Brian Tankersley, CPA, and Randy Johnston, visit the tradeshow floor of the 2025 Scaling New Heights conference, one of the largest educational and networking events for accounting and tax professionals. Watch the video, listen to the audio, or read the transcript. [See Part 1.]
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, Randy Johnston
Brian F. Tankersley, CPA.CITP, CGMA 00:00
Randy, welcome to the accounting Technology Lab sponsored by CPA practice advisor. With your hosts, Randy Johnston and Brian Tankersley, well,
Randy Johnston 00:11
welcome to the accounting Technology Lab. I have the pleasure being face to face with my co host, Brian
Brian F. Tankersley, CPA.CITP, CGMA 00:16
Tankersley, and since we’re in the lab, we’re wearing our lab coats. As a matter
Randy Johnston 00:21
of fact, Brian said, We need lab coats. And I said, Brian, that is such a doggone good idea. And another good idea he had was to shoot a couple of episodes right here in the main conference hall for scaling new heights in Orlando. So we’re at the Marriott World Conference Center. Joe will continue to have this event at this property through 2029, but I’d like to turn your attention to just a moment, to AI. One of the major themes of this particular event was artificial intelligence. Many, many sessions were focused around AI. I had the privilege of teaching an introductory session on artificial intelligence. And I believe that I’ve sat in approximately 10 others myself. And you know, as I’ve watched the AI, I’m just going to start with this and then ask for your color, Brian. But it seems to me, first people are looking at how the Gen AI tools work, the large language models. Chatgpt is very popular with this crowd because the way they’ve been oriented, but we do know that many spoke up and said, Look, we’re using Claude, we’re using perplexity, we’re using copilot, 365, we’re using deep seq, we’re using Gemini, right? I mean, we heard all of these on the floor, so the large language models have a play. But the other thing that became clear is AI is going into the tools, and I had the privilege of being present for the first digits sponsored AI native accounting podcast episode, which was fun to
Brian F. Tankersley, CPA.CITP, CGMA 01:53
watch, be produced. Isn’t our friend Casey Johnson hosting, and she is, and that’s that
Randy Johnston 01:59
she’s a stellar person and knows the content cold. But the key is that we’re seeing AI be put into products, and that works is where we think a lot of this is going. I don’t think you’re going to have to actually buy a lot of AI separately. I think what’s going to happen is you’re going to buy your tools or get your upgrades of your platforms, and AI is going to appear there. So this this kind of multi prong fork, and then the third fork is the agentic AI. We can see that the ability to code certain tasks so they can be done better for you is coming. But I did see misstatements, misrepresentations on agentic AI by various presenters, and there’s clearly differences of opinion, because people have learned what they know about AI, and it’s like that old elephant story. They’ve got one leg, and they describe the elephant by the one leg they have, rather than maybe a bigger picture. I’m not sure I’ve got it right either, but I do understand this three these three prongs are pretty clear coming at us. And from your perspective, don’t worry about missing out, but you do need to be taking action pretty soon. So Brian, I’ve gone on a long rant. Well,
Brian F. Tankersley, CPA.CITP, CGMA 03:10
I think everybody wants to learn about AI, okay, but the problem is, none of us know what we don’t know, and we’re all learning from each other. And there’s a, you know, there’s the part of the other problem is that when it’s hard, one of the hardest things about any kind of technical training is to level, set it so that you have the right level student in the class. Because it only takes one person that thinks the CD drive is a cup holder to mess up the thing for everybody, because then they ask all the questions, and you can’t get anything covered. On the other hand, if you’re too if you’re too simple, you bore people to death, and they walk out of the room. You know, the hard thing about it is that we have a we have a lot of different skills here. We have a lot of different capabilities, and we have a lot of people that perform different functions in organizations. So the things the marketing people need for search engine optimization and content creation and, you know, CRM automation and that kind of stuff is important, okay, but then we’ve got another set of things that are that are really matching up where we’re, you know, with digits and and, you know, your your other, your other, you know, AI based general ledger systems. And that’s there, you know. The, the thing is, I think everybody had an AI story, but it’s it really, you know, it’s everything and nothing at the same time, you know, there’s, it means so many different things to so many different people. And it’s, it’s not, I think, I think, in general, it is changing so rapidly as we learn new things, and as you learn new things, and as scientists invent new things, that it’s you. You know, it’s just hard to nail AI down right now, just because there’s so much to know.
Randy Johnston 05:05
Yeah, now listening to you explain that I had breakfast today with a couple of attendees, and it was fascinating to talk to them, because they explained what they were doing at AI, and that’s the main reason they were at the conference. And I asked each attendee,
Brian F. Tankersley, CPA.CITP, CGMA 05:21
and that’s, and that’s also what we what I got it engaged. That’s the same thing. They said,
Randy Johnston 05:27
Yeah, makes beautiful sense. This particular couple of ladies explained what they were doing. I said, Well, that’s very interesting. What platform you’re doing it on chat. GPT, okay, understand what version are using, $20 version. You do realize that that data that you’re putting in the platform is being compromised? No and so when we got off on a security discussion, we had an additional discussion around. And you do know, since you’re a QuickBooks Online Shop, that there is no opt out right now for the AI training with QuickBooks Online,
Brian F. Tankersley, CPA.CITP, CGMA 06:04
and we like your confidential client data, including payroll,
Randy Johnston 06:07
including payroll. And they were like, No, didn’t know that either. But the reason I bring that story up is I did have the privilege of teaching both Tech Update with some AI and quantum content, as well as the intro AI course. And in both sessions, I commented about the AI training. And my position right or wrong is you should have the ability to opt out, and I just happen to disagree with that. And in both sessions, I wouldn’t call it thunderous applause, but it was notable applause.
Brian F. Tankersley, CPA.CITP, CGMA 06:39
It was very notable, because I was, I was in there. I was serving his fan of white, taking the microphones around for questions, so people can be on the PA and the the thing, the thing I noted was, was how people are just repelled and disgusted when they find out that they can’t opt out of this thing. And, you know, Intuit is not the only one, I’m sure, that has privacy sins that will be revealed. But the problem for the practitioners is that the even though there are disclosures in the privacy policies and the terms and conditions and all those things, they are generally 10s of 1000s of words, and they are written at a graduate student level, and it’s just too so much legal ease that it’s very difficult for people to glean what they actually mean, and everything. And then you get into the middle of those things, and there’s sub processors that can, you know, you could say that we’re doing everything right, and all data is being kept holy, and then we have the sub processor, Palantir, and they’re doing evil things with it, but you’re gonna have to read their privacy policy to figure it out. So it’s fairly opaque, even when people are doing the right thing by the privacy policy, because of the way the privacy policies are structured.
Randy Johnston 07:59
So on a little different tack. It was fascinating to watch Ron Baker and our friends from anchor who came across from Israel for some of the presentations. And they’re incorporating a proposal tool in their platform, which is AI powered. And they had demonstrations of that, and it somewhat supports ed classes and Ron Baker’s new effort to threshold. Threshold is the name of their company, but they’re trying not to use the word transformation, but it truly is a new level of service that they believe is needed. And they’re trying to align the proposal tool, if you will, with that. Well, that was kind of a fascinating use of AI, but I watched what they were doing, I looked at the prompts that were being done, and I realized, oh yeah, that’s fundamentally prompts. And without trying to steal intellectual property, I basically typed up the same prompts, and I used them in my AI course, and I got roughly the same results that we saw with the anchor threshold. So what I concluded, Brian, and this is big ask of both of us, but I concluded, I’m going to have to look at six or seven practice areas for public practice, where we create prompts as examples, very specific in that area. And the same thing for our industry.
Brian F. Tankersley, CPA.CITP, CGMA 09:26
And I think we need to put together the technology lab, AI prompt cookbook, and we need to find some we need to find some vendor to sponsor it. If you’re reading this, you know Barry? What’s Barry’s last name? The our publisher, Barry kit can can get you set up if you’re interested. Barry Strobel, yeah, but, but it’s a, you know, it seems like one of the functions that I see as an educator is that I it’s clear that there is a thirst for a lot of things on AI, and it’s a clear that there’s a third. For things practical, other than creating, you know, mean pictures or deep fakes or other things like that. And it’s, you know, the the knowledge is there, but it’s not getting transmitted for some reason.
Randy Johnston 10:14
And that’s another key word. Maybe there’s a couple other keywords that I was pleased to hear practical because we were told many times your practical strategies make sense to us and your genuine concern. And let’s face it, you and I, the reason we do the accounting technology lab here is we want you to have the best information the way we understand it today, and we try not to be wrong, but some of these observations, there’s a lot of people out there that are real. We call it smoking dope,
Brian F. Tankersley, CPA.CITP, CGMA 10:48
but, but I have to tell you that that I am absolutely passionate about that, because when I took my first job in accounting technology as the as I was an audit manager, and I was also the tech manager for a mid size accounting firm, and I remember the times when I had problems that I didn’t know how to fix, and there was nobody to call. There’s no support system. There was no you know, there was you were just, you just had to improvise. And the thing I will say is that, you know, you and I both are very passionate about, about helping people figure out how to get out of ditch, you know. And I will just say that it’s I feel that that’s almost like my calling, you know. I feel like Kane I’m walking the earth and solving accounting tech problems, but it’s the thing, the thing I like about it is that the thing, the thing I think is critical, is that I think you our listeners, have to know that, you know, it’s if you get stuck, don’t hesitate to reach out to one of us. Okay? Because we’ll, we’ll try to point you in the right direction. Now we’re not going to give you five hours of free time. Hours of free time, but if you have a five minute question, five minute, five minute answer is easy, you know. And again, the the thing to know here is that whenever you’re having problems, keep good notes. Keep you know, if you’re going to correspond with us about it, send us the the ticket numbers you’re dealing with, the vendors you’re dealing with, which actions you’ve taken and everything else, but I think it’s, you know, I remember how miserable it was to have to look at those partners and say, I don’t know, but I will get a plan. And I remember how uncomfortable it was for them and me, yeah, but it was the truth.
Randy Johnston 12:37
And it turns out, one of my other favorite comments from the event, and it happened at engage as well, was, I have question, are you going to use AI to make up an answer? Are you going to actually answer with real facts? And it’s like, that’s kind of an interesting take, because, as an example today, simple request, I’m going to buy new monitors. Can I ask you what monitors to buy. And I said, Absolutely, I’ve researched that, and I think I’ve got good, solid answers, and
Brian F. Tankersley, CPA.CITP, CGMA 13:05
we’ll tell you why, and we’ll tell you why we didn’t pick the others and what the objectives were, and what the what the issues we ran into. And sometimes it’s an issue with Citrix, or it’s an issue with with, you know, with, you know, warranty claims in the past with that vendor or
Randy Johnston 13:21
something, yeah, but it was really fascinating because the business partner said, Oh, he’ll just use AI. He’ll tell you answer it. And it’s like, no, I’m going to actually do work the old fashioned way. And I may make some mistakes, but you know, that’s what we’re trying to do here with these AI platforms too.
Brian F. Tankersley, CPA.CITP, CGMA 13:37
Well, remember, the AI is only as good as the data that was used to train it. Okay? And the problem is, everything you read on the internet is true, isn’t it? Yeah, exactly. And so the problem is that when you’re creating content with AI and you’re just using the AI, it’s like the world’s biggest sausage factory. You will never, ever, ever have a filet mignon or Chateaubriand. Okay? That will never, ever come out of that sausage maker. Okay? Now, it will give you a lot of sausage, and you can feed a lot of people, but it’s not going to give you the, the refined things, and it’s not going to give you the it’s not going to be well thought out. It’s going to be a whole bunch of disjunct things glued together, almost like some kind, almost like a mosaic of tile. We got a bunch of broken tiles. We’ll just throw them together, put some, put some, put some grout in it and call
Randy Johnston 14:34
it time. Wow. Well, I’m listening to you on this because it’s it’s pushing supper time for us. But as it turns out, I was just thinking about the corned beef and the salt. And, you know, I learned lots of different things from Brian when it comes to, yeah,
Brian F. Tankersley, CPA.CITP, CGMA 14:46
so, so, so I have bees, I have dogs, and now I’m trying to, I’m trying to make the world better with corned beef, because I live in a third tier city, and there is no proper, no proper kosher deli where I live. South. And every time I go to the second avenue Deli in New York, or, you know, the Katz’s, or someplace like that, it’s it’s it. It’s a different meat there than it is in Knoxville, Tennessee and flyover. So
Randy Johnston 15:13
he’s trying to teach me, you need this salt, you need that. And it’s like, Okay, I’m getting old. It doesn’t sink in quite as well as you used to, but it’s,
Brian F. Tankersley, CPA.CITP, CGMA 15:21
well, you know, not everybody gets excited about taking, taking a bunch, taking a brisket, put it in it, putting it in a five gallon bucket with a bunch of weird, funky pink salts, and then, then cooking it later, and putting it over cabbage. I mean, it’s, it’s, you know, I try to do everything I can at 110 110% and sometimes that ends up with some weird
Randy Johnston 15:43
results. Well, speaking of weird results, AI can get you weird results. And one of the things that we encourage you to do is understand how to do prompts properly. And so again, not trying to promote our k2 courses, but one of the things I think we’ve learned along the way is how to do proper prompting with our courses. And I did see some other prompting methodologies here that were used, and I’m going to experiment with those a little bit, and Brian and I will work that out. But when we suggest a certain AI technique works properly, you can pretty much plan on it, but still remember, a lot of these platforms have hallucinations. A lot of the products have hallucinations. And as I watched people like digits AI, accounting and puzzle trying to work out how to have accurate results in their platforms, I’ve heard that from the big players too, the Wolters, Kluwer and Thomson Reuters accuracy with AI. It’s a bugger of a technical problem. It is
Brian F. Tankersley, CPA.CITP, CGMA 16:42
because, you know, the in digits announcement, they quoted some statistics, and one of the things they said was that the AI got things right 97% of the time, or 98% of the time. Well, who decides what’s right, you know? And so it’s a, you know, we’re in the early days of AI. Even though it’s been going on for decades, we are still in the early days of AI. I remember in 2010 when QuickBooks Online was the new version of QuickBooks, even though had been around for 10 years already, at that point, it’s, I think we have to acknowledge that there’s a lot we don’t know. There’s a things are changing so rapidly that even if we wrote the perfect AI course today, recorded it today, and delivered it tomorrow, we’d still miss something. This is like walking in. This is like when you go from high school to college and you go to biology, and you thought you really knew biology because you took AP biology, and your head just cooks off. Yeah, because it’s it’s so there’s so many more layers that you never knew. Well,
Randy Johnston 17:46
one of the AI attendees asked me the question, and I said, look, the old methodology I used, which I cannot use anymore, was usually we would teach new concepts over a period of time. We’d teach the what in year one, the why in year two, and the how in year three. Can’t do that anymore with AI. I’m not even sure we can do that three months apart.
Brian F. Tankersley, CPA.CITP, CGMA 18:11
I think you have to just this is what we know, and this is the what, why, how, and
Randy Johnston 18:16
it all comes at once. You just don’t have the time to really cogitate on well
Brian F. Tankersley, CPA.CITP, CGMA 18:22
and understand that that you’re not the only one drinking from a fire hose on AI. We are drinking from a fire hose on AI. Now, you know, again, I’m probably spending over 150 a month in monthly AI subscriptions between Claude and chat, GPT and and, you know, co pilot and and all the others out there. I mean, it’s, I’m literally over, over 150 a month. And, you know, I, I don’t want to, I’m, I’m certainly not one for profit, get spending. But the problem is, I’m afraid I’m gonna miss something, and so I want to have the other engines to try it out. But again, many of you are doing the same thing. And I would just encourage you to, you know, again, look at the privacy piece in particular, because the data that our profession deals with is so sensitive and so so private that you know, again, even even if the likelihood that you know that one per one thing is going to be revealed if it’s one in a million. Well, if you share all of your client data, including payroll information out there, that the law of large numbers will eventually catch up with you,
Randy Johnston 19:31
and especially with these products tracking your IP addresses in your home and your offices, they can tie stuff together, and at the risk of possibly one more sad note in this area, Brian. And this is actually a pretty important sad note. I did listen to presenters who were still claiming fake AI. So as buyers, you’re gonna have to watch this very carefully, because, you know, people would say this is AI, and I’m thinking this. It’s not AI that is machine learning, or that is, you know, and it was clear,
Brian F. Tankersley, CPA.CITP, CGMA 20:05
well, we’re not saying they’re liars, but we’re saying that our understanding of the meaning of that term is quite different than the meaning that they they understand it. To me,
Randy Johnston 20:16
that’s right. And some to some degrees, in some cases, they have marketing reasons to want you to think it’s AI, but, and by the way, it’s perfectly good. Many of you are accountants. You know a lot more about accounting than I do in almost every aspect. But the thing that’s true is, I think I’ve got a pretty good handle on what’s AI and what’s not, but I’ve watched people say this is AI, and it’s like, that is so not AI. So just be cautious. But hey, if it’s an innovative product that you need pretty good with that. So
Brian F. Tankersley, CPA.CITP, CGMA 20:49
yeah, and I’ll say the same thing. I always say, you know, it doesn’t work until it works with your data. Okay, so, so don’t you know, when you get a demo, get a version of the software, throw some of your data in it, see if it works the same way with with it, with your real data in real life. And I know that takes some extra time, but remember, when you pick a software platform where you get married, okay? And you know, when you get married to something, it costs to get out. And so, you know, if you don’t know my world, it’s not either. I mean, you’ve been married almost 50 years. I’ve been married 32 now, you know, but But seriously, it’s, it’s a long term partnership, okay? And you want to know what you’re getting into before you, before you pull the trigger, and so
Speaker 1 21:32
probably why there’s so many prenups nowadays. That’s right. That’s right, you know. So Wow.
Randy Johnston 21:37
Well, here we’re coming to you live from scaling new heights in Orlando. We hope we give you an insight on the conference overall in some of the sessions, as well as AI in this session, and also from the show floor walkthrough. So Brian and I always appreciate the opportunity to share what we’re learning with you, and this one’s pretty close to real time, and we hope that you get some benefit from our time here.
Brian F. Tankersley, CPA.CITP, CGMA 22:01
Thank you. Every day I get to do this instead of working a cubicle for the man, is a win. So I’m so glad to be here and do this with y’all and you know, again, it’s a it’s an exciting time in the
Randy Johnston 22:15
profession. Always a pleasure to have you with us in the accounting Technology Lab. We’ll see you again soon. Good day.
Brian F. Tankersley, CPA.CITP, CGMA 22:22 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.
== END ==
Thanks for reading CPA Practice Advisor!
Subscribe Already registered? Log In
Need more information? Read the FAQs