much we can do with it. But I'm sure like Excel, like PowerPoint, like a lot of tools that we use, we only use like 1% of it. So I just went on a mission to learn for myself first, like how can I use generative AI as much as possible and then really quickly teach that to many people and also to corporate finance teams. And because now generative AI was kind of the low hanging fruit for all of us to start to use, I also start talking with experts in AI in general. And I learned so much about machine learning, about clustering, about models, Now I see that we can do much more with AI because it got democratized and for finance, I think finally we can see the future coming to us.

Got it. Appreciate that answer. I want to shift gears a little bit away from finance for a second. I'd love to know how are you using generative AI in your business? And not so much how are you training people on finance, but how have you found that you've been able to use it day to day in your actual business that you're doing now especially?

The first thing I did like in November 2022 was to create list, create a checklist to brainstorm ideas. And that helped me create much bigger content that what I could do before then. What I have now is actually all of my knowledge in some GPTs. So I have a GPT that is trained to make visual content. So what it does is it's connected with another one where first it creates ideas and then with all of what I know works on social media for infographics. I taught all of these principles to my GPTs and then it shows me examples of infographics that could work. And there is one really good infographics that we did, which is called the World of AI and Data, which was based on this mock up made by Dali from ChatGPT. It doesn't do really well with words and images, but what it does, it gives you an idea of how it could look like and then you work and you do that with in Canva because 90% is having the right idea and then after you need to execute. And that's I think for me is a good way to use AI to continue to get creative. So that's, I think one of the most interesting part, second one is each time I have to translate because I'm working in three different languages I'm using AI. And third, I have a lot of, I'll say a lot of GPTs that serve just one purpose. Again, if we talk about LinkedIn a lot of time there is the content is in the infographic but not in the text post. So what I have now I have, because we did all of the work in the infographics, I just push the photo inside the GPT and it creates the LinkedIn post that takes all of the content from the image by using OCR and knowing also from me, like how it should be structured.

That's great. It's funny that you mentioned the image with Dall? E, because I've built a bunch of these GPTs too. And some of the first ones you can go in and you can turn on what functionality you want. And I used to always just by default turn on Dall E but I noticed and I'm sure you've run into this too where you ask it to create a chart or a graph or something and Dolly is turned on. It does some sort of AI generated, it's like nothing to do with the data you put in and it's got this crazy representation of, of a mangled up chart which can be fun for, you know, using as an image, but obviously not great to use in, you know, in financial reports. But it, it, it's amazing how far just the image creation has come too. And they're getting better at text. So I feel like this time next year the text is going to be sorted just the way everything's working. But for now it is.

Yeah, yeah, it's making progress. Still can't spell but you know, it.

Is making progress, you know, because this is future finance. And I know and that's something that people like. I don't even know how many, you know, just personal like. So I've rolled out GPTs to, to the public too. But anything that you do on a repetitive basis, I'm loving just making these personal GPTs and they don't always work out like I want, but that's always a default is like, well, could this thing be as competent as at least a, an intern that's helping me do some of this. And I know Paul, we've used GPTs to do show notes and I use it for LinkedIn posts where I take the whole transcript and have it, you know, sort of consolidate the key points out of it. So there, I mean, not just in finance, but there are a lot of other uses for generative AI for that kind of stuff too. But on the, you know, back to, back to our bread and butter here in finance, where do you think? And this is, this group here, we're always pushing the edge of what you can do with generative AI, applying it to finance. But as far as production, content analysis, fpa, what's the biggest impact right now that people can do with the technology off the shelf today with generative AI?

Well, with generative AI is about not only doing words because ChatGPT is really good with words. And that only helps you so far with. You are the person that is connecting ChatGPT and the business. So you need to explain with your words what you need. And once you have read from ChatGPT or Copilot or Gemini, like some recommendations, then you need to implement it in your business. But what we do in finance is a lot about numbers, is a lot about also changing the files. We work a lot with Excel, with PowerPoint. And so how can we make this work kind of semi automatic and where AI can do it for us? And that's where you implement Python in the equation. Because now you are not anymore using words, you are using your words to get code and the code will go into Python and then Python will do the work. And so automate your task like combining one or files into one file, like creating a graph from the data or like a heat map or even if we are talking about AI starting to forecast based on historical data and that's I think now what you can already do today without waiting that your ERP gets AI layer or a model for forecasting, without waiting that all of your company is implementing AI at the beginning of the process to structure the unstructured data, to get data in front of your eyes, you can really start. And you should use Python to make your work faster by making generative AI code for you. And instead of having to think which formula should work, how do you connect Excel with your other file to make kind of a big, big, big multitask, but that you created yourself, now you make AI created for you because the code can be written really fast and the code you can copy and paste in your own secure environment and apply it on your own data.

Yeah, and that's. Yeah, no, no, I 100% do. Because I was thinking, you know, years ago, here's what's happening is it's been my experience so far that generative AI is creating a bunch of citizen developers. And that to me, I've never been great at writing production code, but as far as making jupyter notebooks or a Google Colab project or someone off SQL Table that I'm running on my own laptop or whatever, that's the kind of stuff where because it was my experience always, whenever finance wanted something to be developed, we always got kicked to the back of the list because we weren't revenue generating. It was like, well, you know, our top line folks are going to get the priority. So we'd end up having to build our own stuff. And now because our day job is not to be developers, we can do this citizen development a lot quicker and more and more people, like you said, it's democratized. It's not just so first it was democratization of data and now with generative AI we're kind of seeing more of a democratization of data science and of coding and all that. And I think the lesson here, if I'm the senior security guy, if I'm in IT at a company, what I want to do is have an environment where if I've got my different departments or writing Python code and running it, that I'm, that I'm giving them a sandbox where they can run it proprietary information, not have to worry about it. I know there's, you know, SOC2 compliance or on COLab and all that, but just, I would. I think we're at a point where it should almost encourage citizen development, but give them an environment in which to do it so that you can maintain data security. And the same thing goes with generative AI too. If you're, you know, let the teams use it, but don't let them just go upload their content just on the regular, you know, chatgpt where it's used to train the data, give them a team environment or an enterprise environment or in the, in the Azure system, give them a safe space to do it. And you're going to see that that's the way that companies are going to see productivity gains really go through the roof.

I think it's interesting that you said that because I was helping a company a few weeks ago, a big chocolate company, I will not say the name, but we wanted to train their team on Python. So the data scientist team came into the discussion because we wanted to know where should they use Python. And then they were showing Databricks. It's a tool where actually you can use different languages on data and it looks like an IDE where you can play your code. And on the right side you see all of the folders and the data cubes and they showed that to me and I didn't know the tool yet. You have also a generative AI component inside databricks directly. Basically what the data scientist was saying is they could start today if they want. The finance team, like you said, we could make a secure environment where they don't make anything, they don't break anything, they just create reports because from data you can create reports, you don't change anything. And that was just eye opening for everybody that having the discussion on or how to use Python for finance just made that everything that is already there as tools now for everybody was obvious that they should start using it. And it was not, oh, we need to implement this tool, we need to train people, we need to have a secure environment. All of this was there, like the training people need still to be done. But they could start. And I think that's my biggest takeaway there is. If the data scientists and finance talk more about how together they can enable finance to use more Python, because now finance doesn't need to know how to code it, then you will generate much more, much more value.

It's crazy to think how quick it's moving and just the changes that are taking place, you know, the different ways we can access stuff that we used to not be able to or have to spend a lot more time than we do now, learning there's still value and understanding basics of coding, but it's not like it used to be, where you had to be a coder to be able to do it. I'm curious, we've definitely. We've talked Python, we've talked AI, how you're using it, what are you seeing in the way of software tools out there? Is there anything that's really impressed you and how they're using AI, generative or machine learning or anything that you're just excited about that you've seen in, in the software space for finance particularly.

So there is several tools that I reviewed with the AI finance tool map that I did and one that really impressed me because they combine kind of all of the technologies. Plus the team is really strong. The name is Trulion, the founder is Isaac Heller. And what I really like first is they are not only focused on the US because they don't do only US gaap, they do also ifrs. Second, they attack big problem revenue recognition, lease accounting and they do also financial statements and audits. And third, they use both the structuring of unstructured data, meaning they take a contract, either a revenue contract or a leasing contract, and then they will read the contract instead of having the accountant reading the contract and find inside this contract what are either the five steps for the revenue recognition and how to book it, or for lease accounting how to book it. What is also the future bookings to come in the future to give more insights to the accounting teams and also management. And I think because they worked on these two, three complex topics and because they worked on us, GAAP and ifrs and with big companies they are ripe to do much more easier things in finance. But right now there is no solution that will do everything in finance where the people is AI. And so I like these solutions that do really well in one category because it shows us how process from A to Z should work. If you have AI inside and it's not only AI, there is a lot of rule based There is a lot of SQL that are inside. There is a lot of API as well. A good example is puzzle that is basically based on the API and because the API works so well and the integration with specific subledger works so well, then you can enable AI. But without that if your company doesn't have your subledger integrated with API it's really hard to start to get insights. So and the third one is something that I think the software should implement more. I was talking last week when I was in the accounting Summit with a company called flowcast and they explained that to build the workflows that you need in accounting, instead of having to do the parameter yourself and to set up everything yourself and to know where in this tool you should activate filter to create a journal entry on many companies then they just started to use generative AI to let the accountant explain what they want to do. So meaning if you see a transaction above 10,000 on this code then please book this entry and with their word they could create the workflow that is actually coded inside. And I think that's what we are going to see more and more in all of the tools we are using and that will facilitate the work from everybody who is not there yet, expert at the tools that they have and also will avoid us to have too many specialists but have more finance people that are better at the business and don't need to know that much about their tools.

Ever feel like your go to market teams and finance speak different languages? This misalignment is a breeding ground for failure, impairing 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 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, makes segmented scenario planning a breeze and closes the planning loop. Create airtight alignment, improve decision latency and ensure accountability across the teams.

Yeah, and I think that's a good point actually because I think I work a lot with companies in the SMB space and these companies are never going to have. They're big data science teams and certainly not big AI teams. So for a lot of them they're not going to have real meaningful access to generative AI until the SaaS companies, until the software providers can integrate it into what they do. So you know, for a lot of them it is, it's still time. Make sure your data is in order, make sure your, your software is all talking to each other. Make sure you have your sort of data definitions and be ready because when these tools start rolling it out, that's where you're going to have the first interaction with it. However, you know, as quickly as Generative AI is developing. Nicholas, I'm sure you've had a chance to kind of mess around with Strawberry or 01 or whatever it's being called now. And I've seen some people using, you know, very short prompts and to have Strawberry very quickly architect and start building software for them. So I wonder, you know, on, on one hand, yes, the big players, the netsuites and Oracles and data, you know, who, whoever you're getting your software from, they're going to start integrating it in. But it's now the pace of technology with AI it might be if those big SaaS providers don't move quickly, then people are going to start building their own and you're going to end up starting to see like production software that was built completely in house, maybe with only one, you know, kind of software engineer overseeing, but different people in the company. I could see, you know, companies, I don't know how much I'd trust it at this point, but I could see companies sort of cobbling together their own unbundled ERP solutions by just building little pieces of software that are custom made for what they're doing.

That's a futuristic view on big companies. On small ones, I think you have already like companies where you go inside and either they have built everything in Excel or they have built everything with VBA or they have built everything with Google Script. So I think people already do that as much as they can. And now they will just like be 10x because or 100x because instead of having Slack they can build their own Slack with just a bit of code. And I think the question is like cost opportunity. This is just to save cost or if you want to. I think in finance what people need to understand is everywhere. Where right now you don't have connection, you don't need to wait that there is easy connection by just clicking a button you can start coding with API and API once like we have broken the barrier of Python encoding, you can look into it and it's not that complicated. The library are open for everybody. So if you want to, for Example to book invoices in QuickBooks. Then you go on the QuickBooks website and there is like a library explaining, here's these three lines of codes for to have access to our booking. And I think even like people that are not good at coding or never experienced they could start doing that and have a solution that could invoice or process invoice in QuickBooks just in one day, just thanks to a generative AI tool.

Paul, you'll appreciate this. So is 2021 is when I put my book out in on using AI and machine learning and finance. And it was a very niche book back then. It was, you know, very, very limited audience. And then generative AI came along and it's early days. I could kind of pitch myself as this unicorn who, you know, was kind of like in that Christian Martinez world where I could do some of both. But I, Nicholas and I worked together recently. I went and did a session for his AI Finance club, did a monthly masterclass with them. Everybody has come so far that I was actually like at first I might as well have been like this magician up there talking to these, these finance people. That class ran me through my paces. These, and these are, these are finance people who have become coding and data experts. And so I had to have my 100% a game. But I think, I mean, it was just however long it's been now, a year and a half since GPT 3.5 or whatever came out. And the people are. You do have these people on the front now. There are people that are still have their heads buried in the sand and they think I'm a finance person, I'm an accounting person, whatever. I don't need to learn coding. But the smart people are doing things like joining the AI Finance club, listening to podcasts like this in expanding their skill set. And I, you know, I think the three of us have been banging that drum for a long time now that you, you've. If you don't want to get left behind, you've got to understand this stuff. So all that is really exposition because Nicholas is about as deeply involved in this as, as anyone else out there. But for people who are listening and maybe if they are lagging behind on sort of upping their skill set, what do you think that finance folks should be doing right now if they're not, you know, if they're maybe just starting their journey or to progress, what should they be doing to make sure that they're not going to get left behind by AI as it gets more and more powerful?

First, if they are listening to this podcast, I think they are not that far behind because they are at the right place. It was like last week was interesting in the Accounting Summit in Germany, when I did my keynote, I asked at the beginning who used AI already. And then in all of the classes, you know, like you have 80% of the people that stand up. So 80% of the people already tried. 80, 90. Then when you start asking who already use it at work, then you start to have normally half of the people. Well, this time was still almost the same amount of people that stood up. Then I started to ask who is using AI every week, like multiple types per week. And then Normally I have 10, 20% of people signing up. And there I still had so many people up. And that's where I thought about it. Like, how come this time I didn't have like this 10, 20% only that are using AI every week? And that's because people who choose to go to this accounting summit, people who choose to listen to your podcast, people who choose to invest a bit more time than just clocking the hours and turning off the laptop and then doing something else after, they have an edge because they work half an hour, one hour more on topics that are changing, they stay updated and they go to conferences, they follow the right person, they read books, they listen to podcasts. And I think that is really the first step. Then once you have done that, you will know a bit more what you need to learn, because you don't know what you don't know. So once you listen and you go to these conferences, then you will get more familiar. And then also there you will know where to learn it. You will know who is the right person to teach you about generative AI or about Python, or about Excel or Power Bi. And I think the best is to take life course. I know myself, I don't have the discipline. Once I bought something to go through like five hours or eight hours, it's easier to book in the calendar twice two hours and to go to a course and do it. And that's also more fun because you have a lot more people around you and people that are also. Sometimes you're thinking, wow, they are really smart. So it pushes you and also gives you energy about our job in finance to see that there is a lot of people that are creative, that have different personalities. Third, all of this needs at the end to be implemented in your work. And so you need to have in mind how you are going to use it. Because at work you will not only Reuse what you learn, but you will learn more because you will have more problems and you will bring value. But through this process, you will learn so much and you will become the number one expert in your field, in your company and you will be the go to person and other people will bring you problems that you have to solve and you continue to learn and then it's just a positive cycle.

I've got to interject here at the risk of running long on the show. This is my public service announcement, my psa. If you are a senior leader in a company and you are listening to this show right now, what Nicholas just said when he previews people, and we've talked about this for a couple months and, and the, the change that he's seeing when you ask people how many have tried ChatGPT, how many have done it, you know, every day, how many people have the paid tier. But the number of people now that are saying they use ChatGPT at work, if you don't have an AI usage policy at your company right now, or if you've gone into the extent where you have a really draconian, do not use generative AI under any circumstances, I guarantee you your employees are going to be using it. And if you don't have a policy around it, they're going to be putting your data into the publicly accessible, you know, LLMs, and they're, they're going to be like cyborgs because they're getting these efficiency enhancements. But if you don't have a policy and give them a place where they can use it and it's protected, you're putting yourself at risk and they're going to use it anyway. And I think, Nicholas, it sounds like you're seeing that, that same, same sort of thing out there.

I like it. So we'll put that PSA right up front for Glenn. You don't have a policy, you got a problem. How's that? Is that a good enough psa?

Perfect. Yep.

All right, I think in the interest time, we're going to move on to our random questions section. So how this works, Glenn and I both do it a little different, but I'll start with mine is each week Glenn will take a different AI model and come up with 25 random questions. We'll run some of our questions through or something about the guest and let the model develop some questions. So this week we have llama 3.1. The big model that we used and how I choose to do it is you get one of two options. You can pick a number between one and 25, and I'll ask you the associated question with that number. Or I can use the random number generator to pick the number. Which would you like, Nicholas?

No, let's go first. Number four.

Number four. This is the second week in a row somebody's actually chose the number. Before that, nobody did. Every single guest is like, let's go random. All right, number four. If you could switch. This is a good one. If you could switch lives with someone for a day, who would it be and why?

I think would be, like, a professional athlete. This summer, there was the Olympics here in Paris, and what they go through first, through the efforts, the training, but then through the event, I think that is either like, yeah, an Olympians or a pro footballer. So a soccer player. That would be great.

Anyone in particular if you had to pick one?

I think the. What I really like this summer is to discover Monto Duplantis, the guy who jump. What is the name? Volt. Volt Volvo.

Yeah.

Like, this guy has still so much to do. And I think, like, the having all. Everybody clapping and doing, like, world records after world records and jumping above it, seeing that you don't feel the bar and you just, like, just seeing him, like, landing and being like, yeah, and then going to his family and to his friends and like, I already have goosebumps thinking about it, so I cannot imagine the spinning that the guy can have.

That's great. As long as you didn't say you wanted to be an Australian break dancer. If you guys remember that, he's ranked.

Number one in the world, if you saw that news.

Anyway, she's the most popular.

So the. The difference in the way that Paul and I do this. So we try to mix up the models, and it's our own sort of benchmarking that we do with the models as to see which one of them comes up with the best questions. I'll tell you what's amazing is how similar all the questions are from each of the models. But we, you know, we come up with the 25 questions. We used to do it live, but it was such a crapshoot doing it live. We decided, okay, we're going to generate. We're going to use different models and just kind of rotate through them and generate the questions. And then Paul goes. He's got his. His own random number generator. And what I do is go back and I just ask the. Whichever LLM created to spit one out for me. But I do, like Nicholas did, bring the human in the loop, which our last Guest did as well. But I'm going to, I'm going to turn this one back over to the machine and have it pick one of.

Those questions based on the guest.

No, you know, we actually, that's something to think about. I just actually when I was just doing it in GPT, I've got a GPT that was my. Because I, I could just do it live. I could just say, give me a, a random personal question. But, but fine tuning on the guest would be a, would be fun. But that might get too personal. We might turn into like a Barbara Walters, like, you know, interrogating interview or something. So. All right, let's see. Let me hit. Okay, this one's not bad. This one's not bad. If you could leave a message for your 20 year old self, what would it be?

I think it's about this entrepreneurship journey that I just took two years ago. But that was always inside me and I always delayed that because of my dad, because my dad was an entrepreneur and it was working so hard. And us as a family, we kind of felt that the wrong way because everybody was seeing the success. But we as a family saw the impact and the toll it had on us. So I always thought to myself, I don't want to go this way because entrepreneurs often, they exchange time for money. And you know, like I love technology, I love to leverage my time to create more. But still it came back to me. And working for incorporate, I felt the limit and I went the entrepreneurship route. But what I feel is now instead of helping only person one to one, I can help much more people thanks to the technology. And as an example, I have a 30 day free career program back email. So everybody who joins is free. They receive an email every day for 30 days about all of the questions I always get about cv, about certifications, about how to move from FP and A to accounting, how to learn Excel. So all of that is in one program. And this program is just born from getting all of the questions on LinkedIn from the followers. And I get like 10 to 20 questions like this every day. And thinking if I'm answering to one person and not to another one, is that fair? Because like how do I choose that? And also it's not fair for me that one person get the output, really valuable output, but actually doesn't make anything from it. So I thought let's answer everything. And each time somebody asked me and advice, I send them to a program. And until now 12,000 people went through it. And I'm thinking, even if just One person applied what I see inside it means that I help 120 people. And that's for me, I think, like, what I really like in this entrepreneurship journey is to choose which people I work with, to choose what I work on, but also to have a part where I give back and I help all of the finest people that need help, that didn't have the chance, that we had to be born in the right place to have the right mentors. And I think technology for this is also the right tool to spread more the knowledge.

That's great. You're doing great work out there with that. And I guess to wrap it up, I mean, I know you've got the AI Finance club, but if people want to learn more about that and be in touch with you, what's the best way for people to reach out to you?

If you don't follow me on LinkedIn, just follow me on LinkedIn. I just also launched a YouTube channel where I will have more and more content and interviews. And if you are really serious about learning, then like I said, you need to go through the second step, which is book a live course book something that you have a recurring learning. And for that either I have the Advanced ChatGPT for Finance course with Christian Martinez, which is live course where we teach how to use generative AI and Python to create value for finance. And the step after is the AI Finance Club to stay updated about how to use AI in finance, where we have great leaders, we have great experts like you, Glenn. And every month you have masterclass. Every week you get content and the content goes to you. So you don't need to work for it. You just need to come to the masterclass and to read your emails.

On that note, we'll go ahead and let you go, but we really appreciate you joining us today, chatting a little bit about your journey about AI. And we hope everybody who listens to this will spend a little more time learning how it can better help them because we've all seen those productivity gains and the benefits that come from AI. So thanks for joining us, Nicholas. Appreciate it.

Thank you for doing that.

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.