519 Is AI Making Junior Accountants Turn Off Their Brains?

Today, is AI just going to make accountants turn their little brains off and just trust the robots to do all the work? What about the youths? How do we balance leaning into AI with like the foundational knowledge that you still need? Also, did you hear Microsoft launched a copilot function in Excel? It's like AI built into the spreadsheet cell. Also some new tech for cloning your identity. That's specifically for accountants. And the big the granddaddy question, how do we make sure our accounting firms aren't just totally wiped out by AI? I. I think some people are actually coming around to some wild stuff we've seen from AI lately and they're like, oh crap, this actually is feeling a little more real than I thought it would now, maybe a little more concerned. We got a big old AI mail bag, Q and a grab bag. We're covering it all today. K, come on in. This is a Q A episode from the comments y' all have left. Thank you very much. If you've got a question that you're thinking about, somebody else is too. Pop that stuff in the comments. We have a conversation about it. It helps everybody. This question, it was a great one. After talking through Chat GPT Agent mode specifically, which is really, really powerful. After talking through the new chat GPT GPT5, using it for research. A great question here from CBS 450. Doesn't this defeat the purpose of people actually learning? I would just have a bunch of uneducated staff members under me incapable of actually researching anything outside of going to ChatGPT or thinking for themselves. Saving time is one point. But the purpose of a junior researching a tax question is to learn this teaches them nothing except that they are expendable at the sake of profits. Elephant in the room. It's a great question that lands so totally differently with folks. It is such a big spectrum of how folks are processing, assessing this problem right now. One thing first and foremost that I, I think we always have to keep in mind is every big technological change comes with a version of this. This pang of this wasn't my experience. How will anyone else ever get to where I am right now without walking the same steps that I walked? And if you think back specifically through the lens of tech and accounting at all the examples we've had of this, and there's some, there's some big ones. Imagine the first time people were using software to prepare tax forms and not doing it on paper. And I know a lot of listeners are like, I still remember what that was like. I did a lot of that whippersnapper, the first guy I worked for, he said it was essential that I could still prepare a 1040 on paper. And it was a little exercise that he made us do. If you spent decades completing paper forms and then somebody came along and said, this computer is just going to calculate everything for you. You're going to, first off, I'll believe it when I see it. But second, this is what I do. I calculate stuff on forms. If the computer does this, what am I even going to do and how am I going to do it? Well, without the knowledge of how to prepare that form, I mean, even just like putting the accounting ledger on a computer, I had a partner I worked with that always prided himself in sharing that his first job was like footing totals and running these massive 10 key tapes. That was all that he did. And he always shared this story. And I'm like, what do you want us, how do you want us to react to this? Any way besides just thinking, man, you're right, you are really old. But he would very proudly share this story. And this is, I think this is just inherent in change is we only have what our own journey was, to perfection, which is, which is where we are right now, right? And because I can only see of how I got to being to being perfect via my own journey, if someone else isn't going to have the same journey, then I have a hard time ever envisioning them as becoming what I am today. Right? Rising to the level of me. Obviously, obviously this is tongue in cheek, but the reality is I think here that we always have to balance with change is the way that we did something before that is going away and we can see that there will be a new way, but we're still grappling with what that new way even is. And we can't see the nuance in it, we can't see the upsides in it yet. But if you strip this back to the core of the question, how will you know how to do anything if I just does it for you? There's a lot of validity to that. Like there are just things you won't have to do anymore. I mean, when we got software research tools and you didn't have to go to the library, like the library in the office of all the books and find the answer, that was a big change. How are you ever going to learn how to find an answer if you just trust the software to serve the answer up to you? AI is like that on steroids. But even if we look at the example of chatgpt, the notion of chatgpt making us dumber. I can tell you the most happy ChatGPT users today are using it as the ultimate tutor, Learning how to do things they never could have dreamed of doing before, writing macros to solve their workflow headaches. The other day I needed to get an export of every email I had sent out of my Gmail account and chatgpt wrote like a Google Apps script. I had never done an app script before. I just described what I needed it to do and it said, here's an entire app script that you can use and it will get you that. And I was like, I don't even know how to get to apps script. And it told me where to go. I put this like 200 line block of code in and it said click run. And it gave me exactly what I needed in. In an Excel sheet. Not an Excel sheet, in a Google sheet. Is that making me dumber? I. I don't know. I can tell you ChatGPT is giving people powers they've never had before. And if you use it the right way, it is the ultimate tutor to unlock new skills. So it like, I totally get it. When we all stopped working in the farms and the industrial revolution happened and we weren't working fields all day, did our bodies get weaker? Probably not, because they were working in factories. That's a bad analogy. Look at white collar work now where everybody just sits on their cans all day sitting at a desk. We all just turn into, into slugs. As long as our little legs can bobble us to and from work. That's all we need our bodies for as we just jack in to that 17 inch Dell monitor on your desk. The people living that lifestyle now, they have home gyms, they're like, I got to work out every morning so I don't get fat and unhealthy. Meanwhile, the contractor comes over and he's already like 30,000 steps into his day and he's like, I love that for you. I love that you have to go out of your way to exercise. We are entering just like a fundamental, fundamentally different age of how to work and it is going to destroy like rug pull a whole lot of old ways of working. In fact, in this video that they're commenting on, there were some really cool examples of using AI to complete this spreadsheet end to end. And since then I've actually got a number of emails from folks that are like, I just gave it three credit card statements and three bank statements and it gave me like A full, like, completed ledger, probably to the level of a junior person in less than five minutes. Like, I've heard so many stories of that now, where it's like, it's impressive that it's pretty darn good right now, like at a junior preparer level. But the bigger thing is right now, you would never go back to doing it the old way. And every new model makes that entire process better. So we are already closing the door on keying in transactions from a bank statement. If I can chuck a bank statement into an AI tool and it gives me a perfect extract of everything on the bank statement, I am never going to key that stuff in again right now, should that problem exist in the first place. No, the fact that I only have a bank statement, I don't have a bank fee, I don't have a transaction download. And these are many of the problems AI is solving for us right now is things that shouldn't be problems in the first place, but they are. And AI is like the greatest bandaid we have ever had for solving these dumb problems that I wish didn't exist in the first place. But day by day, the old way of doing things, it is. They are going away in record time. And that's going to be a hard thing for people to get their heads around. But if you are the person that did that thing the old way, here's the thing is that insight is incredibly valuable because little whippersnapper coming out of school and coming into the firm now, that just blows past all that stuff that you had to do. They're going to have unique skills. They're going to be very, very good with AI because they just got through school on AI, so they're going to bring some interesting things to your firm. But it is going to be your perspective, having walked the, the, the slower path of doing things in a manual way and all that, it is going to be your perspective that helps round out that person. That's what makes your perspective still so important, is the best version of like, what we do moving forward is the combination of both of those things. I need the youths, I need the people who have spent thousands of hours with AI already coming into my firm to, to help all of us old farts develop, for goodness sakes. I will give a talk about AI to a whole bunch of accounting firm owners. And they're like, I love it, but boy, how am I going to find any time on Monday to do this? Meanwhile, you got a whole bunch of zoomers running around that are, you know, falling in love with their AIs. Both of those things, they are problematic. What we need is, is the balance of both and the perspective of the old guard to say when you come in, if you bypass X, Y and Z, we still gotta somehow get you like this sort of context to round you out a bit. And that's probably not by making them go to the tax library and pull the book on the shelf to do research the slow way. It probably looks like helping them get a broader understanding of like, what are these sources that it's pulling from? Like what are authoritative sources and what aren't. And then when you start leaning into AI more, maybe with the help of some of these folks that are more adept with AI, then what about that? The world is your oyster at that point because you have both. You have the history, the context, and you have AI. This is, this is your game to lose at that point. Buddy and I, I think we are still so early on in this, where there's going to be so much grappling with the speed of change, that this is going to be like a tough thing for us to navigate. It's, it's not too dissimilar to, you know, some folks struggle to make junior team members useful in their firms. And, and we talk about this in the content a lot, the nature of deskilling work. How do we design systems to ensure repeatable outcomes? Do I have a system that ensures we can consistently deliver results to clients within the expected timeframe? Like, have I designed a machine that's capable of that? And that comes down to how well can we delegate, how well can we systematize? Are we overly reliant upon expertise and people that just have experience? Or can we create systems that, that will allow even less skilled people to contribute to that process? That's a learned thing. Some folks will really struggle to get anything out of interns out of junior folks because they haven't invested in how do we train people, how do we develop systems not to be overly reliant upon expertise. All of those same limitations, I would say, apply to AI usage as well. There's people right now who can take AI in its current state and it be a tremendous resource for them. There's other people who can go to the exact same AI and they're not sure how to ultimately save time with it. Right. And so a very practical example, if we bring this back to research is the way that I prompt AI to give me a research answer needs to be the same way I would prompt a junior person to bring me an answer that is provable. I don't just have them research something and have them come back and say, here's the answer, and that's it. If a junior tells me what the answer is, I kind of don't care. I just want to see their work to be able to prove that that actually is the right answer. And so the goal shouldn't be to get to the black box answer fastest. It should be to get to a provable answer fastest. So the folks who are most comfortable with AI right now, they're going to prompt it the same way they would prompt a junior, where they're like, here's what we need to get to. Here is how I want you to organize your answer in a way that's easy for me to review and confirm. Because the goal here is to prove to me what the answer is, not just give me an answer, but if you got somebody that's still starting out with AI, what they're going to do is they're just going to rock up to that prompt box and say, what's the answer for xyz? And it comes back with an answer. And then they're like, well, now what? Now I got to go read all the. All the backup stuff. I got to do all this homework and basically grade AI's paper to know if I even have something I can use. Similar to how turning other human beings into producers that are contributing to the value of the firm. You got to learn how to use AI. It a valuable contribution to the firm. And this is not just something that the big bosses need to be learning. It's something we need to be instilling in our entire teams. And that happens by just using it every day, having it up on an extra monitor, getting the team together every week to share how they're using AI, what's working, what's not working. It's an entirely new approach to work where every single member of your team is on their own learning journey. And we need to be ourselves growing as we grow the folks on our team. Because the upside is so significant. You do not want to wake up two years from now and have an entire team of people that don't know how to use AI. It's going to suck on a number of levels, and you're going to start sniffing around on the job market and be like, who can I hire the knows AI stuff. But the worst part of that is it'll be your fault. These people are clocking in for you every day, and if you don't show them the path for how to continue doing their jobs with AI Man. I don't want to leave that person in a worse place than I found them when they came to work for me. So there's a hundred versions of this conversation like, how are people going to be able to do their jobs if they don't have all the same context that I did? And oftentimes the answer is, I don't know. It does seem like that would be essential to know what you're going to do going forward. But you can also look back at a hundred examples in the past where people would have said the same thing, and we still got to where we are today. Honestly, I think it's an opportunity for the old guard to speak into the new stuff that's happening and round out the folks who didn't get all the same context that you did. Foreign this episode is brought to you by Team up. Who will find you talented accountants in the Philippines. Team Up. Their motto, hire a person, not a company. So Team Up's difference is there for the folks that literally just want to engage the people in the Philippines to come and work for them, rather than like Accounto Corp, who then employs a bunch of Philippines accountants. And then you pay a Counto Corp. And sometimes the problem with Accounto Corp is you're paying as much to Accounto Corp as the actual individual that's on your team gets. And that's kind of a bummer if I'm paying 2000 bucks a month and only a thousand of that's going to my team member. Very common progression that I see with folks building offshore teams. They start with Accounto Corp for the first few hires and then they're like, we just want to do this ourselves. We actually don't like the middleman. And oftentimes it's actually impossible or insanely costly to get your team member out of a counter corp. So then you kind of like Trapped Anyways. Team up, they've got recruitments, recruiterments on the ground in the Philippines who will find talented folks for you. You just work with them directly. Same way you'd work directly with somebody in the States. If you haven't done it before, you're gonna have a bunch of questions. What about computer stuff? What about security? The answers for those usually the same as whatever you do for your, like, domestic team. What about disclosures? If you do tax work? Yeah, you gotta put something in the engagement letter. If you're using em for 1040 work, there's a standalone disclosure that you do. But I did both These in my firm wasn't a big deal and team up. They've helped so many firms go through this that they can give you like great expert answers and best practices for how to navigate any of these things that may be spooky. Even if you just want to have a conversation to learn more, hit them up. That's team up. Learn more. Check out the link down in the show notes. If you've ever seen my tech selection pyramid, it's basically like Maslow's hierarchy of picking out tech in your accounting firm because picking your firm in the right order is very important. You start with the foundational tools and then you layer in the automation stuff on top of it. And wouldn't it be nice if your foundational tools could just do all of that stuff? One of today's sponsors, Client Hub, they're trying to do just this smooshing several of those layers together because they're a practice management system. But they're also doing really cool month end close automation stuff for bookkeeping practices. So you get all the practice management goodies to like manage all aspects of the firm, but it's got a super deep integration with QBO and 0 and has like a step by step process for managing and every aspect of the close. Even like some cool stuff like pulling in uncategorized transactions and you can push those straight out as a client task all inside of Client Hub. Core part of Client Hub is kind of having like this two sided task management system where there's the work that you do and anytime you need the client to do something, you push that task to them. Because the reality of getting work done in an accounting firm get it work together. We're most often blocked by this stuff that we need from clients, right? So pulling that all together into one system and doing some cool month end close automation stuff all in one place. Client Hub is doing a really cool version of that. Now if you're a bookkeeping practice, it ought to be on your short list of PMs to look at. Learn more about Client Hub. Check out the link down in the show notes. Okay, question I got from someone that was actually about a vendor in the tax space. And then we'll get to this new Excel copilot function. That's pretty cool. Basically the company said our approach to helping with your tax workflows we can automate basically the entire 1040 process. And I wanted to flag this as something to just to kind of help your your BS detector. I think I've shared this story on the podcast before but one of the most disappointing conversations I've ever had with a tech vendor was with a pretty high profile company. I met with their vendor outside of a conference in Vegas and quickly became clear just how profoundly unaware they were of what accountants actually do, and particularly around tax workflows. Right now there's a lot of companies that think what we do is put numbers from government forms into our tax software. As if, if only we could perfectly extract the numbers from all the government forms and put it into the tax software, then we will have automated all of 1040 prep. And if you run a tax shop, that is just not what you do. In fact, those returns that come across your desk, we should probably be referring Those people to TurboTax, to H& R Block. That's probably not work that we should be touching. The government forums are like 10% of the equation usually. And so when you hear stuff like this, they're going to automate the entire 1040 process. And I've actually seen this countless times now because we had a lot of folks in the OCR era that said they were going to do the same. They were going to do all the extraction from the W2s and 1099s and brokerage statements and all that, and then they would integrate with your tax software. And really only two or three companies survived that. You had SurePrep, you had Grunt Works, and then you had the first party extraction solutions from each of the tax softwares. But that's not 1040 prep, that's not tax prep. There's so many more things that go into it than government forms. So it is is worth having your BS detector up in, especially in the age right now where people are making a lot of promises. If I hear that, that's a very clear signal that they aren't far enough into this yet to understand the problems that need solving. And the bigger point here is this person's having a conversation with a new AI vendor and they're like, this is really exciting. I think I'm going to, you know, I'm going to commit some returns to them and help them out and help them develop. I am such an advocate of using the platforms that already work. There's already cool AI platforms that will do tax workflow stuff. What is the rationale for going to a brand new company that just started and really doesn't even know they're doing yet? When I got returns, I got to get out the door now. And there are thousands of firms that will stand by other like AI tax workflow solutions don't get sucked into the allure of they finally cracked the code because that landing page copy, it is usually paper thin. And the bigger and more hyperbolic it is usually that is a signal that they don't yet know what the problems even are that need solving. And if we go back to the tax research example where it just spits out an answer that is a great example of something that is not useful for me, a tool like a black box tool giving me a completed return, that's the same as ChatGPT just giving me a research answer where I'm like, well now what do I do? Like how do I make sure that it's right? And you can argue like yeah, but it'll do the review too and all that. And I'm like, well then what am I? Just like a signature jockey? Like I'm not. I can't sign my name on something that I haven't verified. Like that's the entire point of what we do. And it's why the most successful companies in the space right now, they are implementing AI into an actual process where it says here is step one of getting this work done. Here's step two, here's step three. AI is embedded in the getting done of the actual work. You see this in bookkeeping apps in month end close automation. You also see it in the leading tax workflow tools. They are not just doing this one razzle dazzle thing, they are giving you a way to actually manage the process with AI embedded in there to assist in all of the steps that it can assist in. It's not me giving the tool a pile of docs and then giving me something finished back. That's not realistic. And the folks who are saying that like they're just not far enough along yet to realize that's not where we're at today. Be discerning when you're, when you are talking with especially new AI vendors. I always tell people when they come to me with a brand new app and they're like, have you heard of this one yet? It looks so cool. There's definitely a chance that you're the only person that found that apparently. And you know this is really going to change things for you. But the much more likely scenario is it stinks or it's brand new or a bunch of other people have used it and they left it. And it's why my recommendation is like go and use the stuff that works. And we're not talking like old stuff. Like I am very much early adopter, but it's A balancing act here, because if I've got 100 turns, I got to get out the door. A thousand returns, I got to get out the door. I want to run that through stuff that I already know works. I can for sure test new platforms and support them along the way. But like the core getting done of the work, I need reliable tools that will get that stuff done for me. Okay, we got a new Excel. New, it's a month or two old now. Copilot function. So inside of Excel you do equals copilot. And then whatever you put inside the parentheses next to the function is basically a prompt that it throws to copilot and then it gives the response from the AI model in that cell. So, so, for example, I could write a formula in Excel that was, equals copilot, open parenthesis, what's the capital of Oregon? Close parenthesis. And then it would load for a second and then the value in the cell would be Salem. And so this is cool for some stuff, but not a replacement for like math stuff that we already have formulas where it can handle. I would say it's better for more like open ended things, you know, the capital of the state being a good example. Like if you had a whole bunch of states in your spreadsheet and then the next column over you have the copilot formula and it says, give me the capital of this state. And then you point it to the state cell and you pull that all the way down. You know, it'll give you the capitals of all those states. You could also do that for population. Ask it for the population of all that, all those states. Pull it all the way down and it'll give you numbers for all those. You can say. If you're looking at a client list, you can say, see if you can go out and find me the LinkedIn for this person and then give it the first and last name, maybe the company they work for, pull that down and then copilot will actually try to get that stuff. I don't know if copilot, I don't know if it only uses what the AI model knows or if it'll actually go out and do web browsing. But we've had examples of this sort of AI formula in other spreadsheet tools. I could tell you one way that I use this that's actually really, really cool is we've got a big old CRM of like all the subscribers on the newsletter and all that. And next week I'll be hosting an event in Salt Lake City, the Big Salty and What I can do is I've got all of these folks and you know, it's like 20,000 people and we've got all of their, most of their locations. We've got a tool where like it can go to their LinkedIn and if they have their location on their LinkedIn we can enrich the CRM with where those folks are, city and state generally. But that doesn't tell me if they're near Salt Lake City. I can go in and pull everybody who's in Salt Lake City, but there's a whole bunch of folks that are within a one or two hour drive of Salt Lake City but live in other cities and I don't know what all those cities are. And this is an example of where the copilot function could do this. We use a different like marketing tool for this but you create another column and you give it the prompt is this city. And then you point to the city field within a two hour drive of Salt Lake City and then you pull that formula all the way down. The last part of that prompt would probably be only respond yes or no. And it can populate all that stuff. I actually just did that across 20,000 records. It's super, super cool. To be clear, I wasn't using Copilot, I was using something else. But AI built into your spreadsheet, it does open up some cool possibilities. It also opens up a ton of misuse. It's kind of like coming back to AI being a band aid. There's going to be people who can't figure out a formula and they just end up chucking it in Copilot. Like they can't figure out the right way to do it. So they just write a big long prompt into an Excel formula and then point it to a few different cells and say figure it out. And that might work, but a lot of times also it won't work because it's not perfect. It's another example of where it's a cool novel new AI thing but we're actually just blocked on the use cases. Like we've never had something this flexible so you don't even know where to begin to use it. That's the coolest example I've got of going down a list of people and seeing like who's within a couple hour driving distance of a city. The biggest bummer about this new copilot function is you have to have the, the paid co pilot business plan. I think it's like 30 bucks a month. You got to have that to get the Excel formula. And if you've heard anything I've said in the last three years, it is that never buy Copilot before you buy ChatGPT because it is just so, so much worse. Really sucks, man. Microsoft has so dropped the ball on all things AI. It's kind of a bummer. Two more questions in the mailbag. Have I seen It's a software called Arrive. Website is Arrive Accountants. The headline is who would you clone within your accounting firm? A new infrastructure built to enhance productivity, removing the need to hire more labor. There is a tag at the top, right above the hero text with a little award ribbon that says Best accounting firm platform in 2025. They exhibited at Scaling New Heights. I've talked to them a little bit. The quickest way to turn me off a platform like that is making dumb claims. And even just that framing, I don't want to dunk on them too hard here. Even just that framing. Best accounting firm platform in 2025. What does that even mean? Accounting firm platform? If you dig into the copy on software companies websites like this, you can usually see, you usually kind of suss out what their understanding of, of the environment is, especially in the tax base. You will see they make reference to CPAs. When it's like, I think what you actually mean as tax professionals, usually you can, you can suss out how far along they are. So still early days for this tool. But somebody asked, what do I think of this? A dedicated tool for creating AI avatars for accountants. This stuff's been around for a while. We've talked quite a bit about the assistant that I have set up on Delphi. It's basically a chatbot that's trained on all the episodes of the podcast. It's cool, it's novel, but the, the framing sometimes here is problematic because ultimately it can't be a replacement for you and for your identity. If we think about where we're heading and AI being tremendously useful for getting a bunch of work done, what's left on the other side of that? It is us, it is our relationships. That's the most viable thing that we have. So right now, if I deliver financial statements to my clients with a little loom video each month and they can see my head and a little talking bubble next to their financial statements, that's the last thing I'm going to automate with AI. So my framework for this basically is if there's a human element in that delivery right now, I would not replace that with AI. If there is no human element at all, then Maybe, like, maybe it makes sense. And a good example of that is like a chatbot where they can get instant access to your expertise. It's not as good as the real deal, but I also can't talk with a hundred people, 24, seven. And so there are things that, that can do as well that you can't do specifically around video. I think we have to be careful to not send things over video unnecessarily. And I would say most of the time an AI generated video could have been an email. Don't make me watch a video if it's something that I could have read in text. To me, the value of doing tax return deliveries and financial statement deliveries over video is you can more effectively communicate the nuance and also show them as you describe what's happening so they get an understanding of the tax forms or what they're seeing on the balance sheet. That's the value of the video, is the talking with the seeing and also the upside of them seeing your face because oftentimes they'll go a whole year or they'll go a whole month without seeing you anywhere else. And so it's a good reminder of you. And you can kind of express some humanity there. But for example, last time I closed on a house, I got an AI video from, I don't even remember who it was in the process. And I sat there and I watched it and there were no actual people in it. And it was like, just like a custom sort of two minute video that pulled in a bunch of variables from, from the home purchase. And I'm like, I just, I would love to have those two minutes of my life back if there was anything important to communicate to me. Like they made me watch that video as part of a process. I'm like, man, I could have read that in like four bullet points. So I like, there's going to be some cool applications. A very good example. And we will put the AI assistant for this episode in the show Notes. If you haven't played with it before, it is pretty cool. Cause it's trained on all, like all over 500 episodes of this podcast. And what it is most useful for is search is Man. I remember he said that one thing in a podcast a long time ago, but now I can't find it. That is what makes it really valuable. I will even use it when people ask me like, when did I talk about X, Y or Z? Like I will go to my own chat assistant and ask it the question and then it'll give an answer. But It'll give citations. So it'll say, you know, 21 minutes into this episode from two years ago, you talked about XYZ. That's really, really valuable. So some of that is interesting. But even if we start doing more client facing chat assistants, the best chat assistant will be the one that's connected to the client's context. That is their past tax filings, their accounting ledger and all that. I really think it's the practice management systems that would be in the driver's seat to do that. Well, because they have all of the context that can make it a good assistant. But like we're still kind of working through what are the appropriate places to use that and the, the ways that just feel weird where we shouldn't be removing from the equation. You know what I was just thinking about just how much I accommodated my clients. Crappy old banks that just created so many workflow nightmares for us. This is, this is an ad read for Mercury. Just so you know, literally having clients share their password with you. So you could log into a client's account to a bank. Having to have them send you two FA codes, little two FA codes where the text literally says, says never share this two FA code with anybody. So many problems with that. Don't get me wrong. Like, it's an essential part of delivering like bookkeeping work. But then I would be so apologetic, like, yeah, no, I'm sorry. This is such a headache. I wish it didn't suck like this, gang. There are solutions for this. Like Mercury. I literally ditched a traditional bank and used Mercury now myself. The client gives you firm access and then you assign user access for the people in your firm. And it can get super, super granular, like the way that user management actually should be. So next time when you get that stupid situation with a client that is not secure and they're annoyed about two FA Olympics, just remind them, buddy, this is, there's a, there's a solution to this. We can help you with this. We'll help you make the switch to Mercury. You'll have a vastly better experience. Get a bunch of razzle dazzle features your, your traditional bank doesn't have. But then we can lock it down with proper access to whoever needs access to it. So the next time they complain, don't apologize, say, here's the solution. Do you want, want us to help you through it? Think about it. All right? To learn more about Mercury, check out the link down in the show notes. Funny thing happened recently when I was at a conference with one of this Episode Sponsors Makers Hub so Makers Hub, they're like an AI accounts payable platform. The party trick is they're using like AI vision models to do extraction at a level that was kind of never before possible with OCR technology. But I was doing something at their booth. Some of the customers came by like it is one of those tools where the customers are kind of raving fans. But a funny thing kept happening. Like people kept throwing out these ridiculous document situations they have where they're like I got this client, I have X, Y and Z problem. It's a total mess. We spend so much time on it and they just start riffing on it like with the Makers Hub team and they're like oh yeah, no I can see like you do this potentially, like our product doesn't really do that, but actually if we change this or that, this could potentially work. The team is just like, they're just a bunch of like document processing geeks who are also AI nerds. And the product very much is like the marriage of those two things. And so what Maker Sub does best right now? Like the biggest, ugliest accounts payable stuff that you got, man, I had some gnarly clients where Makers Hub would have saved us so much time. If you're out doing conferences at all, you'll see Makers Hub a lot add into a connect. I'm going to be there I think. I'm actually co hosting a thing with Maker's Hub. But if you got any ugly bill pay stuff, class tracking projects, inventory, manufacturing, stuff like that, definitely worth giving Maker's help a look. To learn more, check out the link down in the show. Notes some context for this last question. I shared the story recently of, of. Of someone on, on LinkedIn, a tax pro. They used ChatGPT's agent mode, which can navigate a web browser on its own. They used Agent mode to complete a client's tax return in TurboTax. And so they didn't give it real information, but they gave it some made up information like W2 and stuff like that. And ChatGPT's agent mode went through the entire process in TurboTax of completing the return. TurboTax tried to upsell it umpteen different times. It navigated all of that and in the end the return it prepared on its own in TurboTax was the same as the return that the tax professional generated out of their professional tax software. So this was a very simple return, basically just a government form return. But ChatGPT's agent mode is already capable enough to walk through a simple online tax prep system like that and prep the whole return. Wouldn't it be wild if today your tax software was web based? Intuit Pro Connect. People right now are like, yeah, finally we did it. We're already to the point where ChatGPT's agent mode can do meaningful work inside of online systems. And it's stories like this that people are starting to see and kind of getting their head around the scope of the impact of AI. I think that leads to questions like this. How can the firm remain valuable to clients and avoid being totally replaced by AI tools? They said they're mostly receiving inadequate and surface level answers at this point. That is the million dollar question. People love to sugarcoat this one, man, every conference you go to and every software vendor is real quick to throw that blanket around your shoulders and say, everything's going to be fine, everything's going to be all right. The reality is it'll be something in the middle. Some accountants will not be all right. They will set up camp in a place where AI is capable of coming in and just doing all of that work and then they're going to be in trouble. There's going to be a lot of accounting firms that are going to be totally fine. There's going to be accounting firms that lead the charge into during doing work with AI and it'll be a huge opportunity for them. I think a lot of them are listening to this podcast. But a big thing worth thinking about as we're grappling with like what is our identity going forward is the answer to this question really boils down to what you would say you're in the business of doing. Are you in the business of completing payroll allocation spreadsheets, of classifying bookkeeping transactions? If you're in the business of the doing of the tedium, then yeah, you're probably going to have a hard time in an age of AI. But if we find a better way of communicating what our value is, for example, I would say my role running an accounting firm is, is quite simply to help entrepreneurs overcome blockers to running their business. Blockers related to accounting, blockers related to compliance. I didn't run a bookkeeping business. Like we partnered with small businesses that didn't want to deal with this stuff and we handled it for them on the back end so they could run a more lean team and they could focus on what they were passionate about and we could handle this stuff, handle that back office for them behind the scenes. Many aspects of how we did that work are set to change in an age of AI, but on the other side of AI, are there still going to be blockers for that entrepreneurial entrepreneur to be able to run their business? Are there still going to be blockers related to accounting and related to compliance? Yeah, there's no like super strong AI timeline where you just snap your finger. It does all the accounting, does all the tax, it works out what all of you know, the politicians, how they want to complicate things this year and it just perfectly navigates all that. Like I see a near future world where there are no blockers for entrepreneurs to do what they do. And as long as that's the case, I'm going to still be in business because whatever that means and whatever that looks like, that is my role. That's what I sell. That's the value of what we do. And the even better version of that messaging is the one that's through a specific type of client's lens. Because ChatGPT will always be most powerful in the general applications and the value we can provide is in like the seasoning in of the nuance of. Well, yeah, that general thing doesn't apply in situation X, Y and Z. I know that's the right answer for most folks, but it's not the right answer for people like you. So if we can keep getting more specific about who we're talking to, I think this is also a great way to, to protect ourselves from AI. But how do we avoid being totally replaced by AI tools is by doing things that AI tools can't do. And a lot of that's going to be in our relationships and creating cool experiences for people understanding what their problem is and how we can take them to a transformation that then helps them to overcome that problem. And I don't like to be clear, I don't know that most accounting firms think about what they do this way. And if you want a signal, ooh, this one's going to hurt, Buckle up. If you want a signal, look no further than your website. Is your website a big list of all the services that you do? If this is how we talk about what we do, then when I can do that, or can do 80% of it or 90% of it, then how do we communicate our value? Right? So we got to sort of escape this menu based, we're going to do xyz, buy some hours from my accounting staff sort of approach and instead keep super plugged into the problems of the people that we want to serve and continue solving them, even if the way that we solve them and what those problems are continue to evolve. Let's not forget we're talking about how our business has changed with AI. How are our clients, businesses navigating AI? That's arguably a huge opportunity right now is to help them through that, that transition. I mean accountants were certainly central to the cloud technology adoption and cloud accounting and having sort of a cloud based back office accounts played a huge role in all that stuff. For some reason we always get stuck with these sort of big tipping point sort of changes and we, we hold small businesses hand through it. I think AI is going to be no different in all likelihood. But it boils down to what we communicate we are in the business of doing. We did an episode the other day on, on kind of talking through people's websites. I think the most important thing to put on a website is a really painful pain that somebody's experiencing right now and framing ourselves as a solution to that pain. Is AI going to take away all of an entrepreneur's pains? Sure seems unlikely. And as long as we can keep grounded in that and plugged into to their journey and how we can support a specific type of entrepreneurial journey, it seems, seems unlikely accounting firms are going to be in trouble anytime soon. That being said, the one thing I always try to season in here is there's also a timeline where like that future world is just maybe a little bit better of a world to live in where we're working a little bit less. Like let's not close the door on that possibility. I got little kids and I don't want to root against AI and end up with a bunch of kids that are living in this like work obsessed sort of workaholic work, a crazy number of hours sort of culture. Like I wouldn't mind it if that wasn't how people were wired when my kids got older. So I also don't want to like root against a better future. I mean there's many aspects of this AI timeline that are very problematic. But I also, I don't want to cling to like no, I don't want how it is to change because we can, we can certainly make improvements on, on the way things are right now. The amount of time we spend keying things into software still, the amount organizing we do of documents of PDFs annotating just stupid stuff like that. I hope five years from now we look back, we're like, yeah, no, we don't really miss that. We're actually totally fine that all that stuff went away. I don't know what's your take on this. Drop that down in the comments I mentioned. I'm going to be in Salt Lake City next week. I think we'll have announced it by the time this is this has gone out. Jason on firms.com/events if you're in that neck of the woods, come out and hang out. It's going to be just like a dinner, happy hour sort of thing, I think. Whole bunch of live events we're about to announce. I'm very excited for that. But gang, thanks for coming and hanging out. I'll catch you in the next one.