The Accounting Podcast

Attention: This is a machine-generated transcript. As such, there may be spelling, grammar, and accuracy errors throughout. Thank you for your understanding!

Blake Oliver: [00:00:04] So I can't wait until the I will do most of our bookkeeping and just show me the exceptions. That's going to be really awesome when that happens, and that's going to happen in the next few years.

David Leary: [00:00:14] Coming to you weekly from the OnPay Recording Studio.

Blake Oliver: [00:00:19] Hello and welcome back to the show. You are listening to the number one podcast for accounting and tax professionals in the world, your weekly roundup of news and analysis in the profession. I'm Blake Oliver.

David Leary: [00:00:32] And I'm David Leary. I'm Blake. I'm so excited about this week's stories. We might have the craziest story I've seen ever on the history of the podcast. And I think I've said this before, but this one's really crazy. It's juicy. We have app companies rippling in deal, we have espionage, and we have it ties all the way back into, um, senators and Department of Treasury and Russia. So I'm really excited to talk about that story later.

Blake Oliver: [00:00:56] Rippling and deal payroll companies Air companies at each other's throats. It seems like the corporate espionage angle is fantastic. I have plenty of news about the IRS and of course, a little bit about Doge. I'm not going to let that one go. What is the impact of these IRS cuts? 7000 of those fired IRS workers, those probationary workers have been reinstated. And I want to talk about the impact of the potential cuts, whether or not they actually get reinstated, if the cuts go through. If 20 to 50% of the IRS actually gets cut, like Doge wants to do, what's going to happen to the tax gap? What's going to happen to tax collections? Will we lose more money than we save? There have been plenty of opinions out there, and I want to dig into some of the numbers to hopefully, uh, figure out what's really going to happen. Is this really going to make a difference? You've also got a story, David, about this 1 billion budget hole in LA where I used to live.

David Leary: [00:01:58] City having but billion dollar issues?

Blake Oliver: [00:02:00] Yep. Um, so security is going to do ID checks. I is getting better at longer tasks. We've got some data about that. How is that going to impact accounting and finance professionals. And if we have time, I would love to show an example of of doing some data analysis with Claude, which is really good at it. Now I'm starting to just dump spreadsheets of data into Claude for our own business and get these incredible charts and analysis visually and in written form. So, so much to talk about. We have to get to it.

David Leary: [00:02:35] Consulting stuff to the consulting gravy train of the big four might be coming to an end. And we talked about this a couple episodes ago, but we have to cover that as well.

Blake Oliver: [00:02:46] Yes, Accenture got its contracts, its government contracts cut by the Department of Government Efficiency, and that has sent their stock price because it's a publicly traded company plummeting about, I think, 17% the last time I checked. Not great for the consultants who rely on government money to fund them, to fund them. Welcome to our livestream viewers. We've got Gator NYC here. Good afternoon gentlemen. Good afternoon to you, Gator. Although it's morning here in sunny Arizona, we've got light em up here. Light em up. Made it to the live broadcast. Welcome, Sean Smith. Hello. You caught it live again. Thanks for joining us. And Steadfast Bookkeeping. My first time catching a live broadcast. I still steadfast.

David Leary: [00:03:30] Feel like we can't start until, um. And I'm forgetting his his handle. But till we get our coffees, we need.

Blake Oliver: [00:03:36] We need the coffee emojis.

David Leary: [00:03:37] Yeah.

Blake Oliver: [00:03:38] Where's the.

David Leary: [00:03:40] Handle? Who's handle is that? He always shows up.

Blake Oliver: [00:03:42] I'm blanking out right now.

David Leary: [00:03:43] I know I can't remember the handle.

Blake Oliver: [00:03:45] Well no worries. How about we thank our sponsors while we try to remember that?

David Leary: [00:03:48] So our sponsors I'd like to thank Rate Works and Bluevine this week for sponsoring the podcast. So please, when we read the ads, go to those links, support our sponsors and take advantage of these great offers and products.

Blake Oliver: [00:03:59] Uh, Patrick's got us. Gators got us. It's boring. Accountant. How can I forget? Uh, and vapor has the two coffee emojis. Light em up. Three coffees for us today. I was up at 5 a.m. today helping my son get out the door. He is going on a field trip with his fourth grade class to the Grand Canyon. They are getting on a bus. They got on a bus and they're driving up there, I think. I don't know. Are they there? They're there by now. They're probably there, and they're going to hang out at the Grand Canyon and learn all about rocks and geology. And it's a special trip.

David Leary: [00:04:33] I remember my kids did that. It's a special, special trip.

Blake Oliver: [00:04:36] Yeah, that's that's something that I hope doesn't get cut by the Department of Government Efficiency. The national parks are like one of the greatest services that the federal government provides. And we need to make sure that gets protected. And I say that now as an Arizonan and we need that. It's it's a great public service. And yeah, I'm really glad my son's getting to see it for the first time. Last year was my first time. Um, and it's fantastic. Okay, so we thanked our sponsors. Oh, one other thing I want to do before we get into the news is I want to talk about a recent interview I did on the earmark podcast, where you can get free continuing professional education credit. I interviewed Nicholas Boucher. He is one of LinkedIn's top finance thought leaders. Finance, not accounting. But there's a lot that we can learn from him. And in the interview, episode 82 of the earmark podcast, he demos how you can use tools like ChatGPT and Ta to do financial analysis. You can analyze finances. You can do all sorts of amazing things that you used to have to be an Excel master to do, or you couldn't even do an Excel you had to do in like specialized tools. And what's really neat is that he shows how you can do it one time in ChatGPT, and then you can take the code that ChatGPT generates the Python, and you can then save that and use it basically as an application to get the same result every time with your data. So that's one of the problems with AI is that every time you prompt because it's statistically generating the output, it might be different every time, but that doesn't work in accounting and finance.

David Leary: [00:06:22] So I've been thinking about this a lot. Not not because you're interview, but the whole industry has pushed all these accountants and bookkeepers. You got to do advisory, do advisory, do advisory. There's no money in the bookkeeping or the compliance. But as with AI, I probably think I would be better at the advisory stuff because when you when you interpret the full data, your interpretation Not the wrong word is. Wishy washy is not the correct word here. You have some freedom in your interpretation, and the way you write it up in one time for a client might be slightly different the next time, and that be okay with I. But the the bookkeeping and the numbers and the data entry has to be perfect. And so maybe I isn't going to handle the data entry stuff long term. And it might just handle the advisory stuff. Taking that off your plate. It might not it might do the data entry. It might do both, but.

Blake Oliver: [00:07:15] It might do both. I think it will definitely accelerate the advisory. It'll enable people who want to do advisory to do it a lot faster. But you're still going to have to know how to interpret the outputs. So you still need to know what you're doing. You need to understand financial ratios. You need to understand how to generate a business plan. You need to be able to do it. You're just not going to have to do the grunt work. And that's what Nicholas shows. So uh, go listen to that interview. It's fantastic. It's episode 87 and the the link is podcast earmark CPE comm slash 87. So just go to podcast dot earmark cp com. You can get a free continuing professional education credit by taking a five question multiple choice quiz on the earmark app. You can even listen to the episode in the earmark app. I should probably put that link into the chat and, uh, earn your CPE that way. And that's Nasba approved CPE earmark app. You can get it for this episode that you're listening to right now. In a few days after we livestream this episode, that'll be up on the app. You can go get continuing education credit. We are up to 117,000 CPE credits issued for listening to podcasts and watching live streams. Uh, so thanks everyone who has supported us at earmark. You can Subscribe for $150 a year to get unlimited CPE. But you can also, if you're on a budget, just earn one a week for free by listening to this show or many other great shows like. Oh My Fraud, federal tax updates, and the earmark podcast to just name a few. We've also got what tax Chats, the unofficial QuickBooks accountants podcast. That's a very popular one. Feel free to shout out a few more, David, if.

David Leary: [00:09:09] Unofficial Sage podcast is on there. Um, I'm going to go open up Earmark app because it's easier to see that in the app itself now.

Blake Oliver: [00:09:16] And we'll actually if we have if we have the time, it'll be this episode or the next one, I will actually I really want to make it happen in this episode. We got to do it. I'm going to take our export of our course registrations and completions, and I'm going to use Claude to analyze them. So this is the kind of Excel wizardry that I would not be able to do on my own. But now I can do it with Claude.

David Leary: [00:09:37] Yeah. And we have the right networks or right works. They have their podcast, The Modern Firm that recently launched That's in the app for today from data. Rails is in the app. So there's whatever your choice is. We have 2000 courses so it's everybody can find the CPU that works for them for sure.

Blake Oliver: [00:09:55] Yeah. Whether you're looking for something casual, news based, fun, personal development, sales and marketing or technical, we're getting more and more technical content now on the app too. So we want you and your team to be able to get all their CPP done without sitting at their desk. If you want to do it on the go, if you want to do it on your commute, you can.

David Leary: [00:10:15] And if you want to use it for free, tell your learning development person that we offer a team subscription and get your firm to buy it for for you.

Blake Oliver: [00:10:23] That's that's right. Have a send us an email sales at earmarked me and we'll let us know who we should talk to. Introduce us to your learning and Development manager. We can get you an earmark for teams subscription for your entire firm. It is a steal compared to what your firm is paying for CPE with these other providers? Uh. You can. It's ridiculous. So, uh, let's make that happen.

David Leary: [00:10:50] All right, let's jump into the news. We're going to run out of time. I can see it already.

Blake Oliver: [00:10:54] Yeah, we're we're going to. All right, let's talk about the IRS workers who were reinstated. So go listen to that interview that I did with Jeff Johnson, who was fired from the IRS as a probationary employee. He's a lawyer and a CPA. He got laid off. He told us his whole story in a bonus episode a couple of weeks ago, and it looks like he might have his job back. According to a report by the Journal of Accountancy, about 7000 IRS probationary employees who were laid off on February 20th are being reinstated following court orders. These employees will receive back pay and benefits, but they are being placed on administrative leave and told not to report to work. So, Jeff, enjoy your paid vacation. We'll see how long it lasts. Irs workers represent about 30% of the 24,000 federal workers across 18 agencies ordered to be rehired. So it's not just the IRS. It's 24,000 of these workers that were fired are being the the government is being ordered to rehire them. And it has a lot to do with basically Doge not having the legal authority to order these terminations. That's my simplified understanding of it, is that they didn't really follow the chain of command, and that's what got them tripped up. So move fast and break things. Doesn't always work in a highly regulated environment, such as when you're working with federal employees and their unions and all the protections that they have. What else is new with the IRS? Well, you know, let's let's talk about what will happen if the IRS cuts do go through, like maybe the probationary workers get rehired.

Blake Oliver: [00:12:37] But then the Trump administration and DOJ figures out how to do it properly, and then they end up terminating them. The goal has ranged from 20% to 50% reduction in workforce at the IRS. The numbers have changed, but that's the range that we're talking about. And that's a lot of workers for an agency that before all this had about 90,000 employees. What will that mean for tax collections? What will that mean for audit rates? What will that mean for taxpayers and tax professionals? Wall Street Journal did a story called IRS retreats from some audits as Agency Slashes workforce, talking about how most of these cuts are in the enforcement division, which Jeff Johnson, who we interviewed was in, not the service teams, not the people answering the phone calls, which is pretty smart, because the last thing you want to do is in the middle of tax season, make the phone lines go kaput. You don't want taxpayers and taxpayers upset about that. So they've been cutting workers in the enforcement division. And that's where Jeff was. So the argument against all of this is that, well, the IRS has been underfunded, understaffed for years, and they need those workers in order to keep enforcement levels up, to keep audit rates up, to increase audit rates, to close the tax gap. A huge chunk of tax revenue in this country goes uncollected every year because people evade their taxes. They don't pay their taxes.

David Leary: [00:14:13] Because the IRS never gets around to tapping on their door to get the money.

Blake Oliver: [00:14:17] Exactly.

David Leary: [00:14:19] There's too much work to do.

Blake Oliver: [00:14:20] The IRS does not have the resources to go after the tax cheats, especially at the upper end of the spectrum. The super high net worth folks who are evading their taxes. And there are a small number that evade millions and millions of dollars in taxes. There's actually a story. I don't know if I'll get to it about Douglas Edelman. He was a defense contractor charged with hiding $350 million of income from the IRS over many years. It's like over $100 million in taxes that he didn't pay. And this guy is already a multimillionaire, maybe a billionaire. And he did this.

David Leary: [00:14:57] That's how you get to be that rich. You have to.

Blake Oliver: [00:14:59] Well, I don't think so. I mean, he he got that rich. Well, he got that rich from, like, defense contracts, not from evading taxes.

David Leary: [00:15:07] Yes.

Blake Oliver: [00:15:07] I just never understand why people who are that rich decide. Oh, you know, hundreds of millions of dollars isn't enough. I need to hide another hundred million from the IRS. Like, what are they thinking, anyway? Those are the people that create most of the tax gap. Tax gap ranges. I'm trying to remember exactly what it is 700, 800, $900 Million dollars or billion dollars a year. Billion with a B and like, uh, people at the upper end, it's like 200 billion of that at the very high end. And people at the lower end, it's like 2 billion. It's very much happening with corporations, big, big companies and high net worth individuals. Right. So the argument is that, well, if you cut the IRS enforcement division, then we're going to the tax gap is going to increase. And that's going to offset all of the savings. And we have reported on this in the past that the numbers vary. But in general, a dollar of money that goes into the IRS for enforcement comes back as $12 in additional tax revenue collected, according to studies. So you you put in a dollar and you get $12 back. That's a pretty good ROI. But I saw another stat that makes me kind of question all of this. And this was in that Wall Street Journal article. And the stat is that 85% of taxes in this country are paid voluntarily. So?

David Leary: [00:16:32] So 100% of us are just honest, hard working people that we participate in society appropriately and we pay our taxes.

Blake Oliver: [00:16:40] And it might be way more than 85% of us. It's 85% of the total taxes due. So if if the country as a whole owes $100 in taxes, we collect 85 and that's voluntarily the IRS doesn't go after you to do that. We just do that. We have a self-reporting system. So the vast majority of Americans are honest and pay their taxes. So that 15% gap. What is the impact of IRS enforcement on that gap? Wall Street Journal says that the direct IRS collections so collections from enforcement activities from audits is only adding 2% to that 85% number. So we have 85% collected voluntarily paid properly IRS has tens of thousands of employees dedicated to enforcement, and they only add 2% to that. So I looking at that number, I think about the cuts and I think, well, what's the impact of losing these auditors on the direct collections if, if the only impact is that it lowers, say, collections by half? Well, that's only a reduction of 1%.

David Leary: [00:17:53] Yeah. So what is 1%? Can you put that in a dollar figure for me. Is this 1 billion. So 2 billion a trillion. What's 1%.

Blake Oliver: [00:18:00] So roughly we collect 5.5 trillion in taxes every year. So 1% of 5.5 trillion, whatever that is.

David Leary: [00:18:12] I'm not doing math on the show.

Blake Oliver: [00:18:14] Well yeah. We always get in trouble for this, right?

David Leary: [00:18:17] Anybody in the audience wants to do the math for us? Please figure that out. Now, I understand that 1% seems like nothing, but we're talking 1% of a trillion.

Blake Oliver: [00:18:25] That's a big number. Exactly right. So, you know, it's billions of dollars. And there's something else to consider, which is that how much of the 85% that's paid voluntarily would change if people knew that they were less likely to get audited? And so some opinion pieces have appeared saying, oh, well, Doge is going to save $100 billion. That's how much they've saved so far. But that's going to be totally offset, because now people aren't going to pay their taxes and the tax gap is going to grow and it's going to offset that. I'm not so sure about that, though.

David Leary: [00:19:02] So are you. Because are you making the argument or is this an argument that's out there which is the opposite behavior? If we do less auditing and enforcement, people will play less tricks and, well, they'll just be more open to just filing their taxes and doing it properly.

Blake Oliver: [00:19:18] Well, so what I'm thinking of is a piece that appeared in The Washington Post by Natasha Sarin, who runs the budget lab at Yale, and she argues in The Washington Post that cutting the IRS staff by up to 50% could result in a $400 billion increase in uncollected taxes over the next decade. So saying we shouldn't cut that, you know, it's going to offset the it's going to offset the gains and the. And she also points out that the IRS currently fails to collect about $700 billion in taxes owed annually. So that's the tax gap currently. And she's saying this would increase it. But I don't know about that because again, most people are honest. They pay their taxes. They don't evade them. Um, and it's very high. We have very high voluntary compliance compared to other countries. So I don't buy that necessarily.

David Leary: [00:20:12] And did you say the difference of not cutting it is going to be 700 billion.

Blake Oliver: [00:20:17] Well no. So we currently are missing about 700 billion.

David Leary: [00:20:21] Okay.

Blake Oliver: [00:20:22] Every year it goes uncollected. And that's because people are cheating on their taxes and we don't have the enforcement to make them pay.

David Leary: [00:20:30] So I just brought up the DOJ's website. I want to make sure I'm seeing this correctly. They estimated they're saving us 115 billion, but every year there's 700 billion. We're not collecting. Correct. Like he should go after that. And then he could update his website to 700 billion. But that doesn't make. Yeah.

Blake Oliver: [00:20:48] And the question is how do you go after that money? Everyone agrees that people should pay their taxes. I don't I don't run into people who say that it's great that we have a system where 15% of taxes go unpaid every year. That's because.

David Leary: [00:21:04] Billionaire friends. That's why.

Blake Oliver: [00:21:06] Not yet anyway. Right? Uh, when the podcast gets really big, maybe that'll change. But here's here's the thing. The IRS systems are so ancient and outdated. When I talk to Jeff, he told me about the master data. System that they have to access for all the taxpayer tax returns, and it's a green. Screen and you have to print stuff to PDF. And we talked about how the columns don't even wrap properly. It's very difficult. And so it makes those auditors very, very inefficient. And they can only do like a tiny fraction of the audits that get of the. They can only audit a tiny fraction of the returns that get flagged by the IRS's. Automated systems. So if you actually want to plug the tax gap, if you actually want to go after these people, you can't do it with people alone. The Biden administration and the IRS, with that additional money that they got under Biden, decided we're going to hire a massive number of people to do this. But that doesn't really make sense to me because they're only collecting another 2% with the current enforcement efforts. So let's say you add a bunch of people, how much more are you really going to collect? What's that 2% going to go to 3%. 4%.

David Leary: [00:22:30] But maybe it's.

Blake Oliver: [00:22:30] Not going to.

David Leary: [00:22:31] Wasn't that money also earmarked for build the direct file also improved technology wasn't it wasn't just bodies only. It was to do initiatives of a bunch of things.

Blake Oliver: [00:22:43] So it was it was in name earmarked. But the way the law works, the IRS can use the money however it pleases.

David Leary: [00:22:51] Yeah, once, once they get the budget.

Blake Oliver: [00:22:53] According according to my reading of it. And money's fungible. Right. So, uh, a lot of it was going to go to enforcement, I think about half and then half for service and technology. And they have been they say they've been working on modernizing the tech, but they didn't do it. They had years to do it and they didn't do it. And they've had decades to do it and they haven't done it. And the IRS is now saying, oh, we're going to stop those modernization efforts because our budget is being cut. They lost another 20 billion with the latest continuing resolution. So you know, that 80 billion they got, it keeps getting whittled away. They've spent like 16% of that I think. So my question though is why hasn't the IRS done this anyway. How many people would it actually take to modernize this tax return data system, this giant database that they have? How many people would it take? Would it take a team of ten, 100, 1000? I can't see it taking more than that. And the IRS already has tens of thousands of employees. And what are they doing with these employees? They're putting them on phone lines, the information systems that the IRS is so bad, they have to use people to process the information manually. They have to use people to to handle these tax returns that come in on paper.

Blake Oliver: [00:24:20] So the solution is better information systems, which is kind of always the solution in accounting these days and tax better systems. And they haven't done it. And they could if they wanted to. They have the money to do it. I just think that because they haven't it must be a culture issue at the IRS. It's the people leading the IRS. It's not really a priority for them, and they are looking at the systems and solving the problem with people. And you cannot solve this problem with people alone. You have to do the technology modernization first. That has to be the priority. And I bet you if you did it, you could run the IRS with a fraction of the people they have now, because how many tens of thousands of people are just sitting there answering phones and giving information to taxpayers and tax pros that they looked up on a computer, which the tax pro could just go in and look up themselves if they had access to this system. And that's starting to happen with a few little things, but it's not nearly enough. So that's my view of this, is that if we actually want to solve the problem, it needs to be done with tech and the current IRS.

David Leary: [00:25:32] Not done yet. Right. The IRS has been paying PwC, Deloitte, KPMG and EY for consultation on building these computer systems and it's not getting any better.

Blake Oliver: [00:25:44] Well, because who are they paying this? They're paying government contractors who are paid on time and materials and have zero incentive to actually ever finish the project. It's the system. They should just hire these people. They should use some of those billions of dollars that they still have, and they should go out, and they should hire tech people to come in and build them a modern system.

David Leary: [00:26:06] Yeah. Because ultimately the big four, who all have their own tax divisions and clients paying them about tax stuff. If the IRS got really efficient, you might not need to hire the big four to get your taxes done. Like it's a it's a super conflict of interest. Of all the government contracts the big four should be working on is probably not the IRS modernization effort. It's.

Blake Oliver: [00:26:24] Yeah, exactly. It's the opposite. Right. They have cross interests. They it is not in their interest to make the IRS more efficient.

David Leary: [00:26:30] They're probably purposely putting bugs in to make it worse.

Blake Oliver: [00:26:34] Yeah. And there's proof of this. Uh, CPA trendlines did a story about how the chaos at the IRS is actually an opportunity for tax professionals. And it is if the IRS falls behind on correspondence, if you, as a taxpayer, can't get through to the IRS and figure out how to sort out your issue, you're going to have to hire a professional. And so CPA trendlines is saying it's an ideal time for tax preparers to expand into IRS resolution services if you don't mind dealing with the IRS. If you don't mind that being part of your business, then that's a great business to be in because the IRS is in chaos. And if you can master the tools that are available to you, the tax transcripts tool, the online portals, the direct pay like filing electronically, you can really help people out navigating that dysfunctional system. So from a.

David Leary: [00:27:34] Marketing position, lean into it. Yeah. Lean into it. Into the IRS being chaos and how you can be the those people that haul your when you're going up mountain or they're carrying your stuff up mount Everest.

Blake Oliver: [00:27:49] The Sherpas.

David Leary: [00:27:50] Sherpas. Yes. You could be their sherpa for the.

Blake Oliver: [00:27:52] Tax sherpa.

David Leary: [00:27:53] Your tax sherpa for the.

Blake Oliver: [00:27:54] I bet you that's a name for a start up there.

David Leary: [00:27:57] Yeah. Firm somebody firm name is going to use that now.

Blake Oliver: [00:28:00] Welcome. Adam. Zaki. Adam said that it's 50 billion. The math we couldn't do the 1% of five. 5.5 trillion, he said. Just a reporter, not a mathematician. Thanks, Adam. We should probably get somebody on here to do the math for us live. That would be great. Maybe we should bring back ChatGPT.

David Leary: [00:28:22] We just have a person that sits on the side there, just doing nothing but calculating math for us.

Blake Oliver: [00:28:26] That's right. Fact checking us live. Welcome. Boring. Accountant. You made it five coffees this morning. That's what I'm on about. Hey, David. I have been going on about the IRS, and I forgot. It's time to thank our first sponsor.

David Leary: [00:28:40] So our first sponsor is right. Works. Feeling overwhelmed by rapid changes in accounting. You're not alone with cyber threats at an all time high. Are AI transforming workflows and industry consolidation reshaping the landscape? It's a pivotal moment for our profession. That's why I'm excited about right now, the ultimate mid-year event for accounting professionals, happening May 19th through the 21st in Nashville. This is your chance to recharge after tax season while earning 14 hours of CPE with the industry's brightest minds. Hosted by Right Works, Right Now delivers exactly what you need right now. Cutting edge insights on AI, cybersecurity, staffing, and scaling your practice. You'll experience iconic keynote speakers, technology demos, and powerful networking opportunities, plus an authentic taste of Nashville nightlife. Whether you're a firm owner or staff accountant, right now equips you with the strategies you need to thrive in 2025 and beyond. Register today at The Accounting Podcast. Dot promo slash right now. 2025 and save an extra 15% with code. With code, tap 15. That's The Accounting Podcast dot promo forward slash R-i-g-h-t NOW2025 and use code tap 15. Also go to their website. They use the. Use the URL, go to the website and you can see who these influential keynote speakers are. They're definitely. You don't want to miss them. I don't think we're allowed to say who they are, but you definitely want to go to the website and look at it.

Blake Oliver: [00:30:07] Awesome. All right, David, let's talk about rippling versus deal. What the heck is happening? I saw this press release from rippling, and I thought chaos just follows Parker Conrad wherever he goes.

David Leary: [00:30:23] That was my first instinct that when I saw that. Because if you remember, if we take a time machine a little bit, I would say I was still into it. So we're going 2017, 2018. He had his own company at that time called Zenefits, and Zenefits and ADP locked horns over. I think they were a partner maybe possibly at that time and referring customers back and forth. And then I don't know if Zenefits started offering their own payroll product, and then ADP got mad, or ADP started offering their own HR benefits type stuff, but they were stealing clients from each other essentially what they were accusing each other of. Then they made up and they figured it out, I think. And then Zenefits had all those problems were. Another crazy story. The employees in Scottsdale were having sex in the hallways and they cheated on their insurance exams. So there's just I.

Blake Oliver: [00:31:09] Think it was the stairwells. It was the stairwells.

David Leary: [00:31:11] The stairwells.

Blake Oliver: [00:31:12] And Tempe. Right?

David Leary: [00:31:13] Yeah, it was in Tempe. Yeah. All the ASU kids that graduated, I guess. Um, so.

Blake Oliver: [00:31:18] So Parker Conrad founded Zenefits.

David Leary: [00:31:20] Zenefits.

Blake Oliver: [00:31:20] And there was all that controversy about Zenefits. And then he went on to start rippling, which has been even more successful. Yes, but is now embroiled in another controversy.

David Leary: [00:31:28] Another controversy. So he had a tweet that he put out saying how they are suing deal and the d d.

Blake Oliver: [00:31:36] E e l.

David Leary: [00:31:37] It's a competitor. It's a similar it's um, employer of record company. Both of these companies are employer of record. They took two different paths. Rippling really took the path of when you hire an employee, all the things that have to ripple from that employee, you need to issue them an expense card, you need to issue them a laptop. There are all these pieces you need to put them on payroll. It's all related to hiring of an employee. And deal. They came into the space from the I need to hire employees and a bunch of countries all over the world. How do I hire people in Brazil if I don't have any paperwork to do that? And they become, if I don't.

Blake Oliver: [00:32:09] Have an entity there? Yeah. Setting up your own business in Brazil, not something you want to do if you're a US startup. Yeah.

David Leary: [00:32:15] And so they're both competing with each other on being employers of records. So he puts out this tweet that they're suing them. And the reason why rippling is accusing Diehl of planting a spy in rippling.

Blake Oliver: [00:32:28] Yes. A Diehl employee allegedly infiltrated rippling Dublin office.

David Leary: [00:32:34] Yeah. They hired this person for a management management role in 2023. And at one time, apparently, Diehl had used rippling software. But rippling opted not to renew Diehl's contract. So I don't know if Diehl used rippling internally for their own teams, but there is. So they have been partners on some front or customer of each other. At one time.

Blake Oliver: [00:32:57] Rippling had a deal with deal, but the deal fell apart. Yes, the deal spied on rippling. Okay, so they they put a spy in the Dublin office? Yep. And the spy, over four months, allegedly conducted over 6000 slack searches. And the searches targeted confidential sales information and competitor intelligence.

David Leary: [00:33:19] And most of the time, they were searching to see what customers were switching from. Deal. So maybe you're a customer of deal. And now you're you're having chats and sales calls with rippling. This person's finding those chats and finding who those customers are and then passing that information back over to deal so they can go save those customers so they don't become rippling customers.

Blake Oliver: [00:33:40] Mhm. So how did rippling figure this out.

David Leary: [00:33:44] So they figured it out really by looking at data that's inside of slack which this part of it's a little scary. So if you're an employee and your company uses these chat tools like slack, they can see everything you do, including what you're searching for.

Blake Oliver: [00:33:58] And your private chats, your private messages, everything. Yes.

David Leary: [00:34:02] So they set up a honeypot. And what was the honeypot called? It was called. It was a slack channel called Dash defectors, as in deal defectors. And apparently this person was never actually invited to it. And they ran up.

Blake Oliver: [00:34:17] Against General Ripple's general counsel, sent a legal letter mentioning this channel to only three deal executives. And then within hours, the suspected spy searched for and accessed this non-existent channel. I'm curious, though, the part that's not clear to me is how did the spy find out about the channel if the letter was only sent to three deal executives? Is someone else involved in this that we don't know about yet?

David Leary: [00:34:43] Well, the spy obviously was working under the direction of deal, and I saw a comment that it might be the father of the CEO of deal.

Blake Oliver: [00:34:52] Oh, I get it. The legal letter was sent to three deal executives. I thought it was rippling. Executives. Got it. Okay. Yeah. So they sent a letter to deal, like a legal complaint or something. And then the spy looked for the channel mentioned in that letter. Oh. That's genius.

David Leary: [00:35:08] And then? Then they looked at the unusual patterns of slack activity. Um, and they took the bait. He searched for terms that were in that letter. Um.

Blake Oliver: [00:35:20] Now this is where it gets wild. So rippling. Got a court order to seize the employees phone and showed up. They they they sent court appointed solicitors to replace Dublin office and and cornered the employee who ran into the bathroom. Right.

David Leary: [00:35:42] The employee to flush the phone. Right.

Blake Oliver: [00:35:45] He was going to try to, like delete the information, clear the phone. Locked himself in that bathroom. He was warned about violating the court order. It's illegal to destroy evidence when you've got a court order to obtain it. And he allegedly said, quote, I'm willing to take that risk and then fled the premises. Solicitors heard a toilet flush, suggesting possible evidence destruction. Wild sow rippling filed a lawsuit that happened on March 17th. This was filed in the Northern District of California, San Francisco division. They're alleging trade secret misappropriation, corporate espionage orchestrated by deals leadership, systemic theft of confidential business intelligence. Stolen information included sales pipeline data, proposed pricing, and competitive strategy. Wow. And if true. If true.

David Leary: [00:36:40] If true.

Blake Oliver: [00:36:41] Very disturbing.

David Leary: [00:36:43] And they've actually have had a lot more fighting that's taking place. And there was a even more crazy the story as I was looking into it on January 3rd. There was actually a lawsuit accusing deal of criminal Rico enterprise sanctions.

Blake Oliver: [00:36:59] Rico. What's that?

David Leary: [00:37:00] Uh, that's that's like a conspiracy to commit wire fraud and all that, right? What the.

Blake Oliver: [00:37:07] Yeah, I think you're. Yeah. Yeah. That rings a bell.

David Leary: [00:37:10] So let me recap the players for a second before we get into this lawsuit a little bit. So you have rippling and deal. You have surge capital. Surge capital was a and I don't know if we ever talked about it on the show or not. They were, uh, an investment fund, real estate investment fund in Florida that defrauded a bunch of people at church out of $35 million, per the SEC. You have T Bank, which is a Russian bank formerly called Tinkoff. We have Airwallex, which which is a debit card type money transfer company that a lot of the startup apps use. We have guess who our big favorite evolve bank and trust. We have the Department of Treasury's Specially Designated Nationals and Blocked Persons list, and we have U.S. senators. All of these people are all tied to this mess that's happening.

Blake Oliver: [00:37:58] Wait, so this was a separate this is before the whole deal thing.

David Leary: [00:38:01] This was separate. So this was a this was from January 10th that this this this news came out and I'm going to read this a lot of this is from a tweet from Jason McCullough who has the Fintech Business Weekly podcast. But he did a really nice huge long tweet because there wasn't a lot of news articles about this. There wasn't an Inc. magazine. Um, so the 50 page class action suit was filed in U.S. District Court on in Miami on January 3rd, alleges that CEO of operating a criminal enterprise. And this lawsuit is tied to the search capital stuff. So this other crime that happened, people's lawyers start asking questions and wanting things opened up, and then they discover and they discover and they discover. And what's tied to this, I'll just skip to the kind of a spot deal. Did provide a payroll service and paid one employee for surge capital, but was they say they're not directly involved in the fraud. But I think this is where these things start opening up cans of worms. Um, so there's a lot of allegations in this. Essentially, they were arguing that they didn't follow the proper know your customer and, uh, licensing things for the banks and money transfers. And they were they facilitated money laundering via steal card and deal wallet products. Um, colluded with sanctioned Russian banks, specifically T Bank, previously known as Tinkoff, which is on the Treasury's do not um I called it specially designated nationals and blocked persons list. So they were doing work with this Russian bank. Right. That was already on. Now, apparently they've scrubbed a lot of this. You can't find it on the website or not. And they actually went out of their way. Deal to say how they can help you do things in Russia. So I have a screenshot I'll share on the screen here. So this is no longer on the deal website, but if you want to read it Blake with the the marketing message here.

Blake Oliver: [00:39:55] It says hire and pay employees in Russia hassle free with deal. Your business can easily hire employees in Russia. No more worrying about local laws, complex tax systems or managing international payroll. Deal takes care of everything. In 150 plus countries, the average onboarding time frame in Russia is five days with deal. So they were doing this after the whole Ukraine war stuff started.

David Leary: [00:40:17] And Tinkoff website or the T Bank, they actually even recommended deal as an avenue to evade the sanctions. They recommend this Russian bank recommended to use deal to avoid these things. Now the the lawsuit may misunderstand. It was talking about M2 which is money transmitter licenses and MSB which is money service business requirements apparently payroll processors or these employer of records. They don't have to follow the same registration and licensing as like a PayPal or a square or somebody that's actually moving money around. You have you have different rules and laws to follow, so they may not have to follow those same money transfer things. And they apparently have a subsidiary called D payments, and that they actually have 25 of those licenses now in different states. Um, but they didn't this is not in the lawsuit, but they used to have a thing called the deal card. And the deal card was powered by Airwallex. Airwallex is a middleman app for moving money around the world. If you want to do international payments and things like that, they provide a debit card type platform for that. Um, and Airwallex is bank partner is evolve Bank and trust. So here we go. Another totally questionable behavior happening by somebody that was using evolve Bank and trust.

Blake Oliver: [00:41:39] Remember remind everyone what happened with evolve Bank and trust.

David Leary: [00:41:42] Yeah, so evolve Bank and trust was heavily involved with synapse. They were heavily involved with FTX. They've had all and I think somebody called it out before. Their lack of discipline on tech companies was actually its best feature. That's where all these tech companies use them, because you could be wishy washy in your compliance.

Blake Oliver: [00:42:01] Because they weren't really enforcing the know your customer rules. All that strictly.

David Leary: [00:42:06] And I don't even know, like if that's the real reason. Evolve Bank and trust I think was a historical normal old bank. And then they discovered, hey, we could be the bank for tech companies. And I think they got in over their head is probably the real thing. It probably wasn't any conscious decision. I just think they grew too fast. They took on, they created some APIs. They didn't know what they were doing, and they got in over the head. And that's why they were not exposed in the right word. Um, exploited. Right. A lot of these tech companies kind of exploited their use of evolve bank and trust. So deal did respond back to this. And apparently this is where it ties all the way back to Ripley. This lawsuit was filed by an early investor in Ripley, Thomas Grady. So it's full circle. These companies are accusing each other of so many things, and there's Russians involved and then senators. So deals company advisory board has former transportation secretary and wife Senate, Senate Minority Leader Mitch McConnell, Elaine Chao, former U.S. Senator Heidi Heitkamp, former acting Labor Secretary Seth Harris, and former Mississippi banking Commissioner Charlotte Callie. So there's politics in this. There's Russians. It's it really reminds me, if I think back to the documentary on Wirecard in Germany, there was politicians that were involved. The it was those cards, those debit card money, movement type stuff. Russians were involved. We're going to find out so much more about this story, because every time a lawsuit happens, discovery happens and you get those onions opened up, we're going to see more crazy accusations. And the great part of Conrad being involved, he puts it all on Twitter. He did that with ADP. It's a public feud. So we get to watch this train wreck as it keeps happening. But it's it's crazy. It's a crazy, crazy story.

Blake Oliver: [00:43:56] I can't wait for the made for TV movie. This is going to be spectacular. Who's going to play Parker Conrad? Who's going to play Elaine Chao? We'll see.

David Leary: [00:44:07] And I think it's important to say deal has not been accused of any wrongdoing by the SEC at this point. So no.

Blake Oliver: [00:44:13] It's all it's all allegations.

David Leary: [00:44:15] Allegations.

Blake Oliver: [00:44:15] This is all allegations may be disproven. Uh, yeah. We don't want to get sued by deal.

David Leary: [00:44:21] No.

Blake Oliver: [00:44:24] All right David, that is amazing. I can't wait to talk about this in the weeks that come as we learn more. Right now, I want to pivot to artificial intelligence and a really interesting way of measuring AI. But first, let's thank our next sponsor, Bluevine.

David Leary: [00:44:43] I'll let you read.

Blake Oliver: [00:44:44] Your small business clients deserve banking built specifically for them. Meet Bluevine, the business banking platform created to save you and small businesses time, money and hassle. With Bluevine, your clients don't need a savings account. They can earn up to 1.5% APY on checking balances when they meet an activity goal, or up to 3.7% APY automatically. With upgraded plans and with no monthly fees for standard plan customers. Free standard ACH and competitive rates on same day ACH wires and international payments, they'll save more where it counts. But Bluevine doesn't stop there. Powerful tools like multiple Subaccounts for better budgeting, shared account access with secure logins and automated accounts payable streamline how your clients manage invoices, approvals and payments, and for you. Bluevine is dedicated account and dashboard gives 24 over seven secure access to all your clients accounts in one place. No more chasing down statements or check images with seamless integration with QuickBooks online reconciliations are faster and accounts payable workflows are simplified. Plus, as a partner with Bluevine, accountants get access to a free, upgraded plan with 3.7% APY and a bonus of 200 or more for each new qualified client referral you make with Bluevine. You're not just managing accounts, you're setting your clients up for success while making your job easier.

Blake Oliver: [00:46:10] If you're ready to become a valued Bluevine partner and unlock exclusive rewards just for accountants, head over to The Accounting Podcast Dot promo slash Bluevine. And now AI is getting better and it's getting better at long tasks. This is a fascinating paper from researchers at Metter, they propose that we should measure AI performance, not in terms of the complexity of the tasks it can do compared to humans, but in terms of the length of tasks AI agents can complete the length of time. And if you do this, and you plot out the time, it would take a human to do the tasks that the newer AI agents can do. And you plot that out over the history of these AI agents going all the way back to GPT two, you see that we are actually doubling every seven months. And that is very similar to Moore's Law, which we saw in transistors, where the I don't remember exactly what it was, but for many years, decades, the compute power of chips would double every x months or X years. And we're starting to see that with AI. When you look at it in terms of the length of time, it takes a human to do the same task.

Blake Oliver: [00:47:41] So GPT two, for example, was able to handle a task that would take a human 2 or 3 seconds. Gpt three, you're getting up to 10s. Gpt 3.5. Now we're between 15 seconds in a minute. Gpt four, we're approaching four minutes. An example of that kind of task would be counting the words in a passage. So for a human to just count them all would take four minutes. Perhaps, but GPT four could do that now in a lot faster. So it's not the amount of time that the AI is taking to do it. It's the amount of time a human would take to do it. And it just keeps going up like that. And now with sonnet 3.7, we are at the point where the AI can do a task in just seconds. That would take a human an hour. What would that be? What's an example? Well, a lot of us are using sonnet to write. We write articles, we write scripts, we create outlines, we create PowerPoints. That's all stuff that used to take a long time and now just takes minutes. Or in many cases.

David Leary: [00:48:53] I could see the math on this. I sent you a blog post to review last week for last Friday, and it probably took me 15 seconds to create it with with Claude.

Blake Oliver: [00:49:04] So if this continues, if this continues doubling every seven months, then very soon, within 2 to 4 years, AI systems will be able to complete tasks that would take a human a week, which is hard to imagine. What kind of task would take a human a whole week Writing a detailed research report.

David Leary: [00:49:30] Perhaps I wanted to dig on the deal stuff and the rippling stuff more. But there's not enough time and there's just too much to do. It would probably be a week of research if I really wanted to super dig. Yeah, it would be a week. Whereas on it.

Blake Oliver: [00:49:43] And currently, if you use perplexity deep research, which I use to get ready for this episode, when it came to that story, I mean, it can do that in minutes and that would take me hours right now. So I think this makes a lot of sense based on my experience. Now, it's not just about how much time it would take a human. It's also about accuracy. So it's only about 50% successful. So the task might take a human around an hour, but the AI and the AI can do it in seconds or minutes. But they only have a 50% success rate. And currently, if the task takes more than four hours of human time, the success rate is below 10%. So as the models improve, they'll get better at tackling tasks we can do and they'll be more successful. So currently they can reliably they can reliably complete tasks that today take humans a few minutes. So that's why we're seeing a lot of success in startups that are focusing on building these AI agents that can do small tasks, small task automation, like in accounting, that would be entering a transaction into the general ledger or posting a simple journal entry or, I don't know, adding a new employee to an HR system. Those all take a few minutes. Perhaps I can do that reliably now, and we're going to get up to the point where it can do tasks that would take an hour very reliably. Currently it's only successful half the time.

David Leary: [00:51:19] So if you can break up everything you have to do in a week into 15 minute chunks and start automating those 15 minute chunks. In theory, these agents in AI could start turning those 15 minutes into 15 seconds.

Blake Oliver: [00:51:30] Yes.

David Leary: [00:51:31] So give it all the work you do for a week. It may not be able to get there for a while.

Blake Oliver: [00:51:35] And this is why I'm not really buying this idea that, as Sam Altman has said, that we're going to lose 7,080% of jobs in the next few years because the success rate has to be pretty high before you start to replace humans. And we're going to take a while to get to the point where the AI can do just one task that would take a human a week. But we all know that the jobs that we do are not just usually not just one task at a time. It's a lot of different tasks of different varying complexity and length. So the AI will be able to do the smaller tasks, leaving us to focus on the bigger ones. So I'm, you know, it's going to enhance our productivity. Um, if the trend does continue, though, by the end of the decade, what will it be? I will be able to autonomously complete month long projects by the end of the decade if that trend continues. So something would take a human a month. I would be able to do so.

David Leary: [00:52:38] If we wait another two years, it could redo the whole IRS website and computer system if we just wait long enough.

Blake Oliver: [00:52:44] I mean, you could you could, you know, how many people would it take actually, like if you're listening and you have any idea. I mean, I know this is a big, complicated system that the IRS uses to manage all of its tax return information. This I forget what it's called master data, something or other. Um, it's built on COBOL, which is a language that nobody learns how to write anymore. And it's ancient computer technology. How hard would it be to rebuild that database in a modern cloud based server environment. Especially if you had AI tools to help you write the code, to look at the current database and restructure it and rebuild it and import the data it. It would not take that many people to do it. The resources exist to do it. Now. It's just the lack of the will to do it, the lack of knowledge at the in the federal government to do it. And then when they hire consultants, like you said.

David Leary: [00:53:48] They just never get it done.

Blake Oliver: [00:53:49] Take the money, it never gets done because the whole system is dysfunctional.

David Leary: [00:53:55] I had an article that it wasn't a good enough story to bring to the show, but I'm going to mention it. There's an AI company that has to restate its earnings, and I can't find the article. I don't remember if the company was called, but I'm like, why aren't they using their own AI to do their own through their internal accounting?

Blake Oliver: [00:54:12] Because it's not good enough yet to do that.

David Leary: [00:54:14] It's not good enough. Yeah. No.

Blake Oliver: [00:54:15] Yeah. So um, you know, I can do the simple stuff. The bookkeeping work, right? The data entry work. The the processing of invoices, maybe the approval of expense reports. Um, you know, it's not going to be able to do the complex accruals and the creation of the GAAP financial statements like that is not something we have to worry about. Well, at least for a few years. Um, David, to give everyone a, uh, an illustration of like what is possible right now with cloud and data analysis. I want to do a little demonstration live. So this is Claude. Claude, I if you haven't seen it before, pretty spectacular. It just got web search, which was something that was missing. So now when you, like, do a prompt, you can toggle on a web search and it'll go out and search the web, which something ChatGPT does.

David Leary: [00:55:13] Chat doing that Gemini is Gemini's doing that. Perplexity. They all do that.

Blake Oliver: [00:55:16] Okay, they do that. But sonnet, you know, is is I've been reading about how developers are saying like the code that comes out of sonnet is just really, really good. It's like best in class right now. And I have found with Claude that like the quality of the writing because I do a lot of writing, uh, the quality of the writing that I get with Claude to help me draft and edit my content and prepare for this show is just the best. So I have a spreadsheet. I have to show you the spreadsheet first before we import it. Uh, let me share a different screen. So this is a spreadsheet. Is this showing up yet? Here it is. Got it. Okay, it's on the stage.

David Leary: [00:56:06] Yeah, that's useful to anybody listening.

Blake Oliver: [00:56:08] Yeah. Sorry. Uh, okay.

David Leary: [00:56:10] So the spreadsheet of data.

Blake Oliver: [00:56:11] Well, okay, so we've got this is a dump from our app. From the earmark app. It's a dump of course data. So we've got the ID of the course. We've got the course title the channel name, field of study, the date it was published, the instructor names. And then this is the key information registrations, completions, CPE credits, earned average rating, total ratings and the status of the course, whether it's active or draft or retired. So this has like 2000 lines in it because we have we have more than 2000 courses now.

David Leary: [00:56:44] Maybe 26,000, 2200 maybe.

Blake Oliver: [00:56:48] Um, so that's a lot like how do I how do we, David, figure out what are the most popular courses on earmark? Well, I'd have to, like, create a sheet and do some charts and manipulate this in Excel and try to figure this out. Well, I'm going to try asking Claude to do this.

David Leary: [00:57:04] So and then right before you do that, that would take you how many hours? At least three hours. Right.

Blake Oliver: [00:57:10] Well, if I even had the knowledge, right, I'd have to learn how to do this stuff in Excel first. And you know, I didn't I didn't. I've never taken an Excel course in my life. I've just picked it up on the job. So, you know, I wouldn't call myself.

David Leary: [00:57:22] In that generation that, like, slipped through, like you never learned Excel. You just went straight to Google Sheets or something. Well, did the education system fail you? And you did not learn Excel?

Blake Oliver: [00:57:32] I mean, they don't teach it in accounting. It's crazy. You don't learn Excel when you take an accounting undergraduate degree. At least I didn't. Is that nuts or what? I mean, I.

David Leary: [00:57:44] Can take it as it was more of an information mis type class I think I took. Well, I'll admit I was Lotus 123 is what I actually took. But anyways.

Blake Oliver: [00:57:54] So yeah, it's just it's don't get me started on accounting education. It's just nuts to me that we don't teach Excel.

David Leary: [00:58:02] So so you got this spreadsheet. You just took it and you shoved it into court?

Blake Oliver: [00:58:06] Yes. Well, you know, I uploaded it gently, but yes, it's an exe. This is the export. It's a CSV file. And what's amazing about this is I don't really have to be that specific to start. I can just say, you know, here is a dump. Of course, data from the earmark app, please analyze it for me. And I'm using the experimental. I'm going to turn off web search. And I'm going to make sure that analysis tool is turned on. This is so that cloud can write and run code. And it's going to write code and then execute it. So I just said analyze it for me. I didn't give it any specific instructions. And Claude is saying I'll help you analyze the course data from the earmark app. Let me first examine the file to understand its structure and content. And it just starts writing code. I it's even notating the code. And at first it can't read the CSV file. It struggles. And so it's trying a different approach and it continues to write code. And it's just it's going. It's going.

David Leary: [00:59:15] Yeah. You just see the magic code getting and this is hundreds of lines of code.

Blake Oliver: [00:59:19] Yes.

David Leary: [00:59:20] If you were to be an engineer, if you wrote this code yourself.

Blake Oliver: [00:59:24] And now it's looking for potential date formatting issues, it's looking at the instructors. It's getting the top 15 instructors. So it's on its own. It's looking at the column headers and the type of data we have. And it's predicting what type of analysis we might want to do. And it's going to do that proactively for us. Now, every time I do this I get a slightly different response. And so it's it's worth doing this a few different times and then maybe giving it a more specific prompt later. And it keeps trying. It's extracting insights. It's I mean, look, if you had to if a human had to write all this code, it's just blurring past us on the screen here. This would take a long time, right?

David Leary: [01:00:08] Yeah. If you were to set all this up in Excel and do Visual Basic and make interactive charts and graphs, it's going to take you hours. Yeah.

Blake Oliver: [01:00:16] You can't. A lot of this you can't even do in Excel. You'd have to do it in like a business intelligence tool or something. Now it's putting the code into what's called an artifact. And an artifact is like a tiny app that Claude can run inside of Claude. So. It's still going. I didn't realize this. It feels like this is taking a really long time when we do this live. But normally what I do is I just. I put in the prompt and then I get the answer, like, I come back later.

David Leary: [01:00:52] We're complaining because it took 48 seconds to write a thousand lines of code program.

Blake Oliver: [01:00:57] Okay, so I got an error and now there's a button here to try fixing it. Said unexpected token. Okay. So I'm going to try fixing it. This usually fixes it. I hope this works live. I'd be very disappointed if it doesn't. Let's go look at the chat while we wait. Erica says I had to take an Excel class for my accounting degree. Where'd you go to school? Erica, that's really good. That's nice to hear. Let's see. Boring accountant says why can't I do my pivot tables for me? Probably could. Okay, we got it. The visualization is here. So we've got a bar chart showing courses by field of study. And we can see that business management and accounting are the most popular. We can hover over. And we've got a little pop up that shows the number of courses. It didn't quite get the labels all right on here. It's missing some of the labels but it's got most of them. Finance then taxes, then personal development then communications. It shows you mouse over.

David Leary: [01:02:02] You're able to see the labels. It just didn't put them on the.

Blake Oliver: [01:02:04] Didn't put them on the chart. On the chart, um, it's got our top five channels. So we see the The Accounting Podcast top five by number of courses. The Accounting Podcast. Then for today, unique CPA earmarked podcast. Oh my fraud. We've got bar charts to show that you can mouse over courses by year. So we've got the number of courses for 2022. We can see it goes up each year to 2024 and then it drops down because of course it's 2025 is just started. We can see the distribution of active versus inactive courses. Key statistics. Total number of courses 22 2215. Total registrations 130,642 a completion rate of 89.64%. So that's of the people who registered how many people completed each course. On average, we've got an average rating of 4.87 out of five. And then we've got registrations versus completions as a chart. Okay, so what else could we ask it, David, to do? Oh, it also has a it also has text based analysis as well. It's saying um that let's see our top courses by registrations a very very very nice lady did some fraud from oh my fraud with seven 570 registrations. And then we've got these are actually all oh my fraud episodes. So all top five are courses are oh my fraud with over 500 registrations each. And it shows the top instructors. Oh this is funny. It lists CPA as the top instructor I guess because it's the name then comma CPA. And it's interpreting. It's parsing that as a name. Yeah. Uh, we've also got myself, you, Paul Barnhurst, Randy Crabtree and then it's got key observations. Strong growth trend, high completion rate. What if we ask it what are the most Popular like identify the topics of the most popular courses. Right? Like, you know, based on the course titles, what are the most popular topics? For courses. Um, popularity should be according to completions. How about that? Or let's just say registrations and we'll say based on the course titles and channel names. So I'm curious to know if it can do.

David Leary: [01:04:35] Something like we would almost be impossible for us to do manually this question.

Blake Oliver: [01:04:40] Right, to go through all the course titles and the names of the shows, and then try to figure out what topics are the most popular. I would have no idea where to even begin writing that, kind of doing that kind of analysis.

David Leary: [01:04:53] So. So as it does analysis and writes this. Blake, I know you talked about how there's a little app or a widget that runs inside a cloud. Can you save that? So next time I just say I love that widget. Here's the new data set. Apply what you already wrote against the new data set. Or are you creating a new UI every single time you do this?

Blake Oliver: [01:05:10] So the way this works is it writes new code every time. But what what Nicholas says, he shows me in the earmarked podcast episode that everyone should go listen to. He shows how to take the code and and drop it into like a code editor. He uses Google's as an example, which I've never used before. So you take the the code that's written. He used ChatGPT, not Claude. He takes Python code from ChatGPT, drops it into the code editor, and then you can upload a new data set so it runs the same code every time.

David Leary: [01:05:44] Okay. So you're taking the code from here and having it run in a code only environment.

Blake Oliver: [01:05:48] Yes. Now you know, you also have to validate that like this is all Correct. And how would you do that? Well, you could do some like checksums essentially. Right. You want to make sure that the number of courses is correct. The registrations look correct. You want to spot check this.

David Leary: [01:06:03] And that didn't make up any courses because I think I did it on like in the chat 1.0 days. We I tried to cut a corner and use ChatGPT for this, and we sent a report to somebody and that person came back and said, that's not my podcast title or whatever. It just made up a podcast episode that looked good.

Blake Oliver: [01:06:21] So we've got our analysis back. Top ten course topics by total registrations. Accounting is the most popular.

David Leary: [01:06:28] While Captain Obvious thank you.

Blake Oliver: [01:06:30] 43,000 registrations across 569 courses. Fraud is next, 38,000 registrations across 140 courses. And it gives the top example for each of these. Um, and it says this topic is disproportionately popular with fewer courses but high registration numbers, which is true. Fraud is very popular tax. And it gives the top example, which is an episode that we did recently about Trump's tax plans. Then CPA focused content, then AI and technology, then business topics, then financial topics, then QuickBooks, then client management, then audit. So we could probably dig into this and say, well, you know, we only have 31 audit courses. We need to add more. Maybe if we add more, then it'll become more popular. If people think, this is where I go for my technical content, not just my fraud and accounting and tax and that sort of thing. So then it then it has a narrative of key insights on popular content. Entertaining titles perform well. Courses with creative, unusual or humorous titles tend to have higher registration numbers. And it gives example examples like the one we mentioned before. A very, very, very nice lady did some fraud and dishonest, remarkably stupid conduct. Those are good titles. Case studies are engaging, says many. Top performing courses feature specific case studies of fraud, current events and news popular. So like this is incredible.

David Leary: [01:08:11] So why do I want an accountant? Like. Like I'm a small business owner. Why do I even want an accountant? At this point I could just download from American Express all the transactions I had for a year, and I could upload them in this, and I could get knowledgeable results back. Wait, what are accountants supposed to do? Like, it's like when. When it's not. When they say I is going to replace you, it's not that it's really going to replace the work you're doing. People just aren't going to see the value. Like, why am I going to pay somebody else to tell me this, that I can just have a GPT do it for 20 bucks a month?

Blake Oliver: [01:08:45] So boring.

David Leary: [01:08:47] Accountants answer this. I'm just thinking out loud like, sorry.

Blake Oliver: [01:08:50] Sorry, what's the question? I was distracted.

David Leary: [01:08:51] What if if if I'm a business owner? Yeah, right. And why do I even want to? Why would I engage with an accountant? I could just take the dump of transactions from Amex into a CSV file spreadsheet, throw them into a cloud, be able to ask questions, get answers. I don't need an accountant to do advisory for me. Like, why do I need an accountant?

Blake Oliver: [01:09:12] You need to know what questions to ask. I think many small business owners would have no idea where to even start. Now you could use a tool like this.

David Leary: [01:09:21] And say, what questions should I be? Yeah. Tell me what questions to ask. I could start there.

Blake Oliver: [01:09:25] But you have to know to ask that question. Okay. So I mean, you know, when it comes to like, just analyzing your spending, paying an accountant to take your Amex dump and categorize all your transactions is that's something you could do yourself. You could sit.

David Leary: [01:09:43] Lots of people do. Right. That's a lot of people do. They fax over the Amex statement they handwrite on the lines of the Amex statement what they spent the money on. Yeah, they they fax it over.

Blake Oliver: [01:09:55] If you have a CSV file or an Excel file of like transactions, and you just want to do like a write up, I'd actually be interested to try that now with sonnet. It hasn't been so good in the past, but I bet you that we could get it to if we gave it the chart of accounts, and we gave it a dump of transactions like date description and gave it. Gave it. Yeah. Well, I mean, this is before we don't have a GL, right? We're just trying to do write up work. We're just taking our credit card statements or whatever. And I wonder if how well it would do, applying the chart of accounts to those transactions and how accurate it might be. You could probably get 8,090% of it done, at least good enough to do a cash basis tax return, because the as long as you get the income and the expense, right, you know which exact expense account you choose is not that important because what really matters is the ending net income.

David Leary: [01:10:50] And I think Hector Garcia, I think I don't even know where I heard him say this at. It might have been over a beer, it might have been on a podcast, but at the end, the business owner doesn't really care if if if it's close enough for them to pay their taxes almost accurately. They don't care about it being perfect. And that's the real dilemma, I think, for accountants, is going to be the perceived value of the good and A. It's good enough if I can do it in 15 seconds and it's good enough, I'm going to do that for cheaper. The 15 second option, that's going to be the the real dilemma. I think accountants are going to face is how do you convince somebody that you're more valuable when you're competing against something that costs 20 bucks and doesn't 15 seconds.

Blake Oliver: [01:11:37] Well, it's the same problem you have if you're competing against TurboTax. David, you you You you complained about this. You were getting ready to, you know, you were filling out your organizer that you got from your accountant, and you sent me a screenshot of the organizer, which is basically, it looks like somebody left the screen that your tax preparer has to fill in on their tax software, and they're asking you to put in the numbers, I think.

David Leary: [01:12:04] Safes and safes. If you already have a paper organizer safe, send an app. Right. Let's the firm. When they get safe, send scan that paper organizer and turn it into an electronic organizer. But I think they took their their internal working doc, scan those and put in the organizer. And so I'm, I'm literally filling out an internal working doc that the accountants team would probably be using. I just think it's crazy.

Blake Oliver: [01:12:31] So boring. Accountant in the chat said which instructor was the most boring? And so I asked Claude and it said I don't have data that directly measures how boring an instructor might be. Determining if someone is boring would be subjective and would require metrics like user engagement data, detailed feedback, or specific ratings about presentation style. And it goes on and on. And it suggests ways to figure this out. And I said, do your best. And so the most potentially boring instructors are, I guess, Laura Redmond. See, that can't be that can't be right. Because actually her course is, uh, very you know what I think it is?

David Leary: [01:13:12] It's too new.

Blake Oliver: [01:13:13] It's too new. It's like brand new. Yeah, but it actually, if we know for a fact that her course on QuickBooks has one of the highest registration rates of a course in a short amount of time. So that's pointing out one of the limitations, although I didn't tell it how to figure out what boring is, I it warned me and I said do your best.

David Leary: [01:13:32] We track ratings and reviews, but you didn't provide that data in.

Blake Oliver: [01:13:37] Well, I have ratings. I could say what are the lowest rated courses?

David Leary: [01:13:43] And then you could turn that into based on that. Who's the lowest rated authors?

Blake Oliver: [01:13:47] I guess you could try to figure that out and to avoid, uh, to avoid any embarrassment in case one of our courses is the lowest. I'm not going to show that to our audience. You can do your own research. Um, so there's been some chat going on, chatter here in the live stream about quality. And this is a pushback I always get when I show I because it is imperfect, like we said, a task that takes an hour. It's only 50% accurate. Uh, tasks that take seconds. It's now really, really accurate. And so it's not really a question of perfection. It's like with self-driving cars, you might never get to 100%, but at what point do you get to be better than a human? And so this is something that we should care about as accountants, and we should not strive for perfection. And we actually are taught not to strive for perfection. It's one of the principles of accounting. To paraphrase it, it's that the cost of the information, the cost to obtain the information, should not exceed the value of the information obtained. The cost to obtain the information should not exceed the value of the information that you obtain.

Blake Oliver: [01:15:02] So it's the same thing with this. You know the if you can do work that a human used to do, and you can get it to very close to human accuracy, you know, you have to decide what is that threshold, what is that materiality. We also have that concept in accounting. You don't have to be 100%. You can just be within the threshold of materiality and say good enough. And so is that legal? Well it depends. There's a lot of gray areas, right? Like I said, you know, the IRS doesn't really seem to care if you code every transaction correctly on a schedule C, because what they're really looking at is the net income, as long as the deductions are legitimate. It does it matter if an expense is in one category versus another category? Most a lot of tax practice practitioners would say, I don't really care all that much, and they don't focus on it, because if you try to be perfect, you're going to be way too expensive. And the client doesn't want perfect. They want.

David Leary: [01:16:16] It stuff done faster.

Blake Oliver: [01:16:17] They want it done, and they want it done to a standard that doesn't expose them to too much liability. And what that is, is very individualized, both for the professional and for the client. That's how I look at it.

David Leary: [01:16:34] And it's probably going to be a while before we have any data like that, where there's actually a natural experiment of, hey, these, these 200 people all used AI to create to submit their taxes and, you know, give it a five year run and they didn't get audited or whatever the percentage is versus humans that have submitted returns for five years, that that study. There hasn't been a natural experiment like that to figure that out.

Blake Oliver: [01:16:58] If you look at the quality of human tax preparers, the quality of tax returns that are prepared by people, uh, the range and quality is enormous. And you know how you know it, because you see all these tax pros complaining about how bad the returns are that they're getting, that they have to fix.

David Leary: [01:17:15] In the mid 90s, there was a lot of news articles and magazine articles like that where they would give a the tax information to 5 or 6 different tax professionals and then to TurboTax or H&R block, and then they would get everybody would have different results like nobody's numbers were ever identical, which is argument about the What is the quality like? What is the right answer? Well.

Blake Oliver: [01:17:36] The way you would measure that is you would look at the the liability calculation, like what is the tax due at the end. How close are they. Does the I calculate the correct tax due within a materiality threshold that is acceptable. And if it does then you should probably use the I because it's a lot cheaper. And if you're willing to like play, play, uh, you know, push the envelope further, then you're more willing to use it when it's less accurate. Uh, it.

David Leary: [01:18:05] Goes back to that answer, right? It depends. Right.

Blake Oliver: [01:18:08] Yeah.

David Leary: [01:18:09] Ultimately.

Blake Oliver: [01:18:09] So, you know, don't tie to the penny, guys. I know it's like there's this urge. You gotta tie everything to the penny. You really want that? But that's not what your clients want. And that's not what the system needs. Uh, it took me a long time to get over that, but I'm still guilty of it. I still like to have systems that tie out perfectly and beautifully, but I know in the back of my mind I know I'm wasting my time. I shouldn't be doing it.

David Leary: [01:18:38] Chasing pennies.

Blake Oliver: [01:18:38] Yeah, yeah. So I can't wait until the I will do most of our bookkeeping and just show me the exceptions. That's going to be really awesome when that happens, and that's going to happen in the next few years. I'm convinced it's.

David Leary: [01:18:50] Got to start with expensive cards first. Just the expenses going through. It's got to start with that before it does the rest of the accounting.

Blake Oliver: [01:18:55] It's going to do the simple tax returns too. I mean, I can't that'll be great because that will eventually make tax filing essentially free for taxpayers. Whether or not it's the IRS with the system or it's TurboTax, because the more efficient the system gets, the more you know, the cheaper the competitors will show up that do tax returns with I for a fraction of the cost of TurboTax. And that will force competition and the price will go way down.

David Leary: [01:19:20] Yeah, you're right, they're probably TurboTax is probably going to face the most competition it has in the next 5 to 6 years.

Blake Oliver: [01:19:27] And they're in trouble, I think because I just I just did our partnership return on TurboTax business. That interface has not changed since last year. And it's it's not great. I mean, it's I'm I'm filling out like a little survey about income and I'm basically putting income numbers into different buckets. And then I do the same thing with expenses. And I fill out this form and then it generates the return. But it doesn't help me tie out the schedule L or M or whatever. I have to do that myself. And I'm just sitting there thinking, what is Intuit doing? There? And somebody's going to figure out how to make that work and to, to handle these, like, I mean, you know, we have a small partnership there and it's just cash basis. And so tying out those schedules shouldn't be that difficult. I think I could figure it out. And because it took me like an hour And, uh, we're at the point where I can do an hour long task at 50% accuracy. But in what? In how many months or years will it be able to do that with 90% accuracy or 99%? And then at what point is it just doing everything? It's fascinating. Well, David, we've gone way over today, but, uh, I've really enjoyed it. Yeah. Thanks, everyone.

David Leary: [01:20:51] This episode in 15 seconds at an hour and 20 minutes.

Blake Oliver: [01:20:54] Well, that's one thing that I really sucks at right now is is generating podcast episodes. Uh, we tried doing it with Google's Notebook LM, but.

David Leary: [01:21:05] Every episode just sounds the same.

Blake Oliver: [01:21:07] Every episode sounds the same. Yeah. Although maybe people complain about us that way, too. We'll see. Uh, don't forget, you can earn free continuing professional education for having sat through this today. You deserve it. And you can go get CP for the past episodes that you have listened to as well of our show, as well as All my Fraud. Go to earmarked app, create your free account, earn a free CPE every week and if you want to support the work that we are doing, sign up for $150 a year. We're going to have that go up, so sign up before the price increase comes. Um, and what else?

David Leary: [01:21:53] I just blanked out. All right. Oh, well, the accountant too. We are having an accountant two viewing party. I'll put the link into the.

Blake Oliver: [01:22:00] Movie The Accountant two, the sequel to The Accountant. We are going to. We have rented a movie theater in Tempe, Arizona, and we are inviting accountants from all over the Phoenix area, perhaps beyond. If you would like to come to join us at this event, and you can watch it on the day it comes out, it's going to be fun. We're going to pack that theater. So David is going to find the registration link. But you know, the best way to get it actually is to download the earmark app and go to the events.

David Leary: [01:22:31] Teased in there. Go to the.

Blake Oliver: [01:22:32] It's in the events section at the top of the app. You can click there and register.

David Leary: [01:22:36] I'm going to comment now. I teased it. Now they have to go to the app instead and delete the comment.

Blake Oliver: [01:22:40] So awesome! Thanks everyone who joined us today. Hope to see you around soon. Send your messages to the The Accounting Podcast at Earmark Me. We are the accounting podcast at Earmarks Me and connect with me on LinkedIn. I am also on X.

David Leary: [01:22:59] And thanks for the reviews. Some people after last week came in and wrote us a couple five star reviews, so we do appreciate the support. Continue keeping on keeping on. We do love the five star reviews. We'll have to read those maybe next week.

Blake Oliver: [01:23:11] Next week. All right everyone, thanks a lot. See you around soon. Bye.