Artificial Intelligence Podcast: ChatGPT, Claude, Midjourney and all other AI Tools

Is Artificial Intelligence Going to Make Excel Obsolete with Christian Torres

Jonathan Green : Artificial Intelligence Expert and Author of ChatGPT Profits Episode 362

Welcome to the Artificial Intelligence Podcast with Jonathan Green! In this episode, we delve into the evolving role of AI in data management with our expert guest, Christian Torres. Christian is a talent with a unique flair for using AI to streamline processes within Excel, leaving behind the need for cumbersome spreadsheet expertise.

Christian shares his insights on how AI is transforming the landscape of data management by enforcing better data hygiene. He underscores the importance of having well-structured data before leveraging AI tools for automation and efficiency. With AI's assistance, the focus shifts from merely using Excel to optimizing it for more powerful capabilities without an extensive background in formulas.

Notable Quotes:

  • "You can't manage what you don't measure. And that's very true. You're not gonna understand your processes unless you actually collect the data." - [Christian Torres]
  • "The hype is that AI can do everything; in reality, it's about organizing your data before jumping to analysis." - [Christian Torres]
  • "You might have a beautiful dashboard, but if you don't understand the numbers, you've missed the point." - [Christian Torres]
  • "We're in a digital hoarding era, storing masses of data, often without understanding its value." - [Jonathan Green]

Christian emphasizes the necessity of defining the problem and understanding the potential of AI to provide solutions that enhance existing processes. He highlights how AI is not about replacing systems but about making them more efficient and effective.

Connect with Christian Torres:

Christian shares exciting developments in how AI can be integrated into everyday processes, particularly in Excel, to empower users with new efficiencies and capabilities. If you're interested in how AI can enhance data management and business processes, this episode is a must-listen!

For more insights and to connect with Christian, visit his platforms and explore the potential of AI in revolutionizing your data operations.

Connect with Jonathan Green

Is artificial intelligence gonna make Excel obsolete? Let's find out today's special guest. Christian Torres. Welcome to the Artificial Intelligence Podcast, where we make AI simple, practical, and accessible for small business owners and leaders. Forget the complicated T talk or expensive consultants. This is where you'll learn how to implement AI strategies that are easy to understand and can make a big impact for your business. The Artificial Intelligence Podcast is brought to you by fraction, a IO, the trusted partner for AI Digital transformation. At fraction a IO, we help small and medium sized businesses boost revenue by eliminating time wasting non-revenue generating tasks that frustrate your team. With our custom AI bots, tools and automations, we make it easy to shift your team's focus to the task. That matter most. Driving growth and results, we guide you through a smooth, seamless transition to ai, ensuring you avoid policy mistakes and invest in the tools that truly deliver value. Don't get left behind. Let fraction aio help you. Stay ahead in today's AI driven world. Learn more. Get started. Fraction aio.com. Now Christian is someone who is very bad at spreadsheets. My favorite thing about AI is that I no longer have to learn all those formulas. Like the main thing I'm always typing in is how do it's happened. Yesterday I was talking to chat, how do I move duplicate records in a Google sheet? How do I, if I have first name, last name in one column, how do I split those into two columns? And in the past you had to know so many co, like there was such an expertise required for Excel, and now I see it as you don't. You can use it without that expertise, but also the top more powerful things are possible. So I'd love to get your perspective as a pure expert. Yeah, absolutely. This is a thing I deal with every single day. I definitely use tools like chat, GPT and copilot to enhance the work that I do in Excel. But when I was. Basically given that question on whether or not Excel was going to be at risk because of the advance of events in ai, I really thought about it and I said, no, it's gonna actually force good data hygiene. What I mean by that is, yes. Chat, GPT and any other of the tools can help you perform function to figure out how to do things in Excel. But really what we're talking about is the concept of garbage in garbage out. If you want a tool like chat, GPT to help you build out a robust tool in Excel. You need to basically have your data properly structured. You need to get rid of all the errors and fix some of those things upfront before you even introduce that to ChatGPT. Because let's say for example, you have data all over the place. It's not in a structured table. They're across different sheets. You don't have things properly named, and then you just ask chat, GPT, please help. I need help with a formula. Okay. It might help you with that formula, but it's way better and easier, and you're gonna get so much more out of it if you actually take the time to figure out, oh. It can help me if I create a table, give it a name, and then refer to the actual column. So now you tell Chad, TPTI have an Excel workbook named workbook 1, 2 3. And on sheet one I have a table called products. And in that table I have a column called like description product id. I have, sales numbers, cost. I have all these very specific named elements in the sheet. And then you tell Chad GPT what you want it to do, and now it can actually just build the automation script. It can understand the context of what you're doing. So what I'm actually finding is that the use of AI is forcing people to be better at organizing their thoughts, organizing their data, and kind like pre-processing it before they jump to the advanced level of trying to have Chad CPT build like a custom formula. Yeah. That brings up one of the biggest issues, which is the over hype of ai. A lot of people see ai, like the enhanced button in movies, Bush one button, everything is enhanced. Yep. And we're now entering this era of digital data hoarding. We transcribe every meeting and we just store all on cloud drives. And I sometimes think about how much of the data and data centers is just trash, it's just hoarding. We store everything. And it's the same reason you see someone who like, oh I they, why do you need broken hangers? Just in case. What's the scenario right there where you need that? And you go what's the scenario where you'd need a transcript from a meeting where nothing happened 17 years ago? What if, and we're so prepared for what if, and it happens every once in a while. Every once in a while I'll have to go find an email from 10 years ago, but I've never had to find a transcript other than if it's a contract negotiation. Then of course there's a meaningful conversation, but how many of our conversations are that? So we. Store all of this data, and I think we have to get through this era where we think that just having data is meaningful. It's like people that buy libraries by the foot, like they just have tons of books behind them and it's just the spine in the book, or like just being near it makes me smarter. It's like osmosis doesn't work that way. That's like magic. So I do think you're bringing up something really important, which is that we have to start thinking about usefulness, which is, now you can use an AI to help you organize your data before you analyze your data. You can't go to that process, but you still have to answer the questions and say what the data is and how you wanna organize it. So even if you're not using like Excel language and you can use natural language with ChatGPT BT to create spreadsheets, you still have to know what the data means and what you want from it. And I think this is one of the biggest issues we have right now. When you have too much data, it's hard to search. So you have a message library. You have to go which floor of the library do I start on? It gets really complicated. I find that this is the thing that people find least interesting. What's the most important? Which is how are you gonna organize your data? What are the categories you're going to use? So important. And every time I get a new computer, I have to start off with these are the different file folders, sub folders in my downloads folder. And then I set up rules to have videos to the videos folder and PDFs to the PDF folder.'cause otherwise you just have this massive. After two years, 10,000 files in a single folder that are useless. Yeah, no, that, that's a great point. And in the world of data there and management in general, when you're doing, making data-driven decisions, there is a a saying, which is what you can't manage, what you don't measure. So you can't manage what you don't measure. And that's very true. And that's, I see that all the time. You're not gonna understand your processes unless you actually collect the data. You're not gonna know where things are going unless you have the numbers. But there's another perspective there if you switch it around, why bother measuring what you don't intend to manage? And the same concept could be applied. Why are all. What I find really exciting is developing workflows that take that information that's captured by some kind of meeting ai. There's Otter read ai, there's plenty of tools out there, there's native ones within Google and teams and things like that will transcribe it for you and even give you summarized reports with action items and sentiment analysis. But what's the next step? So the next step is actually thinking about a workflow that integrates AI tools into something that has a purpose. So once that transcript is generated, then maybe the synopsis is automatically emailed to certain key stakeholders, or it like creates or suggests a future follow up meeting. Like something that triggers action, not just it goes to archive and you feel better because. Yeah, I think that we like to skip the planning phase, so we go which data do you wanna keep? Let's just keep everything and we'll figure out later. Yeah. And I see a lot of businesses do this in a lot of areas and. It's leads different failure. It's like I see people do this. They build a TikTok following of 500,000 people. They make no money because the audience is broad and it's a non monetizable audience. They didn't do the planning phase first. They go, why don't you get followers and they'll figure out how to make money? And then you find out, no, you have to do what you're gonna do with it or what type of audience you want, and who's your customer. Otherwise you're gonna waste so much time. With a broad audience. And when your audience has nothing in common with each other, then no one wants to advertise with you because you don't have a specific audience. And this is what I see happening a lot, which is we go now that we have AI and automation, we can store everything. We'll just figure out what to do with it later. And the problem is it takes so much longer. One of the biggest issues I deal with is clients will ask me to accelerate a non-existent process. So with AI, I can take a process you have and make it faster, but they'll say, we don't have a process yet. So I go, all I'm gonna do is crash faster. It's like I'm missing that critical component, like accelerating something doesn't exist. It's two different phases, but we often skip the process design. I say it's better to design a process manually or in your life, and then move it into the computer and then move it into ai. Go through three phases, then you're gonna do a real success. But people, unfortunately, the hype of AI is so big they go, no, I'll just push the enhance button and it will do the thing I want. And it's actually, there's. AI is not quite there yet. You still have to know what you want and I wonder how often deal with this where people say I just gimme the data, and you go what? Show me an example of the correct output. And they go, I don't have one. Like anytime someone says I, I'll know it. When I see it, I know. I go, great. It's gonna cost you three times as much. It's gonna take three times as long because I have to guess and keep guessing until I, get lucky and create the output that you were looking for from the beginning, but that you will know it when you see it. Yeah, I literally dealt with that exact situation yesterday. It's part of my role. I help with internal tool development, and a lot of that has to do with moving, migrating, analyzing, interpreting data. And so we're in undergoing a migration from one CRM to. Essentially, I was asked the question, can you scrape all the data? Can you export and take all the data out of one and get it ready so that it can go to the next one? I said, yes, but I need to know what the proper import template looks like. I need to know what are the fields that we need. I need to know what are those processing steps. Otherwise, yes, I can just grab everything. But it gets really complex once you get into the multiple dimensions of data, because you could have data over time. You could have data drilled down by by certain criteria personnel department, et cetera. I need to know where this is going. I need to know the intended output, and then I can fill in the gap in between. So I often tell clients, show me where you know what you have, and show me what you want. And I can, figure out the magic in the middle of the Excel voodoo get there. And it's a journey. Sometimes clients. The most successful projects that I've actually worked on, and the most happy clients are the ones that have a process already that they're just doing manually. They're doing it very tediously. They know what the, why they're doing it, and they know that there's value and what they're doing, but they also know that they could be doing it better and way faster. And then I come in and say, ah, okay, so this is what a human is doing. You're doing this and this. Now I get it. What if I told you that I can take that hour and a half long process and turn it into one button that when you click, it does everything for you and it's just oh my God, I didn't even realize that was possible. But it's so much more exciting because what you're doing is you're accelerating something that they already know works for their business and has an intended purpose. Yeah. This is so important. It's such a step that people are skipping an AI right now, which is that if I don't have. Raw input and ideal output. I can't solve for the middle of the equation, which is the formula or the process or the AI element. My favorite thing is when someone will come in and go, yeah, I have a really complicated process. I have it completely codified. Here's all the steps. Everything's written down. Sorry for making your life so hard, I'm make no, you're a dream client because you know what you want. It's so much harder because. If you don't really know what you want. And this is why a lot of other people that are in service industries, like logo designers are always mad'cause they go just make it kinetic. And you're like, what does that mean? Like exciting but not too exciting. And it's like futuristic but not too futuristic and use all these really cool words that don't, aren't visual words. And it's really challenging. What they're really saying is just keep making logos. And when you make what I like, I'll tell you. And that's why that's an area, like every logo designer deals with that. That's why ai, that's why it's really good to prototype with ai. So then you can go, okay, here's a rough idea of what I like and that's so much easier because you have a sense of what you want. But this is something that's really common, which is that because people don't know what their problem is. Like most people who come to me, they think it's an AI problem, but it's an automation problem or a systems problem. 10% of the time it's ai. Rarely do I get to stretch my AI muscles.'cause most of the time they just have never dealt with an automation. It's almost always the data's here, we need to get it there. And what you brought up is really important that if you don't know which data is meaningful. So I worked on a private resource. Someone goes, I need metrics. I go, which ones? He goes, all of them. And I was like, metrics isn't a mass now. There are so many things. I had this, redirect, like a link redirect on my website for a long time, like I can make short links. So you go to serve master.com/x and serve master.com/y instead of a really long link. And I was, my website was getting slower and sore and I found out that this plugin was logging. Like 50 columns of data every time someone clicked a link. Like what browser they were using, what time it was, what's their time zone what was their IP address? Like, all this information that was like, what are you doing? And hundreds of thousands of rows of meaningless data. And I was like, why is this even an option? What person wants to know the IP address. If everyone's ever clicked a link on their website, like I have no idea why the data would be meaningful and I didn't notice it for years.'cause it takes a long time to get so much data that starts crashing your website. And they were like, you have one table that's 20 gigabytes. I was like, what? I have 400 blog posts. And so 90% of the website was this one table, like 90% of the data I was paying for. And we have this other thing that's missing I kind of wanna bring up with you, which is that we. We design a process, we set an automation, and people often don't check. It's like they set up a system to create backups and they never test to see if it will actually, you can actually reverse engineer, turn a backup into a live system until after you've had an emergency. And like I'm a super parent person. I generate three backups at my websites at least once a week on three different tools, on three different platforms. So one of them will hopefully work, right? So it's like the host does one, have a third party tool to have one built into the software. So it's you just, we have this tendency in it. I dunno if this is a Western thing only, but where we. Have a plan for an emergency, but we never check to see if the fire escapee actually works. Until it's too late. And how do you deal with like systems in that place? Like where there's a check element okay, we're storing the data here, we have this process, let's check if it's working. How do you handle that part of the process? Yeah. So when I'm, let's say as part of the development process, when I'm building an internal tool I often tell my colleagues like, I want you to try to break this. Often they'll come to me and say oh, I think I'm so sorry. I think I may have broken your spreadsheet, or I think I like, found something wrong. It's no, don't apologize. What I want you to try to break this thing because don't worry, I already have a backup and a backup of a backup. So what I sent you is one instance of this tool, like you can't actually break the system. You can only. Break that instance, that like copy that you have because right now, before we even consider building this out into a higher level whether it be a power app or something like built into a front end web app. A lot of the stuff that I do is essentially an MVP for something down the line. So we have a iterative process where people will, say, Hey, this is a very inefficient process. Can we find a tool that will fix this? Sure, no problem. Build in Excel. And then that'll be the MVP product. And let's test it. Let's break it. Let's see what we like, what we don't like. Let's see what else we need to add to this thing. And then in the future, we can consider, all right, maybe this needs to be added into our actual software bundle. But when we do that, I make copies and I basically plan for failure. I assume that I'm not gonna get it perfect the first time. I assume that. Macros. You're human. You're going to miss something. And so I always go through that testing phase and I release it into the wild. Say, go ahead, please try to break it, because I wanna find all those areas where it's gonna fail. I wanna know like where things are going to not work as intended. And if you just plan for that and assume it's gonna happen. You're not gonna get caught off guard. You're gonna find way more bugs earlier in the process. And then the final product when you get to it, whether it be like a month or two down the road is gonna be way better than if you assume it's good to go. Let's hope it works. Here we go. Yeah, that's really, I. Critical. One thing that's happening a lot now in the AI space is that every idea is getting funded and there's a super high failure rate for AI businesses, of course, because it's the hottest new thing. And there's this step that everyone skips, which is the what problem does it solve step. And as someone who's analytic and comes from an annual mindset, this is really important to me. And every time I build a tool, it's to accelerate a process that exists or to solve a problem. But we see a lot of AI companies building stuff that's really cool. And it's that's awesome, but you're gonna go outta business because Cool. Especially yeah, when you have a ton of money, when you're come, it's the heydays or the salad days. You have tons of money, you'll buy all these tools, but then you go, this isn't doing anything. This isn't boosting our revenue. This isn't making a difference. That's the first thing that gets cut. So the cool tools might last for a couple of years, but you have to. Really focus on these two critical components. I think that's what's really interests you about the way you build, is that it's either accelerating existing process, it has a purpose that we already know is there, or it's solving a problem. So it's moving away from a bad thing and moving towards a good thing. What I see a lot of conversations I have is everyone wants to solve the wrong problem. Everyone wants an AI to handle on their social media. I go, oh, great. How many? What? How would you feel? If you start talking to someone on social media, you realize it was a lot. Would you buy from that company ever again or is it over? Yeah. And I say, or they wanna do AI phone calls or they wanna do, it's always things that are like super boring and say, oh really? How often from social media do your customers know that things that are important to help you plan the future product? So you wanna, you're losing that channel where people let you know, Hey, something's broken. Something, your website's not functioning. It happens to everyone. Every a link breaks or there's some PHP update, something breaks. Sometimes your customers become the QA person that finds it because you've tested it, and then an update or a WordPress plug or something breaks it. If you remove the ability for your customers to communicate you through a channel, the more channels you remove, the more likely you are to have a fatal error and not catch it until it's too late. So I always worry that people, a lot of companies try to like trying to solve. Like human communication. Oh, the one thing I order a roof is talking to my customers. And I'm like, isn't that literally the most, that's the most important part of your business? And it's there's this micro creon book airframe, and it's like you never, even if you export the wings, you never let a leather company build the middle of the plane, because that's the key component of your business. You build airplanes and it's the same thing as if you get sale, if you output out, send sales to one an AI and have another AI building your software, you don't have a company anymore. Once they figure out they don't need you and you're giving up the most important thing. So yeah, I can have an ai. I guess sometimes people think that's what this podcast is. They go, oh, it's an artificial intelligence host. No, I'm a person. It's the topic, not the mechanism. Yeah. But someone asked me that. I go, oh, that's a really interesting question.'cause they read, you read a word differently. Oh, that's insane. Who would, nobody would watch it. There's a lot of tools now that will generate AI podcasts. Again, no one listens to those'cause it's not interesting. We don't wanna listen to two robots talk to each other. That's why adults don't watch very many cartoon movies.'cause it's not realistic. Like, why do I care which cartoon dies in this fight? It's two cartoons fighting. They're not real. So I think that this is so important to Island too, which is that. The process elements. One thing that I've seen a lot of if you wanna start a business like 30 or 40 years ago, you just wrote out a really detailed business plan. Went to the bank, got a loan, they reviewed it to see if it was good enough, and then they would loan you money or not, based on the quality of your plan. We don't need to do that anymore. And now nobody has a plan. I so many people skip this phase, which is the, how do you know you have a good idea? Phase. And I think it's a critical component and it's also it just iterates further and further down. So someone will say, oh, I wanna accelerate this process. What does this process matter? I don't know, but it might like, oh, great. And that happens a lot. Like a lot of times because we built our businesses and we haven't been fastidious along the way, which is what's your, who's your ideal customer? I can always tell when someone, the broad or the answer is that the more in trouble they. If we just get 1% of all humans and like 1% of all humans, yeah. Name name a product that 1% of all humans use any product for any brand, right? It's just not there. And that's what's really important. And it's like that we think that AI can solve a problem, but we're solving the right problem. It's the strategy component. And AI can help you strategist and help you figure out if you have a good idea of all those pieces. If you do it, we're like, that's boring. Let's just do the exciting stuff. So I'm interested because you have, there's two camps of people that come to spreadsheets, people that are passionate about it and people that hate it and keep no records. So what do you do when someone goes, Hey, we have no records, we have no process. Here's like a log into our software. Can you just fix this for us? So in those instances, I actually go back to my roots, which is quality management. Like I was a quality control technician and really focused in on Lean Six Sigma, lean manufacturing, things like that. And there's a framework called the DMAIC process. And the first thing you need to do. Define the problem. You define the problem, then you figure out what are the things you need to measure. Then you figure out like, the analysis phase, how are you gonna analyze the things that you capture? Then you improve the process and then you figure out the proper controls. So DM A IC and that process works whether you are working on an AI tool, a spreadsheet, it could be you any a new business by defining what measurements. Even more important to figure out what the problem is because when you identify what it is that you're actually trying to solve for, then you can hone in on, okay, what metrics do we either have that we're not really paying attention to? Or what metrics can we generate? What metrics can we extract and start collecting to establish a baseline that will help us? Create some kind of analysis or improvement down the road. You can't do that unless you have the measurement, but you can't measure unless you know what it is you're trying to understand or fix. And so that's the process that we walked through. And to your point around ai, basically short cutting that analytical process and then people not really paying attention to. We have so many we have so many instances where we'll have spreadsheets with data, and all we wanna do is automate it. We wanna basically say, can we create an AI that's going to just build this cool dynamic dashboard? Okay. Do you understand the data that you're asking the AI to analyze? Because if you don't have a fundamental understanding of what this data actually is. Then you're gonna miss the entire point. You might have a beautiful dashboard with, charts showing you that things are going up and everything looks good and green, and heat maps and everything. But if I ask you a question about the actual raw data. Are you gonna look at me like I'm crazy? Are you gonna understand what these numbers actually are telling you? And that's why in analytics you can have descriptive analytics that describe what's going on. You can have diagnostic analytics where you look at, okay, what's the source of the problem? Is there a correlation between this and this? But really it's once you get into predictive analytics and prescriptive analytics that you start asking this question, what am I actually trying to achieve here? What am I trying to do? And what do I do once I get this variable? So the AI spits out X. Okay, is that good? Is that bad? Do I need to do something about it? Am I like, okay, or is this like a red alert? What does it mean? Yeah. I think that's the critical component is that if you can't tell if. The response is good or bad. Or if you can't tell, then there's no point in generating it. And I thought it was funny when you said that. I was like, how many people ask for that? And then they don't even know what dynamic means. So it sounds cool. Yeah. So I think this is really gonna help people who are thinking about AI processes and hopefully this want people reach out to me, I'll work with me, is they'll first thing, oh yeah, I need to know what I want and be able to describe it and to know if I've received what I'm looking for. Because if you can't tell, if you've received a good output, then you're in this really. It's an impossible to solve situation. So this has been amazing. Christian, I really appreciate your time. For people who are wanna improve their processes, especially with numbers, especially with Excel, and realize that their numbers, their metrics, their data is not where it needs to be, where's the best place to find you and see the amazing things that you're doing online right now? Yeah, so I've got a couple of platforms. You can always reach me on my website. So it's stark analytics.com. Yes, stark as in Iron Man, I'm a very big Marvel guy. People ask me like why he's my favorite hero and I said one, he wasn't magically gifted powers by the sun or something crazy like that. He's not an overpowered character. He literally used his brains and ingenuity. To create a solution back in a cave. And what's really cool is you look at the evolution of all the different suits. That's what we do with our own systems, right? We start off with a really ugly, clunky, hunk of metal, and then eventually we refine it and we make it better. So I just love that like journey. So it's stark Analytics. And then I also have a YouTube channel, freak and Freak is basically like a little bit. Customer facing sub-brand where I just do short training videos to help people learn what's possible. Like I'm really at the point right now where I'm shifting away from what can I build to, what can I teach, what can I empower people with? And so a lot of the stuff that I'm doing now is, skills training workshops. I have a, an accelerator workshop where I basically cover. The newest features in Excel to how do you integrate AI into your process with Excel and empowering people to use these tools to do things better because, sure, you could call me in and show me your process and I build something for you, right? I can give the person a fish, but if I teach them to fish, that's even more impactful for them and in the future. On the website has details on there, learning channels, sheet freak, TikTok, Instagram, all those things. I'd love. Connecting with people and talking about this subject nerding out on it. And yeah, I've pretty much made myself available and accessible through every channel. And it's so much fun. It's so much fun, like hearing the stories that people have when they embark on these journeys, things that they've built that they get excited about, and it's yes. Okay. I want, I wanna work with you and I wanna show you how you can take this to the next level. That's amazing. So we'll make sure to put all those links below the video and in the show notes. Thank you so much for being here again for an amazing episode of the Artificial Intelligence Podcast. Thanks. Thank you for listening to this week's episode of the Artificial Intelligence Podcast. Make sure to subscribe so you never miss another episode. We'll be back next Monday with more tips and strategies on how to leverage AI to grow your business and achieve better results. In the meantime, if you're curious about how AI can boost your business' revenue, head over to artificial intelligence pod.com/calculator. Use our AI revenue calculator to discover the potential impact AI can have on your bottom line. It's quick, easy, and might just change the way. Think about your bid. Business while you're there, catch up on past episodes. Leave a review and check out our socials.