Brian Tankersley, CPA, and Randy Johnston, discuss the recently held CPA thought leader and think tank symposium, which included discussions of upcoming trends and technologies critical for accounting and tax professionals. Watch the video, listen to the audio, or read the transcript.
Or use the below podcast player to listen:
===
Transcript (Note: There may be typos due to automated transcription errors.)
SPEAKERS
Brian F. Tankersley, CPA.CITP, CGMA 00:00
Welcome to the accounting Technology Lab sponsored by CPA practice advisor. With your hosts, Randy Johnston and Brian Tankersley,
Randy Johnston 00:10
welcome to Technology Lab. I’m Randy Johnston with co host Brian Tankersley, we’re coming to you live from Las Vegas, Nevada for from AICPA engage. But what we want to talk to you about in this episode is the CPA practice advisors, accounting leaders, Think Tank symposium, which was held on Sunday, June 8. It was an all day event that brought together the 40 under 40 winners that could attend, as well as other thought leaders in the profession, and the discussions were well thought out and well structured, but we thought some of the content could be helpful to you. Now we’re following Chathams rules, as Gail Perry asked us to do nothing that was said in the event can be used without permission to attribute. So we’re being very conscious that those discussions were kind of like behind closed doors discussion, and we’re kind of opening the door a little bit for you to peek in. And the insights that I learned, and I think that Brian learned, were pretty, pretty good. So we’re going to start on this session with the AI topics that were covered, because AI continues to be hot. But the people that were in the room probably pushed AI even further with agentic AI and the discontinuance of applications. I mean, it got pretty strong. So Brian, what would you start to call out for our listeners in terms of real key topics?
Brian F. Tankersley, CPA.CITP, CGMA 01:46
Well, I think generally with with AI in the profession. You know, the one of the things that’s interesting is there are a number of practitioners in the 40 under 40 that really are thinking differently about how AI affects their practice. You know, just about everybody has a story about how they’re using it to summarize large amounts of data, to to extract certain pieces of data from things. I’ve got to caution everybody that there is a significant privacy concern associated with it. You know, we don’t want your AI, your your confidential data, your clients confidential data to be used in training other people’s AI models and or to be embedded in those models. So we have to be very careful with that. But it was a it was a very interesting, interesting day, and and I will tell you that I tried very hard to match make between two of the 40 under 40, the two people that were really outstanding practitioners, I’m not going to name them under Chathams house rules, but I told them that I thought they needed to get out there and create a layer of AI on top of the application software that They use so that, so that they could, they could, they could be almost, almost application independent. But it was a it was an interest, you know, so so that way, if they wanted to plug their own plug attack software, the AI would just do the talking to it, almost like we do with Enterprise Architecture in Fortune, 500 companies with SAP at Salesforce and all the other applications they use. But it was, it was interesting that we had a lot of discussion about how AI was going to affect the profession, and my assertion was that I’ve had about 10 things that were going to eat my job since, since the 90s. They included adding machines and accounting machines, mainframes, personal computers, client, server, the web, cloud, big data outsourcing. And now a however, one of our one Randy and my friends, that that was not part of the of the group, actually, we had an interesting discussion yesterday where he said that AI is different in that it will actually do things for you. So it is something you can instruct and get to do things. And he made the argument that it was, it was different than than some of these other technologies that we’ve had in the past. You know, it remains to be seen how it’s all going to work out, but there’s been a lot of great discussion about what AI is going to do for us and what it’s not going to do for us. You know, I actually saw, actually saw Jeff Gramlich from validist, and spoke with him a little bit. They have an AI engine that will ingest pretty much any trial balance and put it into their platform. You know, those are the kinds of innovations that are that we’re seeing today in these, you know, at the conference, and we discussed a lot of those things. But again, the the we really didn’t reach a conclusion as to where it’s going to land up.
Randy Johnston 04:53
So I appreciate your kind of observations on that. And likewise, you pulled out the conversation we had. Yesterday. But, you know, just a few summary items that I would say agentic AI is big. That’s where everybody believes it’s going to go. Number two, that AI will be embedded into products that we use, rather than the general LLM platforms that we’ve been using, the chat, GPT and so forth, they’ll still get used and maybe we build our agents on those. But it is also, number three, clear that about every task that we do today, we can pump through the AI engines and they can get better over time. But number four, even when they write content, they don’t do a very good job of catching the tone and style. So therefore, many in the group cited that they’d seen AI used by people that they knew, and they knew that that person had not written it because they knew the style. Now maybe that gets better over time and maybe gets more muted over time, but I thought that was an interesting one. Another attendee said, All apps are dead. There’s no reason for any more apps. And that was rather interesting in terms of a big picture strategy, but also that APIs were no longer going to be needed. Now, I think that was a little misdirected, but I understood what they were trying to get to because they thought that the AIS could just go inside the systems and grab the data, not really knowing the mechanics behind it. But bottom line is, I get the sentiment, and then next that most of the applications would be very data centric. And one of the concerns and questions was, how do we select a foundational skill for ourselves and for our future team members, people that we’re hiring or people that are in school? What’s going to stick in the long term? Because things are changing so fast, perhaps iterating every few months, or even slowest, every year, what skills are really going to be needed? Now I haven’t even summarized all the thoughts that these sessions generated for me, Brian, that’s where you and I are lucky to be able to sit in listening to thoughtful people thinking about their future in the future of the profession when it comes to these topics.
Brian F. Tankersley, CPA.CITP, CGMA 07:21
But can we agree that calling the cable company still sucks? Okay, so as much as we talk about agentic AI, there is still a need to talk to a human being in many cases, and when you and so this, this inter human communication, I don’t think AI is going to going to ever completely replace that, because I, I have never felt heard or understood, really, at a at a molecular level when I talk to any kind of AV, IVR system, you know, like you have at the cable company or the phone company or cellular carrier or somebody like that, you know. And, and I, I think, I think the thing that we all have to remember is that accounting profession is about trust, and trust is something that we ascribe to humans, but we don’t ascribe to ascribe it to systems in the same way, nor do I think we should ascribe it to systems in the same way, because systems can be much easier, much much easierly. It’s much easier to redirect them. One assertion I would say about AI and the accounting profession that I made and I, and I’m, I’m a firm believer in this, and I want to just throw it out here is that I think the big changes that affect the accounting profession are going to really come when the counter parties on the government side and the user side get AI at the level, at the level that we’re talking about today, and it’s going to drive changes in the kind of information provided, and the detail of the information provided, as well as it’s going to make it much harder for people to hide behind the the anonymity of a 1040 or an 1120 or 1120 s, you know the it’s going to be much easier to pick out the exceptions And then to then to subjectively evaluate those. And I think realistically, the government side of that, getting getting fixed, is going to be much more interesting. You know, if you look at Europe, and you look at you look at what they do with value added tax today, what they do in Canada with value added tax, with their GST and HST, those, it’s it’s much harder to cheat in those in those systems, and so it’s a, I think there is a big reckoning coming as a result of AI. And, you know, we kind of discussed that. But, you know, again, the if we think about the, if we think about the, I guess the foundation. Of accounting, you know, we, you know, the I think the accounting professionals are going to get there one way or another. They’re going to get there, whether it’s by client demand or whatever. But I think the government side getting there is where the most change is going to come from. And so it’s going to be a very interesting ride to see how a Elon less Doge continues on.
Randy Johnston 10:25
Yeah, well, you know, we did talk about the governments like France and Australia preparing returns, and we talked about how the tax code was so Byzantine that it was hard to get real answers, as opposed to accounting software, where debits and credits really resulted in Financials, where that you could publish and why debits and credits were hard enough to pump into AI, with systems like digits AI and many of the others that we saw introduced here, the AICPA engage. But the fact the matter is, those tax codes are all still pretty soft and and you know this one I can attribute to you, Brian, because multiple times you’ve said people will make mistakes, and they need accountants to help them overcome some of those errors that they make along the way. And needing this human to talk to, I think, is where the value is going to be, because if more of the functions that we’re doing, plan accounting services or audit or tax or wealth management or business valuation become automated, we’re still going to have to have that value judgment, And I just don’t want advisory services to become commoditized, which is possible. But you know, when you were talking earlier, too, one of the attendees said something to me that is, you know, I’m still just shaking my head over I’m not saying it’s wrong, right, because that’s what you’re doing, is you’re learning here. But they said that many of the younger team members that they have are using chat GPT as a cycle analysis tool. So they’re putting their problems in and letting chat GPT tell them what to do. And it’s like, oh, my that is well beyond, yeah, that’s beyond what I thought people would do. And when you look at what’s happened with, you know, the most recent model of Claude, where, you know, it started sending threatening emails for termination when there was, you know, asked to shut down. I mean that those types of things, these AI models still are not well understood and but I think the main takeaway was, even though there was a lot of buzz around Gen AI, there has not been Very many fiscal models showing the profitability of using AI. They see the soft benefits, but the business cases are still pretty slim for illustration, which takes us back to AI being built in products and possibly a lot of hype around these. You know, my estimate of the 200 vendors on the show floor at AICPA engaged this year is that at least half 100 of them are talking about something that they’ve got that’s AI in their models. And I just don’t think that’s necessarily true
Brian F. Tankersley, CPA.CITP, CGMA 13:34
well. And the other thing we have to remember here is that everything we do in accounting is subject to a cost benefit analysis. You know, we don’t, we don’t spend $5 chasing a nickel. And the thing we have to remember with AI is that the processing cycles that are required to make that work are are much more expensive than, say, just traditional processing cycles and so, you know, takes me back to Fred around you at at their chief evangelist, and his his idea that that different models and different different technologies are important as you’re trying to optimize things. Because AI is not going to solve everything. Certainly, it can, just like you could have, you could have a concierge solve every problem you have, and it would cost a stupid amount of money, okay? But you know, when we’re trying to optimize for these things, I think it’s critical that we understand that AI is but one tool in our toolbox. We have to queue and create this, this system that we have that’s going to give us accurate information. Now, remember, we don’t want to spend more on creating that information than it’s worth. Nor do we want to you know, and part of the value of it is in its current. Currency. So if, if we have more current information there, there is much more utility to that, as long as it’s accurate. And so I think AI can help with that. But, you know, it’s, it’s, we’re at the beginning of a very long road down this thing of AI, you know, Intuit has actually announced that they have some new features that are coming out in in QuickBooks Online here in the fairly near future, related to AI and, you know, I and let me just, let me just go out here and venture a guess that those are also going to be accompanied with new, new increases in price on QuickBooks Online to just, you know, to justify those AI things. But it’s, I think it’s, I think it’s important that we understand here that we’re at the beginning of a very long journey down, down this technology and things are going to evolve. You know, when I started in the accounting profession and I walked in the front door of Cooper’s library in August of 1992 one of the partners, Tom Jennings, great guy, used to sit and smoke at his desk every day, you know. And he would go through two packs minimum of cigarettes, sitting there having conference calls with people. He was actually a consulting partner, where he was who people went to when they had gap issues that they didn’t know how to fix, almost like a referee or a judge. And one thing I’ll say here is that I think the human element that he taught, that he taught me, is is is critical in this, because there’s still a lot of subjectivity in accounting, you know, and again, especially with respect to tax, two plus two is not always for in tax. Sometimes it’s five and sometimes it’s three, depending on the loopholes and the exceptions and things like that. I mean, look at the credits we have on electronic electric vehicles. You know, you have a $50,000 vehicle with a $9,500 tax credit on it, or whatever the number is, and now it’s a $40,500 vehicle. And so it’s, I think, I think it’s important to understand here that AI is going to be helpful, but the humans still have to make decisions about risk and about about the validity of information and about the credibility of information, and those humans themselves also have credibility that they’re trying to protect. And so it’s a the equation when we start plugging AI into this, AI is certainly a tool, just like caffeine is a tool when you’re trying to stay up and work more hours. But it’s not the ultimate solution. Okay, it’s just part of the process that we have to price in and figure out and and see what things are going to evolve into.
Randy Johnston 17:54
Yeah. So you know, just thinking about the Intuit AI model, they’re just now coming out, saying it’s been under development for six years and just talking about more of the capabilities of their model. And of course, we’re watching the and have, in a Future Tech Lab update for you some of the apple announcements from the Worldwide Developers Conference with the evolution into their, you know, 26 naming system, but all of their incorporations of Apple intelligence and of chat GPT directed in some of their platforms, so things
Brian F. Tankersley, CPA.CITP, CGMA 18:30
let me inject in here, that app that that into it does not have any way for you to opt out of Your data being used to train this AI model, okay? And that is a critical issue that that we have to address and think about here as a profession, cause you know we don’t need I think that you know, our Code of Professional Conduct says that unless we’re authorized to do so, we cannot reveal confidential client information, and I think revealing it too into it, and putting it into their systems potentially is a is a revealing of that information. So we have to wreck them with this whole privacy situation as a profession and and understand that this is not something that comes at a zero cost. You know, you and I had a number of discussions yesterday about about privacy being dead, and how maybe we just need to hold our nose and go with it on some of these things. But I think we have to at least say out loud in this discussion that figuring out the privacy is really important because it’s become clear that Washington’s not going to do it for us, and we are the We Are the bulwark out there as a profession protecting the privacy of our client information. Now we can go along with what the pool kids in DC. Say, or we can have a standard that’s higher than that, that protects our clients, and we have to make decisions surrounding that, or Silicon Valley is going to make it for us. And you and I both know how that’s going to come
Randy Johnston 20:13
out. We do. So you know my two takeaways, and you really got to one of the two was, what do we do about this privacy security, because holding the nose is probably, you know, I just don’t want to give up on it yet. But, you know, I’m beginning to see we don’t have a lot of options on this because of the devices we use our phones and other things. So we’ll talk more about that in a future technology lab, because that’s a current troubling thing, and I want to be more thoughtful about it before we talk to you about it. And But the second thing is, it is clear that getting an initiative to create agents in your firm is actually a good use of time, and likewise, in a future technology lab, we’re gonna have to talk to you about the creation of agents, because we think there will be platforms that do it for us. Maybe these will be inside tools. But the fact the matter is, the agents that you create will help you and your team for your clients very specifically, and that’s where we think the value of this future AI is going. So Brian, any other closing thoughts?
Brian F. Tankersley, CPA.CITP, CGMA 21:23
We live in interesting times, you know, I think about the changes I’ve seen over my 33 year career thus far. And you know, where, where? Literally, I didn’t have a work computer for the first year. So I was a Coopers Lybrand, and they were ahead of everybody else with technology. You know, in my market anyway, and and so I think it’s important to acknowledge here that change is always part of accounting there, there people will always say that accounting is going to disappear, and that’s not the case. Okay? You know, we still have police. We still have law enforcement. They still have to investigate, okay, and and we are the investigators of business that figure out what’s right and wrong and sort out the winners and the losers, and figure out and help people navigate and figure out how to, how to, how to steer those businesses forward, and that role is not going away. Okay, so I think we all need to calm down a little bit and get and get ideas of how we’re going to do things. I would suggest that you start with ministerial tasks with AI. So if you have the Microsoft 365 copilot, get it to summarize your inbound emails when you’re behind and tell you what to do, get it to write responses to messages for you where you need things done. You know, I used, I used generative AI a while back to get to get to get a thank you note written for my son, because he spent three hours trying to figure out how to write one, and he was clearly stuck, and I had had chat GPT write one for him. I actually had to write five, and I picked out the best of the five, and I put it in front of him, and I said, Write this down, and don’t, don’t say a don’t say anything different. This is a perfect this is a perfect item. And I think we have to think about that. Is that AI is going to be invaluable when we get stuck. It’s going to be invaluable that ministerial tasks that are that are not the thing that we’re paid to do, but they are the necessary tasks that we do as part of it. And so I think it’s a I think it’s a matter of us reckoning with this new tool and figuring out how to use it. It is going to neither create nor destroy the entire accounting profession, but it is going to have an impact on it. And I’m I’m quite excited to be living in this age.
Randy Johnston 23:53
Well, interesting times are your words and others that I’ve heard, uncertainty and unsettled at the risk of ending on what sounds like a negative note, but there is confusion about what lies ahead, and that’s okay. We’re going to help you figure it out here in another technology lab. Good day.
Brian F. Tankersley, CPA.CITP, CGMA 24:17 Thank you for sharing your time with us. We’ll be back next Saturday with a new episode of the technology lab from CPA practice advisor. Have a great week.
Thanks for reading CPA Practice Advisor!
Subscribe Already registered? Log In
Need more information? Read the FAQs