Behind on AI Adoption? Because the Second Best Time to Start is Today!

Ted McRae - Certified Prompt Specialist will be joining Matt and Dan to overview strategies accounting professionals can utilize in their practice to incorporate AI.

QB Power Hour is a free, biweekly webinar series for accountants, ProAdvisors, CPAs, bookkeepers and QuickBooks consultants presented by Dan DeLong and Matthew Fulton who are very passionate about the industry, QuickBooks and apps that integrate with QuickBooks.

Earn CPE through Earmark: https://bit.ly/QBPHCPE 

Watch or listen to all of the QB Power Hours at https://www.qbpowerhour.com/blog

Register for upcoming webinars at https://www.qbpowerhour.com/

00:00 Introduction and Welcome
01:06 AI in the Accounting Industry
06:06 Housekeeping and Session Details
15:45 Understanding Large Language Models (LLMs)
24:10 Practical Applications of AI in Accounting
32:46 The Cost of Private AI Environments
33:10 Handling Personal Identifiable Information
33:44 AI in Business: Practical Examples
34:21 Different Types of AI
35:24 Generative AI: Creating Content
37:17 Machine Learning: Continuous Improvement
38:23 The Future of AI: Risks and Opportunities
43:33 AI Hallucinations: Understanding the Concept
48:59 Effective Prompting Techniques
51:42 Upcoming AI Cohort and Final Thoughts

Transcript

Dan DeLong: 0:32

Welcome everybody to another QB Power Hour today. Of course, clockwork. As we go live, my, my machine, my other machine restart and me if hear if you hear anything, they tear apart the RV park that we're, so there's Bobcat heavy machinery behind me, so I apologize for that because that's, I can't move, a construction job. Unfortunately, I don't have that much full inmate in the park here. But Matthew, how are you today,

Matthew Fulton: 1:12

Dan? I'm doing good. But first real quickly, luckily AI has done such a great job of doing noise reduction. We don't hear any of that stuff. Oh, cool. It's good seeing you recently out in Florida and I'm excited for this conversation today, Ted. I'm glad you're here with us. We were pre-talk a little bit earlier and I think this is gonna be a, fun episode for sure.

Dan DeLong: 1:37

Yeah, we you guys. We always, this particular topic of AI in general has always has a lot of unknowns and, I don't know, maybe fear, scared of the machines type of thing. And that's, what we wanna try to unpack here today because as Ted was saying right before here, it's pretty amazing how little, the adoption of AI in the accounting industry has has occurred. Ted what, were you thinking about that?

Ted McRae: 2:20

I was just, for a. Industry that has so much technology. I can't think of any industry besides, actually, I can't think of any industry that has so many dedicated apps and technology and all this stuff. Holy smokes there. I, there's so much technology to learn. I cannot believe that, that something like this is that what people are afraid to, get into? And it's so underutilized, even if a lot of people get into it and start using it, it's so underutilized for what it can do. It just amazes me. And I think a lot of it, obviously we all have our, real lives that we go into, but it might just be fear and that's and I can tell a story later about my fear doing it and what stopped me. But I. That's all we were talking about really is just how underutilized it is, I believe, in this industry.

Dan DeLong: 3:24

Yeah. And the subtitle of this particular episode is because the second best time to start is today, like that when's the best day to plant a tree was 20 years ago, but when's the second best time it's today? So this is really what we're focus on here today is about a way to, to jump in to the to the end of the AI pool, without losing your float, my name.

Matthew Fulton: 4:13

Hello everybody. Matthew Fulton here with Parkway Business Solutions and the Facebook group QB Community Live. Really, again you're talking about going backwards and planting things, and I just, I'm gonna use a quick second to say, I just imagine this AI image of Bill and Ted going into their phone booth, going backwards in time and planting trees for all of us would've been great, but yeah, that's me. C Community Live Parkway Business Solutions.

Dan DeLong: 4:39

Alright, and join me again, not with Maker code, but but as a certified engineer specialist today.

Ted McRae: 4:48

What did you call me yesterday? I wanted, yeah, go ahead. I wanted

Dan DeLong: 4:50

to call you a

Ted McRae: 4:51

ologist. I'm a ologist, not a proctologist. Thank you. No, yeah. And I'm a Did you find

Dan DeLong: 5:00

The the latex gloves or No?

Ted McRae: 5:02

No I forgot the latex gloves. They're actually, they were all too small for my hands, no, and I'm a self-proclaimed propeller head. Anything, technology, app-wise, everything. I just completely love anything that's gonna build efficiencies, and that's really how I've tried to build my career is how do we build efficiencies for people that, especially folks that are trying to do this whole value-based pricing, right? So efficiency is key in our industry. The faster you can do something the better it is, as long as it's accurate, right? Because speed is important, but speed without accuracy is horrible. So how can we be more efficient but still contain and keep that accuracy?

Dan DeLong: 5:49

Yeah, there was one of the values that Intuit was, I can't remember what the value was but it was work fast and and they ended up changing that value to work smart, work fast. Because you can do dumb things fast.

04.29.25_-_behind_on_ai_adoption__29_Apr_2025_restream: 6:08

Yeah.

Dan DeLong: 6:10

And that's how AI is fitting into this picture. It the speed of which this technology is coming in and really disrupting a lot of things. I think that's part of the fear, I think that of people feeling like it's gonna leave them behind, with, with ai. So let's, before we get into meat, let's do a little housekeeping details about the QB Power Hour. It's every other Tuesday at noon Eastern. It's eligible for CPE Maker Hub. Used to be our CPE sponsor. So appreciate what you did there before us Ted, but now we're, using earmark for that. So after the session, about five days later, there will be a. Course to take over a earmark to get the CPE credit. And I'm sure Ted is appreciative that he doesn't have to do all that. It was

Ted McRae: 7:11

a Dan, it was a good transition period. Yeah, it was a good, it was good helping you out in the transition period, but yes, it was a very a lot of work to create 88 CPE certificates and making sure that everyone was on, was there on time and stuff. Exactly.

Dan DeLong: 7:31

We do have an earmarked channel for QB Power Hour. You can always check the, link out there. If you have specific questions about the topic that we're being talking here today, please put them in the q and a. You should see a little q and a button at the bottom of the webinar. That helps us keep track of those questions because as all of the AI bots that are joining us are announcing that in the comments. Those comments just seem to go by and scroll by really quickly. So appreciate if you have specific questions, please put them in the q and a so we can make sure, we can either address them live or follow up. And but if you have just general comments, please put them in the chat. And then we also have the links for the handouts, which I was about to socialize, but again, my computer restarted. Hopefully we've got those in the social media, live stream as well. We also do have a QB power hour store. So you two can have a pink hat like me. I need to order one all. So we're gonna start off with our first polling question as a setting the stage and setting the, The foundation of what we're gonna talk about here as a self-assessment. What is your comfort level with ai? Anything from novice where it scares me to, I'm an AI ninja fully embracing this technology and incorporating the advanced features. And Ted can you talk about your journey with, ai? Like where how you, became a ologist.

Ted McRae: 9:12

A ologist? Yeah. So I started, you know what the funny thing is, so I was hearing about this AI thing, and this was way back when, a few years ago, when AI actually was only up until, I think it was September of 2021. So anything new we were in 2022 or the late ends of 2021 when I started using it. So anything new I. It wouldn't, it had no idea. You couldn't ask any who won the World Series or what are the a You can't do anything like that. And so when I first started out, and probably there's a lot of people like novice, it scares me. A lot of people or myself in particular thought there was a specific like magic to asking the prompts. So I searched all over the internet when I first started, using chat GPT for chat, GPT prompts what are the, how do I do prompts, how do I do this? And what I soon found out after searching for weeks and weeks is that there are no speci anyone can give you, oh, here are 10 prompts that you should use, everyone should use in accounting. But realistically I. That's not really what you should be doing in chat GPT. And there's really no way, there's no specific here's exactly how you should do it. There's a framework. And so that's what I didn't really realize until I became a, ologist, is that it's dirty work. No, I'm kidding. It there, it's not about

Dan DeLong: 10:57

digital.

Ted McRae: 10:58

Yeah, it's digital work. It's really about where you wanna end up what you, want. And you gotta think of it as if I were to ask Matthew Hey, can you help me create a webinar that I wanna do next month? I wouldn't just tell Matt that'cause he would be like, yeah, sure here, and he would just tell me everything that he thinks that I should do. But if, I tell him, Hey Matt, I'm doing a webinar next month for a specific type of person and it's at this conference. Here's all the research on the conference. And then I do even more things like that. Now Matt has a framework to work, or Matthew has a framework to work with where we can create awesome content for that. But if I don't give him all that basic stuff, like having a conversation with him, he's never gonna know. So that's really how I started out with it way, way back when. And I've kept up with it and I hate to say it, but I use it probably 50, 60 times a day.

Dan DeLong: 12:06

Yeah. It's now it's it's part of your, almost your everyday thing. The whole idea of the. Dyslexia where, you know, if you don't have your phone to do basic math skills, you just Yeah. Purge that out. But I've got a calculator on my phone. So we're gonna be doing Ted and I are gonna be doing a four part weekly cohort series over a school bookkeeping where we're gonna be talking and unpacking a lot of these things from the Mr. Rogers level of ai. And we'll, we maybe go up to Sesame Street or maybe electric company if we're, lucky.

Matthew Fulton: 12:48

I'm sorry, Mr. Rogers is going to, the ologist is gonna be the first episode of this. Yeah,

Dan DeLong: 12:53

the French. He'll be happy taking off more than his shoes for that. But what we're gonna be doing there is we're gonna be talking, and today we're gonna be talking about AI and large language model statistics. We'll just really keep this, as a fireside chat type of thing where we'll talk about what we're, gonna be unpacking in the cohort series which will be called prompts for practices, crafting effective prompt for accountants where to go to register. And then we'll have a q and a session about that. So let's let's talk about, oh, Noah, did you share the results of the first poll? Because I thought that was really oh. Did I Just, so it looks like about 60, almost 70% are in the novice, the basic level of, of the com comfortability with with ai. So let's talk a little bit about, the, impact of, AI in, in small business as far as an understanding of how businesses are using AI and the large language models. I really like this this slide that you created here, Ted, about.'cause that's part of the thing for me was what is clawed, what is all of these other, the Gemini or Bard, I think they called it originally and now it's got a new name to it. It's tr trying to keep all the players straight. But talk a little bit about this this Ted

Ted McRae: 14:32

Yeah, so there's a few things here that really strike me as, especially in this industry, chat, EPT is like the Intuit in cases of this whole ai large language models. They were pretty much the first, the Kleenex, right? Yeah. They're the Kleenex. But the thing is they are, in my opinion the, most reliable, I never stray. Personally for me, I never stray from chat GPT except for perplexity. And so there's different types of AI that people need to to, actually keep in mind. There's, AI that can generate speech for you. 11, 11 labs, which I can give it a, I can give it a script and I can either choose my, the audio person or I can upload snippets of myself and create audio. There's like Gemini, there's Microsoft. So it all depends on what you're doing it for. So like Google Gemini, if you put it into the Google suite of tools works perfect if that's what you're gonna use it for. You can also attach chat, GPT, to, to Google as well. But it, so it all really depends on what you're using it for, but. The thing is really quickly here is that some of all these other ones while you might like to use'em and they're fine, they may not have the same technology as like just a regular old chat GPT. So it really matters what you wanna use it for.

Matthew Fulton: 16:18

Ted, can I just real quickly have you step back just a little bit and can you give us a quick definition of what a large language model is? What is an LLM and why is that important in what we're talking about?

Ted McRae: 16:31

Yeah so a large language model takes context from a person, right? And then will spit out what it an answer a real person would. So it takes your question, spits it out and, contextualizes it. Like, a per, like you would in a regular conversation. Large language models are huge data sets. And I could really talk about like the prompts, what it does. But Matthew, what it really does is it takes your sentence and then it breaks it out into little chunks. All right. So it's like for instance, the bill is due on the on May 1st. So we'll take that and it actually chunks that out, the bill and then it says is due, and then it has the date. So that, so it separates all that stuff out. Yeah. So it understands. Then puts it back together so it can understand what you're talking about, the date, actually referring to a date in the future that this document. So almost like Matthew, do you remember when we were, not you and I together, but when you were in grade school and you had to cut a sentence apart and find out where, what is the adjective, what is the noun, what is the verb? And all of that stuff to, to show the context of a sentence. So large language models do the exact same thing, but in a more granular process. So they, they chunk that out and right now I have old man brain, so I can't remember the name of the, of what it's called. Oh, tokens. They chunk it out into tokens. Okay. And then each token, they analyze that token. And so what large language models have, it's just a huge data set, and now it goes and it returns it to you like a normal person would.

Matthew Fulton: 18:27

So another way to kinda say it'cause Grace asked a good question earlier of what exactly can AI do in the ac in the accounting to assist in our work? That's obviously a huge question, but a quick answer way. I feel when we're talking about LLMs large language models, while we the, word language is in there, it's, it is, it's truly large data sets. So shameless plug here, maker's Hub has taken the ability to, when you scan in a whole bunch of different documents in different formats, there's this large database to try to understand where valid or important information is on a sheet Yep. Can extract it and then understand the words that are on there to place it appropriately into the right fields for your accounting platform. Same thing, where there's a lot of companies working on like check images now, trying to be able to pull that type of stuff. So the more data you have, the more it can reference to try to give you a better result. Is the intent of it all, is that a good way to. Summarize it. Yeah.

Ted McRae: 19:27

Yeah. And like the really cool thing that you can do with like chat GPT now and all the other ones is, and this is what I did, I created an accountant a ai, so it's just a, all he is his, name is Milo. And Milo knows a lot about accounting because every little thing that I see on LinkedIn, like people will do infographics and stuff like that, I just copy and paste it into Milo. So like, how to analyze the spread, how to analyze a balance sheet versus a cash flow summary, how to ba versus a, a p and l and all of these other things, income statement. And so I just upload as much as information on accounting that I could find into Milo. I also have him look at d different websites. Then now I can do things in a, in accounting that maybe we couldn't do before, but I don't use that as the true source. So people will be like, I don't trust it. You shouldn't, that's the point. But it helps you get to where you want to be faster.

Dan DeLong: 20:41

No there's a lot of words here about the statistics usage and adoption. Any, anything you wanna encapsulate, take away of all this? Of course you can download the slides and take a closer look at some of these statistics,

Ted McRae: 20:56

Ted. Yeah. I think that what people need to realize on this is that the businesses don't think of it as, it's gonna take my job away. Think of it, especially like I said, being the propeller head that I am, what can we in this industry use it for? And we all know that there is a, there is a, gap between like how many hours we can work and how much we, especially if you're if you ba bill by the hour, you can only bill so many hours. But if you're, but if you're value-based, the faster we can do things, the more hours we can bill. So if I can take a, if I can take a balance sheet and ha have the first pass of analyzing it, go through a AI before I look at it, it's almost like Dan, like what you say, having an intern look through my stuff. I wouldn't just send my interns report to my client, but the intern's gonna look in and look for anomalies for me and things that are like red flags. Then I can review them with the data, with the mindset of an accountant to put my spin on it. And then I could even put it in a and say, okay, chat GPT create the the summary report for my client for this. But that doesn't mean I'm gonna send that straight to my client. I still need to review. Now I take that summary report and build my report off of it. Yeah, but how much time does that save us? It's astronomical when you use it as just the first set of eyes and create, instead of staring at a blank piece of paper, I always tell Chad PT to write down a summary of my ideas. And then I don't start with a blank piece of paper. I just save myself three hours probably. And then the other thing, 49% of companies are currently using it and 30% intend to use it in the future. So for all of you, I was amazed at how many people said it scares me, is what'd you say? 60% of people said they're in the first two buckets.

04.29.25_-_behind_on_ai_adoption__29_Apr_2025_restream: 23:11

Yeah.

Ted McRae: 23:12

We're like five years almost into chat. GPT artificial intelligence right now in large language models. It's time, like you said, Dan, it's time to start thinking about it. it's,

Dan DeLong: 23:25

It's the new disruptor of, the day. You look back at all the different things that dis disrupted the accounting industry. Cloud accounting is the the latest up until this point, Excel, right? You can go back to all of these things where you think about, I'm sorry. Oh my God, it's gonna change. It's gonna change the it's gonna take my job, right? But it's not, it's just another tool that can assist and to your point, provide you more leverage to do the things that you wanna do. Faster, right? Think about

Ted McRae: 24:04

it. They're in Dan, they talk about'EM as ERA technology eras, right? Yeah. So we had the, really, the first era was the www era where we connected everyone online. And that was a game changer for all of us. And then there was the, then there was the cloud era where we actually started sharing data with each other. Then there was the mobile era where now we're connected to our mobile phones, and now we're in the AI first era. And so this is going to, it's not going away, and this is where it's really gonna be. Yeah. Thank you Matthew. You're right.

Dan DeLong: 24:41

Thank. So this one what, accountants are, using AI for, right? You've got the basic level where it's creating some content. So starting you off with that three hours of savings, like you were just talking about, instead of staring at a blank document. Get it to start with that rough draft and then the intermediate level. So you wanna talk us through this these, different levels?

Ted McRae: 25:08

Yeah. So if you think about it at the basic level, and it seems like probably 70% I don't mean to, if it's less, that's fine, but it seems like 70% of the people on here are at the very basic level using it to or look at an email. Is it right? I want to create an email for this and that, or I need to create a blog. Which people are probably not using it the right way, even if they're using it in the basic level. Now the, intermediate level, we're allowing large language models or AI chat, GPT, whatever you wanted to. Whatever you want to call it. We're allowing it to take it up.'cause we're starting to trust it more. So we're allowing it to take us to a different step where, alright, I want to create maybe Dema, generate a demand letter, but I'm still not using it to its full potential. Or I'm generating a letter to a client that I need to fire or I'm maybe creating small easy SOPs, little things like that. So that's more of an intermediate level. And then maybe you're going in and if you know what A GPT is, we're gonna talk about it extensively in in our school of bookkeeping training that we're doing, but creating GPTs or using it to ask questions. About a product. How do I create a formula, an Excel formula that sums all of this really cool stuff and it just spews it out for us and shows us how to do it? That's pretty intermediate, right? Or how do I create a button in HTML? That's pretty, easy. But you can say, how do I create a button that does, that, goes to this specific link with these colors, and when I hover over it, it changes colors and when my mouse clicks on it, it shakes. So you could have it do more than just create a button in h TM l you can have it do in single thing and create the webpage for you that the button goes on. So, intermediate is just really basic. A tiny bit above the basic. We're getting into knowing how to create somewhat crafty prompts. The advanced level we've created Milo, like I have. Who is a, an accountant in my little chat, GPT world. And Milo knows a lot about accounting, on a pretty awesome high level because I start, I feed in every single news thing that I get about it. I have him look at websites all sorts of stuff. So anything I want to ask Milo or analyze or Milo, what, you know what it's, today is the end of April. You're as an accountant. What are you doing today? And Milo will tell me what the average accountant is doing. So now me as a marketer knows what to actually talk to the accountants about. So now we're getting more at an advanced level, right? Milo can analyze reports. I can have Milo generate demand letters. I can have Milo actually generate them based on a level of friendliness, aggressiveness, and how late it is. Milo, I can say Milo generated demand letter from my client with an aggressiveness of seven'cause he already knows of one to out of seven, a friendliness of two, and the, it was due on this date. And here's, and I upload the invoice and he'll generate a demand letter that is more than just you need to pay this is like the third or fourth time. And then you can also do some pretty advanced. Or semi advanced coding at that level where maybe you're even, you're doing more than J you're extracting databases and you're doing really cool things like that because you can ha you can do JavaScripts, you can do all sorts of stuff within within chat GPT. And in fact, my daughter's a a software engineer and they are not allowed, they have a program that scans their they use artificial intelligence to scan their backend code for minute errors. They're not allowed to use large language models to create code, but they use it to scan the code for errors. And then at the ninja level you're really, you're using it to. Think about it this way, if you could create a GPT and we, like I said, we'll talk about it, but A GPT is memorized information that you put in a little in a, little area of chat. GPT. Think about this. If I have a GPT where I can upload all of my SOPs, I can upload all of my company, like hiring information and all that stuff. So when I have a new employee come on board that has the questions on standard operating procedures, they can just go into my GPT that I did. And I said, so if I, if we have a client that refuses to give us their bank account information for QuickBooks Online, what do I do? And it just spits out all that information that employees just saved. A a lot of time having to go and ask people around the office. What they should do in specific areas of the business. So now I'm like, all right, how do I do X, Y, and z? And the chat GPT just comes out and tells me how to do it based on my company.

Dan DeLong: 31:02

Yeah. Good question here in the comments how secure and I think that was part of what you were alluding to, where you don't want what you are training it on to be training everybody else. Which gives that specialness, I guess to Milo over sharing it with, with everyone in the

Ted McRae: 31:29

world, right? Yep. Yep.'cause Milo is basically an extension of what Ted wants. So we, if we each had our own Milos, it would have, its, it would have its own characteristic of the company that we work for. I can have Milo be Ja, Jason stats. And react to me like Jason Stats and I actually have a Jason Stat bot if I ever want to create a video that's halfway funny I put it through my Jason Stats script builder and it builds that script out. So I can have I can have chat GPT review, YouTube videos or, things like that. I actually had chat. GPTI have a Ron Baker chat, GPT for value-based pricing, believe it or not. So it's, there's no end to what you can do.

Dan DeLong: 32:24

And how does someone keep that from being leaked? Is be like how do you,

Ted McRae: 32:34

your chat, GPTs, you can either share publicly, or keep them private to yourself. Now if somebody gets it I really don't care because I don't have a, I don't have any, I don't put social security numbers in there and, stuff like that, right? Which I still I still would never do, I wouldn't upload a report and say, analyze all of my employees, give me a social security list alpha or numerically from smallest to large. I wouldn't do anything like that. I would think of chat. GPT. If you were in a doctor's office and had to follow HIPAA rules, would you put, what would you put into a computer system outside the doctor's office? I.

Dan DeLong: 33:17

It's a good good rule of thumb.

Matthew Fulton: 33:19

There are certain service providers out there that allow you to spin these up into your own closed environment, but it becomes like the cost of it is exponentially higher. So that would be something you'd research. Once you've really developed something and you've got a tool if you really needed that private information there. The key thing really is personal identifiable information. You don't want to specific, you don't wanna have that in there. You don't want the liability of that would be the key, I'd say.

Ted McRae: 33:52

Yeah, I would, I I. I would think of it, Matthew, just like any other thing, are you gonna, are you gonna put like your employee, or not your employees, but all your client's data up in an unsecured Google Drive that maybe has a shared link to somebody that you don't like, a deep folder that might be shared with somebody. You wouldn't do it. And I wouldn't do it on here either. Agree. But the thing is realistically we shouldn't even be using it for that. What can we analyze with people's personal data? Like a company, I can understand a company, alright, I'm uploading a balance sheet for a company. Okay. It's just nu it's really just numbers, right? It's not like I can take that now and infiltrate the company and pretend like I'm the CEO and access his credit card information. For the most part. That's not what you're uploading anyway.

Matthew Fulton: 34:48

Good point.

Dan DeLong: 34:51

When I asked AI. How many different types of AI are there? It spit out, I think 16 different ones, of how AI is used. So we want to just talk about the different types of ai. So we've got traditional generative and, machine learning. So Ted, how would you, qualify or quantify the different type of AI technologies?

Ted McRae: 35:21

There's traditional, which is like the chat GPTs and the copilots and things of that nature. They're just like let's talk about it this way. Traditional AI was like, free chat, GPT. There's, they have limited scope. It doesn't learn. So that's just like what traditional back in the day was. Now, if we talk about, and then you can talk about traditional AI is almost like chat GPT right now. But again, without the subscription, right? Because it's not gonna learn from what you do. Then there's generative ai. And generative AI is, I use it all the time. I can you and I just do this with chat. CPTI need an icon for, so my Matthew, you know the arch out on the Channel Islands, right? That, yep. The little, okay. So I wanted to create an icon for the arch, okay. For Ventura. And so I just uploaded a picture and I said, here's the company I want to generate this icon for. And I want the theme to be coastal, so use coastal colors. And so it generated a logo, and I shouldn't say icon, a logo. I. For that. So I uploaded just a regular real picture. So generative AI creates content, creates images it can create all all sorts of stuff for, you. So generative, you give it an idea and it's really creating like an image for, you or I play the guitar. I can have it create a song for me based on an, accountant. I actually, one, one time I was doing a a webinar for AI and I had it create a song for accountants around April 15th. About how, about like how crappy their clients are when they get right around that time asking for things or not giving them the right paperwork that they want and stuff like that. And it wrote out the song, it go, gave me all the chords. It told me how to strum it and everything. And I sang that song and I sang that and played that song on one of my webinars.

Dan DeLong: 37:49

And then the machine learning and then machine learning. Yeah,

Ted McRae: 37:53

data. Data. It's like my milo, right? My Milo learns. I have to tell it to learn. But my milo also learns when I just type things into him as well. He just keeps learning. I can hard chord, hard code learning into Milo, or as I'm typing, Milo learns as well.

Dan DeLong: 38:16

So this would be just to put it in a QuickBooks, speak. This would be like the bank feeds as they are trying to homogenize all these transactions. And it will start to recommend. The things that you wanna categorize it to, or something in like Maker Tub where you put in a, purchase order or a document and it's different for one over the other. Where it starts to learn the sales

Ted McRae: 38:45

tax. Sales tax should go to the sales tax chart of account. And not just not just to cost of goods sold or whatever you wanna put it. To put it to. Yeah.

Dan DeLong: 38:56

Which I think this leads to a question that was posted in here. What is the possibility of AI basically becoming self-aware very or, becoming corrupted based on the things that it learns?

Matthew Fulton: 39:11

Skynet is real. Yeah.

Ted McRae: 39:14

Yeah. You know what's the funny thing is I was listening to a podcast and they were talking about. These guys that are doing the large learning models, didn't they watch all the movies we watched growing? I know, right? Come on. Literally out whole lot of movies saying this is a bad idea. Yeah, and guess what? They know it too. They're like, yeah, we know that this is a possibility, but they're, but obviously if it's we're in a race, we're in an AI race right now, is it going to be the United States? Is it gonna be China? Who's it gonna be? That is, is the AI dominant country and it really matters.'cause ai, if you think about ai, what we do just in accounting, think about what we're gonna use AI for in defense. In the future, what we're gonna use AI for in all sorts of like negotiations and stuff like that. So

Matthew Fulton: 40:16

yesterday I was listening to a podcast as well. And then they were specifically talking about ai. They had mentioned that as an example, having AI research, AI itself and how to do certain things. This kinda leads us towards the different levels or layers of where it could become a problem. But currently. AI can complete a human task that's, say, that's like a, one hour task. It can complete it faster with a 50% accuracy. And the imp it is improving and doubling in time, like how much it can do every four months. So when they talk about artificial intelligence, they also talk about human artificial intelligence. In other words, artificial intelligence becoming as intelligent as humans instead of, I'm not gonna get the term right, but like super intelligent the, next level above. And they compared it to a toddler versus an adult. So a toddler trying to teach an adult something is not gonna happen very easily. The adult can even placate the toddler and make'em, oh, you're so cute. You're teaching us such wonderful things. I. So we have to be, the concern is when we go beyond human artificial intelligence, are we gonna actually understand what it's truly capable of or not? And is it gonna go Skynet? Nobody really knows, but we gotta, I don't think, I don't think

Ted McRae: 41:46

we'll have to, I don't think in our lifetime we're gonna have to worry about it. But it's, the fastest it's actually increasing in knowledge is amazing. I'm amazed every time I log in, what I can do with chat GPT every fricking time I log in and I do something. Not because it's actually changed, but because I just thought of a different way of using it and I'm like, holy smokes that. And then like, when you can do things like copy and paste images and it can read it. Then store that information. That's why I said, every time I see anything about accounting, infographic, whatever, as long as I, as long as it comes from a source that I trust, I just copy and paste it into Milo and have Milo learn it.

Matthew Fulton: 42:33

Do you ever talk to yours? Little tip I like to use is on my chat, GPT, you can hit the side and ask it. Can you explain how AI can be used in accounting today?

AI: 42:45

Absolutely. AI can be a game changer in accounting. It can automate repetitive tasks like data entry, invoice processing, and reconciliations, which saves time and reduces human error. It can also analyze large volumes of data quickly, helping identify trends, anomalies, and potential risks. AI powered tools can assist in forecasting, budgeting, and financial planning by providing more accurate predictions based on historical data.

Matthew Fulton: 43:14

Thank you. Plus, sorry to interrupt you, but can you tell me, does that mean AI accountants are gonna take over the world someday in the future?

AI: 43:22

No worries. It's unlikely that AI accountants will take over the world. While AI can handle many routine and data.

Ted McRae: 43:29

Yeah, it it's pretty, of course they would say

Dan DeLong: 43:33

no.

Ted McRae: 43:35

Hey so Linda Russell, how you doing? I, Hey,

Matthew Fulton: 43:39

Linda. Yeah.

Ted McRae: 43:40

Talked to you a few times before. So try using perplexity.ai. So Linda says, I'm using chat, GPT as my search engine, pro tip perplexity.ai. It was designed to be an AI search engine, and we'll give you better results than chat. GPT, just ex FYI.

Dan DeLong: 44:05

Yeah. And, I want to talk about this whole concept of AI hallucinations, which was really like the catalyst of this, talk that we're having, this course that we're creating because, we, we've all seen like the news of hey, AI took the bar exam. Or AI can't, couldn't do math or couldn't do simple math. Where AI hallucinates the truth. Sometimes like it could have just hallucinated that response that, you said when you, asked it, is it gonna take over the world? Maybe that's what you think that you want to hear. And this whole idea of of hallucinations isn't necessarily a flaw but more of a feature. Ted can you talk a little bit about the, concept of AI hallucinating and just making stuff up.

Ted McRae: 45:09

Yeah. What ai hallucinating nowadays is more of, I don't know if you've ever heard of, pep Cap. Problem exists between keyboard and chair. So honestly, AI hallucinations prior to 4.0, chat GPT was because it just didn't have the data right? And it just tried to make up whatever it is. Now, realistically, it's because you like what I said to you guys? Hey, let's before we got on here, I was telling Dan and Matthew, if I told them, Hey, let's all create a blog and we all typed into chat GPT create a blog about accounting and chat GPT, it would might, if we just said that it might give us all the exact same, roughly the same information. And it might not even know what we're talking about. Is it talking about like account chat, GPT taking over accounting? Is it talk? So AI hallucinations are a not so well crafted prompt now, except for like it used to be. It used to be like you would ask chat GPT how many Rs are in strawberry? And it would tell you two, it didn't matter how many times you told it. There were three, it would tell you two Mine at least now, and I can only go off of mine, says that there are three. So it is starting to learn, but go. But it is the, in most cases now it's the user.

Dan DeLong: 46:54

Yeah. And it's this George Costanza G here about it's not you, it's me. It's if you're not getting the outputs that you want. You really need to change the input that you're giving it.

Matthew Fulton: 47:09

I'm gonna disagree in one area specifically, which is when you're creating an image and you want words in there.

04.29.25_-_behind_on_ai_adoption__29_Apr_2025_restream: 47:14

Oh yeah.

Matthew Fulton: 47:15

Like you can type out exactly what you want it to say and it will still always spell it wrong every time.

Ted McRae: 47:25

I never, I always, but Matthew, you can correct the chat, g you can correct it to create the actual words in the image. If you tell it to focus on the words in the image.

Matthew Fulton: 47:40

I years sounds like it's got a better education. Mine, I think kindergarten or something. So maybe I don't know how to spell it. Could be. So it's probably my prompting.

Dan DeLong: 47:49

Yeah. Like I, I saw a post where somebody had searched Google just did a Google search for BS in the computer. And Gemini, the AI overview that came up actually said. There is, there are bees in the computer that have been there since the year 2000. And, of course that's the, poster child for, okay, this AI doesn't work at all because here it's saying that there are bees inside of your computer.

04.29.25_-_behind_on_ai_adoption__29_Apr_2025_restream: 48:23

Like honeybees.

Dan DeLong: 48:25

Yeah, like real bee, like buzzy bees. You can google it, you can google it now. And the AI overview does not say that.

04.29.25_-_behind_on_ai_adoption__29_Apr_2025_restream: 48:31

Okay.

Dan DeLong: 48:31

But, there what it was doing is it was searching the internet and it found this company that created a blog article about such, things. And it was an it, an fool joke, right? This one things where. It found that article and therefore provided you that as truth. And that's something that AI can't do yet, right? Like you cannot measure the value of truth, even though it found an answer for you. Yeah, there were bees in the computer. But in this case it was totally a joke and it can't figure out sarcasm or truth or things yet until you, unless you tell it. Now, apparently somebody told the AI overview that is not a true thing and it doesn't pro offer you that as true. That's funny. So let's talk a little bit about some prompt examples, right? Where they're ineffective and maybe effective, right? So in this example, make a report about accounting, how that is an ineffective prompt.

Ted McRae: 49:49

Yeah. So make a, this is almost like the, what we talked about the blog, right? We all, it's very vague. You have no idea who am I making the report for? What, is the report? Am I talking about accounting or am I talking about numbers? From from something that you want to give to me. So you're there, it lacks any context whatsoever. And can I just share something really quickly? I wanted to show you guys something. Matthew, this will show you, allow multiple, I'm just gonna replace the current share. Sure. Is that okay? Yep.

04.29.25_-_behind_on_ai_adoption__29_Apr_2025_restream: 50:27

I stop sharing.

Ted McRae: 50:28

So look at this. I said can you create an image of the bookkeeping department with QuickBooks online on the computer? And look at what it created for me. So accounting, a bookkeeping department actually spelt it ex. So look at how good it's getting now. QuickBooks, it's spelt wrong. I could, redo it now and I could say you spelled QuickBooks wrong up here, and, but it like small, like it says profit on loss, but it's still pretty darn good now, right? Holy smokes. It's, really, crazy. Go ahead and share again, guys. I'm sorry.

Matthew Fulton: 51:07

That's where, so I'm sure I've, other people seen these reels, right? Where they'll be like, show us a unicorn with great, beautiful wings and make it magical. No more magical. Make it extremely magical and they just keep building on top of it, on top of it, on top of it, on top of it until it's just absurd. That's some fun stuff you to do.

Dan DeLong: 51:26

Yeah. So when here we're talking about, being more effective in your prompt, create a one page summary report on the latest accounting trends for small businesses focusing on automation and AI tools. So just by adding more. Context to the input, you're going to get a, more specific, answer from the from the output, right? And then also

Ted McRae: 51:53

Dan, we were talking about basic and intermediate levels. That's more, that's an intermediate prompt right there, right? That's not even an advance prompt because there's still, it's contextually wrong still but it's better than the other one, honestly. Yeah. And that's what we're gonna talk about in our little prompt thing as well.

Dan DeLong: 52:14

Yeah. So in the month of July Ted and I are gonna be doing a four, four week cohort where each week we'll break down four different modules going from basic what is it to intermediate advance and drafting prompts. And each week we'll also have some homework. So because it's at the school. We'll have some homework, some assignments. There will be a cohort I can't say it, cohort only community where you can, put in questions and we can talk about them in a, weekly follow up and q and a session. We can do a group discussion about the prompts that are being created and the outputs. And Ed will be there as the subject matter expert and theologist of the group to be able to really dissect and and allow us to really unpack that prompt and, why maybe it gave us the the output that it is. And I'm gonna try something really silly and allow you to pay for it what you want, right? There is just a a small setup fee to make sure that it is, that, that you're able to attend and we can keep the lights on, but whatever you feel the value of this sort of thing would be is entirely up to you. If you think it's just worth the$97, that's great. That's fine to you. If you think it's worth more than that's fine as well. And hopefully at the end of it you'll find that it was worth what you paid for it and we'll be able to level set that. So there's a link there to be able to do that. So any any questions or con thoughts about what it is what it is that we talked about here today? Matthew, I know you were traveling the globe, before all of this. And what do you in your journey with adopting. Your, thoughts about this technology.

Matthew Fulton: 54:24

I, it's been extremely helpful. I would, I actually probably put myself in between the beginner and intermediate just because there are some more advanced functions I've used by accident. But it's not something where I was always using. It's consistently every single day. I've definitely been using it a lot more. I like being able to talk to it like I was just demonstrating and ask questions because I can research different things while I'm doing stuff around the house, have this conversation be learning at the same time. So I think that's pretty powerful. But I do also wanna help spread an important message for open ai. They've asked you to stop being so polite to the overlords of AI because it's costing'em a whole bunch of money. Every time you run those extra prompts by saying thank you and your response back to,

Ted McRae: 55:09

But Matthew, I always say please, just in case AI takes That's

Matthew Fulton: 55:13

exactly right. You can, you.

Ted McRae: 55:17

That way. I, that way I know that when they take over, they're like, Hey, that's the guy who always said, please let's not make him a slave. Let's make him like, run the slaves or whatever.

Matthew Fulton: 55:28

I love it.

Dan DeLong: 55:29

Yeah. There's some good questions in there. When are we doing this? So this is in July. We'll be having the the weekly sessions for content on Monday at noon Eastern. There'll be for an hour, maybe an hour and a half. We're baking in some extra time'cause I'm sure there'll be a lot to talk about. And then the follow ups are gonna be on Thursday for an hour at, 12, I'm sorry 2:00 PM Eastern Time. So whatever that is for for you. And they will be recorded. So if you can't make it for that session they'll be there for you, to review. And I'm sure we'll have an AI summary created from all from transcript because the course and, a lot of the content was created by ai. So we've we've actually leveraged AI to create a course on ai. So we're teaching humans how to use AI with ai, which is,

Ted McRae: 56:32

and you know what Dan, there's one person that, this is actually a pretty good comment says it might be possible that AI could replace the jobs, but for those who will update themselves and adopt related technology, AI can't replace it. Yeah. There's more likelihood that AI is gonna replace your job if you don't use it. Because you're gonna be obsolete. You're people want you're not gonna be able to do, so if you think about the competition in the industry, you're not gonna be, you're not gonna be able to do it as quick as a guy down the street that actually has adopted ai. So the more you can adopt this stuff, the better it is for your practice. You might be obsolete if you don't adopt it because somebody else did, and they're gonna do it better.

Matthew Fulton: 57:15

Not, it's what Uber did to independent taxi drivers.

Ted McRae: 57:18

Exactly. Yeah. They didn't update their technology and stuff like that. And now they're, who takes a taxi nowadays? I only take a taxi if it's more convenient, one sitting right there and I'm like, all right, I'll just take a taxi. Yeah. But then the taxi driver is always arguing with me that, oh, you don't have cash that you can pay. Exactly. Yeah. You know what? My wife let me have$150 for this trip. I'm not gonna spend it on.

Dan DeLong: 57:47

Yeah, the I think Joe Woodard said it best at the at the last scaling new Heights, which if you're going, you won't see me there.'cause unfortunately I won't be able to go, but you'll see Matthew. But he said AI's not gonna take your job. It's the people who adopt AI that are gonna be taking your job. Yep. So if you're not going to see, a robot, come into the space. It's the person that is adopting the is bringing that, Milo with them as part of their team. Yeah, a hundred percent. All right. We appreciate you joining us here today, Ted, and for those of you that, that might be considering joining the cohort, we look forward to continuing this conversation over at School Bookkeeping. And we really appreciate you all joining us here this week. And I'm glad that the construction stopped, as as I think there was some rain. So they went for lunch, I think. Nice. So we will see you next time on the QB Power Hour, which we'll, be actually talking about keyboard shortcuts. That's what we thought we were gonna be talking about today, but we, I didn't even know my schedule, you wanna check it out on the QB Power Hour. We have, the upcoming events and topics there. So we appreciate you joining us here today, and we'll see you next time on the QB Power. Have a great day, everyone.