90 Days to a Better Tax Season with AI

Rebecca (00:04): Okay, it’s October 15th and we want to talk today about what we should do between now and the end of the year to kind of prepare for this next tax season in this new AI world. So whether you’re trying to be a fully AI native firm, for next tax season or you just want to figure out how to incorporate some AI into next tax season, we’re gonna share what our thoughts are. about what you could do to prepare over these next few months.

Charlie (00:32): I mean, in my mind, we have about three months, right? I mean, I feel like, you know, January fifteenth, things start ramping up, whether you got ten ninety nines or you got Q4 payments or whatever. I think now is a great opportunity to spend some of this downtime, like really going heads down, and trying to figure out what do you want next year to look like? And, you know, I’m slightly biased, but in my mind, the best thing someone could do between now and January 15th is go deeper into AI and start learning more about AI, how to use it, getting more confident using it, getting more comfortable using it, and then really trying to figure out kind of how they can fit it into their existing workflows and their practice. And even potentially maybe uprooting some existing processes or practices and implementing AI and kind of doing things differently.

Rebecca (01:26): Yeah, And I think also that’s kind of internal and then external. We need to prepare ourselves for, you know, how we’re gonna interact with clients, the ones that stick around and still want a human involved in bookkeeping and tax work. how do we kind of get get ourselves acclimating, get ourselves used to them bringing us AI stuff? You know, we got a taste of that this year and probably during this extension season of people, giving you AI kind of things. so, how do we get acclimated so that it doesn’t feel like a shock, or uncomfortable when that happens and kind of put ourselves in their shoes and educate them a little bit on, what’s helpful for the tax preparer, what’s not helpful, and maybe even give them some tips on what they could do with AI that, might benefit them or help the tax season kinda go more smoothly.

Charlie (02:20): It’s interesting because I think a lot of people, like you said, did get that taste of clients using AI, clients responding with pretty obvious AI generated responses, clients sending pretty obvious AI generated work papers, documents. And you know, if we think that’s bad right now, or bad in the sense of it’s occurring more and more, I mean, just imagine what it’s gonna be like six months from now.

During actual tax season, when these models have improved over a period of six months, adoption has increased over a period of six months. I mean, the models right now are already really good at both tax and accounting tasks. and they’re just gonna get better. And people are gonna feel more comfortable using them on the client side, and it’s just gonna be 10x, like we like to say, probably as as far as the scale.

Of clients using AI and us getting sent AI generated or AI assisted documents, emails, messages, whatever it may be. And, you know, in my mind, there’s just absolutely no reason to try and fight it. Like it’s inevitable. It’s gonna happen. and so, I approach it, differently, where like I want my clients using AI. I I’ve said this before.

I spend time with my clients talking about AI, even doing like AI advisory with my clients as far as like, hey, I love that you’re using AI. Like I’m I’m all for it. let’s get you set up though, so that you can actually use it better and get some of this real time visibility that you want. And you know, what does that look like exactly? because you know, a lot of accountants, I feel like kind of have this dismissive tone. When it comes to their clients using AI. And

Rebecca (04:07): Yeah.

Charlie (04:08): it’s not like the clients are coming from a position of malice in the sense that they’re like, I’m gonna send my CPA this AI generated quote unquote slop in order to mess with my Accountants. No, I mean they’re genuinely, I feel trying to be helpful. And they’re like, hey, I use AI to compile all this for my schedule C. Here’s this work paper. or I fit out this document ahead of time, and you

Rebecca (04:34): Yeah.

Charlie (04:34): know, they may have not have done it correctly, they may have missed some things, but how do we get them set up so that the output and the things that they’re sending us is actually useful and helpful for us.

Rebecca (04:44): Right. Yeah, and I think part of our job has always been being educators, you know, like educating our clients. There’s so many things that we have to teach our clients and explain to them to to help them understand the process and what it all means and what tax implications are, and how to run a business. You know, we’re we’re always teachers. And so this is just a new thing that we have to teach our clients. And I think that like the key is to remember like if they sign an engagement letter, you know, this January and and they say, Yep, I wanna come back this season, you don’t have to like they want to work with you. They’re not replacing you with AI, right? So it’s the clients that leave, you know, they did replace you with AI perhaps, if we wanna be kind of dramatic about it.

Possibly they did leave because they want to try it on their own now. that’s going to happen, right? But the ones who stick around and still give you something AI generated, They must still see some sort of value in you. so I think we can let go of the frustration and like I don’t think we need to be offended when they do that. I think we need to do a little bit of education the first few years of this, because they don’t know, we saw an example of a client gave, a tax preparer a fully filled out eleven twenty s like thinking they were being helpful. But like notice they still gave it to the Accountants. They didn’t give it to the IRS. So that wasn’t them saying, I know more than you and I’m replacing you. It was just like, like Claude said this would help, here you go. And then our job is to educate and say, well, we need to know where these numbers came from. So can I have a balance sheet and and kind of work work out that process with them in this, you know, new world that we’re living in.

Charlie (06:33): Yeah, I mean this is a great point of and something I’ve been talking about a lot is like this accountant client hardness. Right. So like for my clients, the ones that use AI and pretty much every one of them does, I sent them a prompt that was like, hey, set up a new project in Claude or Chat GPT, whatever you use, and just label it accounting or whatever it is.

And here’s the instructions that I want you to basically paste into that project. And in there, it includes how to send things to me, like how to write an email to me. Yeah.

Rebecca (07:03): Mm my gosh, that’s genius.

Charlie (07:06): So like whenever a client is working in this project and it has a question or whatever, I’ve basically now pre-populated my client’s workspace, this project, so that if the client’s in there. And it’s like, okay, I want to draft a message to Charlie about this. Like that’s built into the instructions of the project that the client’s working in. So I mean, that’s really been helpful in the sense that I’m not getting back a wall of text from

Rebecca (07:32): Right.

Charlie (07:33): a client. Like it’s built to be short. It’s talks about like what the client’s trying to do, what question they’re trying to ask. Yeah. Yeah. It’s not, it’s not

Rebecca (07:39): citing it’s not citing the IRS code, you know, it’s

Charlie (07:44): it understands who it’s talking to, being

Rebecca (07:46): Mm-hmm.

Charlie (07:47): being me, if the client’s gonna send me a question. And it knows you know, our engagement, what we’ve done, how long we’ve been working for. it has access to the client’s financials that I’ve delivered to them. and the client has granted all this access, right? The client’s

Rebecca (08:01): Mm.

Charlie (08:02): like, hey, I want to use AI as well.

Rebecca (08:04): Right.

Charlie (08:05): and you know most of my clients Like what they want is just this real time visibility and a deeper understanding of their taxes and accounting that they’ve never been able to get before. And now’s the first time in history that they feel that they’re able to do that with using AI and they don’t have to bug me with every single question that comes up into their mind. Even if in their mind they may feel it’s a dumb question. But they don’t they don’t know, right? some business owners are a lot more in the weeds. when it comes to the taxes and accounting than other business owners. Others are like, I don’t care. I mean, I talked to a a guy yesterday, and he was just like, Man, I just do not care about this stuff. Like, just tell me what I need to do. you know, I’ll trust you to let me know. He was like, but I’d like want to kind of understand it at a high level, right? Like I just want to know what can I and can I not do And why, you know, if I can’t do this when I hear all my buddies talking about them doing it, can you just tell me why? Like, is that too much to ask? and it’s not. And you know, so that’s kind of the route that I’m taking is that they just want real time visibility, they want a better understanding, and they want some want more than others. And so like I’m really trying to meet my clients where they’re at and then get them set up accordingly.

So they can kind of like do this self-inform and even kind of feel like they’re staying in the loop in their own accounting and taxes. while still me doing, you know, doing all the work. I mean, they’re still paying me to do the work.

Rebecca (09:39): Right. They they haven’t fired you to do it all themselves. There there’s there’s a gray between our clients like, you know, not using AI at all and just a hundred percent going all in on us and our clients leaving to replace us fully with AI. There’s a middle. and I love that idea because you know, a lot of us are gonna be sending out, you know, seventy two sixteens for the first time if we’ve never had anyone overseas, like this is our first time. really disclosing and and having getting consent on that. And some people, you know, are feeling nervous about that that maybe some of their older clients aren’t going to be comfortable with them using AI tools. But if you frame it to them of like, hey, we’ve taken care of the security side. So you don’t have to stress about that. Like you can trust us. We we’ve done our research and we know what we need to do to to keep your data safe. But let’s talk about the exciting things that Now that we are using AI in our firm, we can give you things that we weren’t able to give you before or it was a lot harder to give you, or we would have had to charge you a lot of money to give you. So, you know, let’s open a conversation. what could we be giving you now? Like what style would you like to receive, your tax return summary in? what can we do for you to help you? feel like you’re getting exactly what you want from us now that we have tools that we can use. And then on the flip side, like you’re saying, if they’re gonna use the tools, how can we support you using those tools? So, you know, I’m picturing that project that you’re having them set up where all the accounting things go in there as well as instructions from you. You know, maybe you even when you send them their financials, you tell them like just throw those in your in that project so that that context is there, you know, like we We could be giving them things and telling them this is what we recommend doing with it, so that you have this one kind of harnessed place where the rules of how this should be talked about are there, but then also like the documents are there. so I think if someone is listening and is like, wow, I really want to set that up, I think some of my clients would be really blown away, like especially the ones who gave you AI. slop and stuff this year clearly they’re comfortable using AI with their tax information. So, you know, I want to try this with some of them and and tell them how to basically never send me an email that annoys me again, but also like make them feel supported and educated on this. you mentioned the instructions that you gave and that’s genius. And of course we all want to give our clients that prompt to paste in. but like You’re not you’re smart, but like you’re not like we’re all pretty smart, right? So could we use Claude or Chat GPT to help us figure out what those instructions should say, or is this like you were just able to figure it out on your own what it should say?

Charlie (12:26): I mean, no, I definitely worked through this with Claude. I mean, I just I saw it in my own work of

Rebecca (12:33): Mm-hmm.

Charlie (12:33): using AI in the sense of how important it is, this the concept of the harness, the concept of the instructions, the context of like telling Claude how to operate and seeing the success on my side. And so in my mind it just I mean, I know it will work. on on the on the client side. It’ll work for the client as well. We’re all using

Rebecca (12:55): Mm-hmm.

Charlie (12:56): the same models. we’re just using them in different ways. and I’ve said a million times like these things are just tools. You know, you put a sword in the hand of of a child and you’re not that that worried, you know, that they’re gonna hurt you. But if you put it in the hand of a samurai, you know, you’re a lot more concerned. I’m not saying clients or people or children. But the point is is that it’s a skill, right? Like it really is a skill. And I think that the more that you can learn that skill and and train, you know, I mean practice, put in the reps, the more you learn, the better you get using it. And then you just start kind of having these ideas

Rebecca (13:35): Mm-hmm.

Charlie (13:36): that you can, try out and see, okay, that’s working. for my client for these instructions, it’s like, It’s a V zero. Like, hey, let’s try this out and let’s see how it works. You let me know from your side of things how it’s working for you. I’ll let you know how it’s working for me. And then we’ll continue to iterate and improve. And I think that’s the best thing about AI is it’s so good at iterating, improving. You can just give it feedback and say, Don’t like this. And here this is what I don’t like about it. How can we make it better? How can we fix this? And it pretty much always has an answer. Sometimes it might go backwards. It may make a change. I’m like, no, that’s not it either. Well, okay, why isn’t that it? Why don’t you like that? Because of X, Y, and Z, tell it and just continue to refine and refine the process. and I do think, yeah, you have to get your clients need to be on the same page, you know, with your clients. And I think that makes things a lot easier when you’re working with clients is be on the same

Rebecca (14:34): Yeah.

Charlie (14:35): page.

Rebecca (14:36): Right. If you’ve decided that your firm is going to start using AI and you’re, committing to that future, which I think is wise, your client list may evolve because of that. And that’s okay. Like our businesses change over time. All businesses change over time. They’re not like a static thing and things things may be changing in your firm and and that’s okay. I think maybe starting with for those instructions, the things that maybe annoyed you about this year of what clients gave you, that might be a good place to start when you’re talking to Claude or Chat GPT about, you know, how do I write these instructions that I want to give my client? And maybe you have like like a a general one for everyone and then add some custom stuff to them as well. But you know, start with the things that frustrated you and tell Claude, you know, I didn’t like that they did this. I didn’t like that they did that. This is how we’re gonna be working together. This is the what the engagement is that they’ve signed up for. What parameters, what harness, what rules can we put in place so that they can paste this into their project and I can get actual helpful things from them and they can get helpful things from me. You know, yeah, I think there’s a lot of potential there for a smoother tax season. and I, imagine a world at some point where where we complete some financials and our agent is able to give them to their agent and their agent can put it into that project, the accounting project or something, so that there is like this one consolidated space where, you know, things are going.

So yeah, this this is really interesting to think about, So if we’re gonna prepare right now for the next tax season, I think some of that, like before you can write those instructions with Claude, I think we have to work on our mindset and just make sure our mind is ready for what this all means. So, you know, where would you begin, Charlie, with the mindset side of just accepting that you want to use AI, but you don’t really know what that means, but you know that you need to do it.

Charlie (16:37): Yeah, if it were me, if I wanted to basically become AI native, which you know just you know basically means more confident, more capable, more comfortable using AI between now and you know, let’s say January fifteen, let’s say three months. You know, you have

Rebecca (16:55): Mm-hmm.

Charlie (16:55): three months basically from today would be a good goal to set. Again, I’m big on starting goals and then working backwards.

Rebecca (17:03): Yes.

Charlie (17:05): I mean, some people are still working today. Maybe people take off tomorrow. first thing that I would do would probably be I mean, first thing I would do would be to download Chat GPT app on your phone if you don’t already. And I would go on a long walk and I would use the voice mode on the app.

And I would pick one of the voices that you like, because there’s some you don’t we all have different tastes. I really like the guy that I I say guy, sorry. I really like the voice that I talk to on Chat GPT. and I would go on a long walk and I would say, you know, I would talk through it. I’d say, This is my goal, right? Like I didn’t have the best tax season, right? Things are getting more difficult, things are changing because of AI.

Clients are obviously using it. I’m maybe not using it as much, but I see all these other people that are using it. And I want to get to a point in three months from now where I am a lot more capable, comfortable, and confident using AI. And I really want to leverage AI in my firm and also personal life, so that I can I can maximize everything that I’m doing. And I would, you know, just go for a long walk, like an hour walk, and I would just kind of brain dump everything, starting with some of the issues, right? Some of the things that I don’t like about the current operation, some of my current frustrations, whether,

Rebecca (18:32): While it’s fresh, right?

Charlie (18:34): yeah, while it’s fresh, while it’s in your mind. you know, whether that be your type of clients, the interactions, your software stack. So I’d kind of start with, you know, the things that are top of mind that you don’t like. And then I would make sure and tell it you know, the my setup. Like, okay, I have this type of firm, I have this many employees, I have the remote, some are in person, we have this is our tech stack, like this is how like high level of how the firm operates. We don’t have to get into super granular, but the more details I think you give it, the better. And I think you start off this conversation with. with whatever model you’re using, by basically telling it like you want to brain dump, you don’t like really want to get deep into it yet. You just want to

Rebecca (19:24): Yeah.

Charlie (19:24): give it information. Right. And so, it’s kind of like a it’s like a therapy session with the with

Rebecca (19:29): Yes.

Charlie (19:30): the AI. And talk about where you want to go, talk about, your frustrations, talk about where you’re at today, talk about how much you’re using AI. I would tell it typically how you use AI, it’s just like, well, I only use it in the chat. I don’t work in projects or I do work in projects or I use cloud code or I use the web app, I use the desktop app, whatever it may be. Tell it where you’re at today and how you’re using it today. and some of the success and failures that you’ve had by the way that you’re using it. and so yeah.

Rebecca (19:59): Yeah. You could easily also say, like, I don’t even know what I should be telling you. Like, can you ask me some questions and store the answers somewhere so that every time I open a chat with you, you know the context of who I am. You know, I think sometimes we don’t even know like what we’re supposed to tell it. So if you give it some background on who you are and what what your end goal is, you could also say, Can you ask me some questions? That would be helpful to Answers too.

Charlie (20:28): Yeah, I think, and maybe I’ll even try and write something like this up, but like a starter prompt, right? Like a prompt that you could just copy and paste into Chat GPT. And then you can say, all right, and then you can start the voice mode and say, okay, let’s begin, basically. And it’ll have that context of like what this session, what this thread’s

Rebecca (20:48): Yeah.

Charlie (20:49): gonna be about. but you know, high level, that’s what I would do to start like day one. I would have this conversation, brain dump everything. And then I would basically take that transcript and I would tell it, like, okay, let’s summarize in detail this entire thread and let’s break it out into the relevant sections and parts of what we talked about. And then let’s start going deeper into like each of these sections.

And then I would go deeper into each one, like the frustrations, the workflow, the current tech stack, all this, the use of AI and how you’re using it. I mean, this will take time. Like you really do have to kind of spend some time on the front end to save on the back end, but it really, really is important to put in the time. And now that we have time. in these last three months, I think it’s definitely worth spending a week just just like chatting high level about things. Like I wouldn’t try and set a three month roadmap on day one. I would try and get organized and kind of get it get all my thoughts out and have the AI kind of understand where I’m coming from, you know, like why I’m trying to do this, where I feel like is my weak areas my strength, etc.

And then like once it has that information and already knows your goal, right, then it can try to fill this gap.

Rebecca (22:10): So what I’m hearing you say about this conversation we’re going to have with Chat GPT and dig into this, what’s interesting is my mind, as you’re saying this goes to couldn’t any business owner do this? Like anyone who owns a business and wants to figure out how AI fits into it, like maybe this is a great place for everyone to start.

Charlie (22:31): Absolutely. I mean, this exercise is not only for accountants. I mean we have other businesses you and I do, and like I’m gonna do the exact same thing. you and I talk all the time about what are we gonna do over the next three months for this podcast for ALFA over at TCR Like what are we gonna do? And this is exactly what I’ve done, I go on these long runs and I’m talking to ChatGPT the whole time and trying to talk about, you know, okay, where do we want to go? where do we want to take it then? Where are we at today? And then how do we fill this gap? And then I start high level and then I get back at my desk and I’m like, okay, let’s summarize this entire thread in detail. Let’s break it out into the different sections. And then let’s start going into these sections in more detail. And then once you kind of start doing that, Then you can start trying to plan, okay, well, what should I do this next week? What should my goal be for this first month? Right? What should my goal be for the two months? start, with the goal, work backwards, and then plan forward is how I kind of approach these.

Rebecca (23:36): What’s interesting is you’re not saying like the first step I would do to be an AI native firm, by January is not like to take a course or to like research Chat GPT, like how do you like you’re just at you’re just saying you should just talk to it. And I think what’s cool about that approach is then like you know, as you’re having those conversations and you’re starting to plan, steps of what we need to do to reach the goal. you’re building your knowledge of how to use AI just by practicing through talking to it. And then also you’re building like your very own custom plan. Like, you know, I I could tell you a plan of, you know, you should increase prices, you should drop these clients, you should stop using this software. What you know, we could tell you things to do that are generic.

But if you have this conversation with your agent and your AI, it’s gonna be like getting for the first time ever, like your something that’s exactly built and designed for all the nuances of your situation. so I think that’s pretty exciting.

Charlie (24:40): that’s what’s so great about it is that it’s really you know one of one plan right and I have like direct say into what this plan eventually entails with the help of of AI and a lot of back and forth right like it’ll say things all the time it’ll suggest things and I’ll say well I don’t I don’t like that I don’t want to do that or I’d rather do this. one thing that I think you could also do in parallel is just start obviously using AI on a personal level. And I’ve said this a a bunch of times too. using these personal agents, whether it’s Grokbot, whether it’s Dots, whether it’s Muse, whatever it is, just start handing off more and more to them because that’s a great way to kind of see what they’re capable of doing.

I get, kind of proactive nudges from my personal agents all the time. And they’ll be like, Do you want me to do this for you? And I’m like, huh, I didn’t really think about that. But yeah, I mean, usually they’re all like low stakes items. I mean, I’m not doing anything crazy with them. So it’s like I’m not too concerned. If it does do this, and like what’s the worst that can happen? It marks an email read you know, when it was unread You know, start low stakes.

And then another great thing, and this is how I’ve really, I think, really learned a lot was building a custom app. And I think when people hear like building an app, they roll their eyes, they’re like, okay, vibe coding an app. But if you really build just a super personalized application, it doesn’t have to be a mobile app, right? I’m not talking about launching an app in the app store.

I’m talking just building a basic web app or whatever it may be that does like one tiny thing that you’ve always

Rebecca (26:29): Like a tool.

Charlie (26:30): wanted, yeah, like a tool. And you know, for me, when vibe coding did come on the scene last summer, I mean, I started building all kinds of apps. I mean, I remember I built a you know, a a workout tracker app to

Rebecca (26:42): Mm.

Charlie (26:42): where I now I work out at home, but I used to go to a gym. And one thing that I always hated was when I was tracking my workouts, you know, there will be like three different types of a machine. You know, there’s like three different Lab pull downs. And

Rebecca (26:55): Mm-hmm.

Charlie (26:56): they’re all like slightly different or they’re a different model. And so 90 pounds on one does not equal 90 pounds on this other one that looks the exact same. And

Rebecca (27:04): Uh-huh.

Charlie (27:05): I remember like, well, I can’t really track my workouts that grade. And so I built an app. And what I did was I inventoried every equipment in that gym. And I built my own inventory of the equipment in the gym. And I took pictures of every equipment and I uploaded to this app. And so then I could track my progress based on the exact machine of this gym, not just a generic, you know, Lat pull down. And

Rebecca (27:31): Wow.

Charlie (27:31): you know, that was just low stakes. but I learned a lot of you know, how do you build an app. I there’s a lot that goes into it. You got to host it. You got to figure out how do you log into it. You got to figure out setting up auth. You got to figure out how to set up the back end, the front end, and your UI preferences. And you kind of learn a lot by just kind of working through something like that. and then like once you get it working and it’s like, this is actually this is actually pretty cool. Like what else could I build? You know, what else could I do? And so I think, you know, It’s interesting because a lot of people that I feel like we talk to do have this like aspiration of building an app. And

Rebecca (28:09): Mm-hmm.

Charlie (28:10): some people like look down on that or kind of dismiss it. But like I mean, if you’re trying to build an app to sell to thousands and thousands of people, that’s a whole different ballgame. That’s a whole

Rebecca (28:21): Yeah.

Charlie (28:21): different conversation. But if you’re trying to use AI to build only an app that you use and to help you solve one specific problem that you have, I think it’s a great way to learn. a lot when it comes to using ai to really do things how to

Rebecca (28:37): Yeah.

Charlie (28:37): to set this up how to build this what are the testing that needs to be done what are considerations that need to be done how do you host it how do you deploy it how do I log into it okay I have

Rebecca (28:46): Mm-hmm.

Charlie (28:47): to set this up how do I set that up and you

Rebecca (28:49): Yeah.

Charlie (28:50): just really work through AI and I think that’s that’s really key is like working through things and over a long horizon on something and sticking with it and seeing it through and kind of grudging through some of the the more challenging parts.

Rebecca (29:07): Yeah. And I think, you know, you you mentioned the other day on the ALFA call that you can also take screenshots of things that you like and say, like, I want it to look like this, or I I don’t like this feature about this, but I do like that feature. So, you know, think about that when or just go to this website and go look at it because I do like what’s going on over there, or the these are the problems I have with this existing thing. so you know, if you could you can use ideas from From elsewhere to help you. I think for me, a tool came before an app. and I didn’t even know what a tool or an app was. I just knew that I had a spreadsheet called the client analyzer that you could, you know, input all your client names and what services they have and what you’re currently charging them for those services. And then it would do some formulas of like, If you increase the price by this much, this is how much your revenue would go up if they say yes and kind of different scenario analysis. And I just wanted it to look a little prettier. And someone said that Claude was good at Excel. So I downloaded Claude. I don’t well, I don’t even think I had the desktop app. I went to the website and created a free account and I gave it this spreadsheet and said, I’m about to present something in TCR about this, and I just want this to look prettier. And that’s all I said. And it made this like really complex, like a million tabs. It had like a readme thing on the first tab. And it was doing like amazing things that I had never even thought of that that we could analyze, you know, with someone’s current like book of business. but at one point I said to it, you know, this feels like There’s so many formulas and it feels like there are a lot of errors because I kept catching errors and it kept saying, like, you’re right, I did make that error. So, you know, I had to review the work and test it. And, I thought, let me tell it like my frustrations with what we have. Like we’ve built something cooler, definitely, and prettier, but it still feels like not user-friendly, and that’s frustrating me. So then it said, well, here, click this.

And it made this what looked like a website to me, which I learned was called an HTML file, that was interactive. I could click it. I felt like I was just on a website, you know. I like I did not know what HTML was. and it was like the same concept as my spreadsheet, but it was like so pretty and easy and like buttons and and no formulas anywhere, nothing that could break.

But it gave me the exact same end result, but just in a way more digestible way. so then I said, I love this. How do I give this to someone else? Right. That was the next step of I want to share this. This is great, you know? and I remember also having to do a presentation at a conference, and it gave it to me in this HTML, like website type form. And I was like, what about we’re not using PowerPoint? And it was like, no, like.

You just have this little website essentially that is your presentation. So then it kind of got to like, okay, well I have to give the slides to, the conference ahead of time. How do I how do I do this? Or how do I share my this cool tool that you made with others? And then my mind like started going, like, well, what other tools can I make? Like, you know, now I’m seeing I can make these little websites that are interactive. and I know websites not the right word, but that’s kind of what it looks like. It you know, it’s And I think some of them you don’t even need internet to be using it, But from there it told me like, well, then you would want to host it and you can use this kind of domain or that. And that’s where I got started getting uncomfortable, right? Because it was no longer like I put something in and magic comes out. It was saying like you’re gonna have to decide whether you’re gonna download this software or log into this account, you know. And that’s where I had to say, like, look, I don’t know what any of this means. You what do you recommend? Like walk me through it. I need you to really hold my hand. I took a ton of screenshots of each screen I was on and had it help me through. Now I probably would press the computer use button and just have it set everything up for me and I could just watch it do it. But at the time, you know, it was kind of cool to work my way through it because I started learning how domains work and the DNS records and what an HTML is and and that it can have this generic name or you can tie it to your domain and have it be part of your website and what a subdomain is. And then I was like, wait, could I work on my websites here? You know, so I gave it my current website.

And then it started showing me what we could do to it and it was updating it in real time. And then it, you know, and it all just kind of like flows from there. Then it said, well, if we get this GitHub thing, then you don’t have to keep dragging and dropping the new HTML file into Netlify, then you would just be able to just say, Yes, I like it, and it would just magically appear on the website. so then I had these tools kind of become part of my website and Then I could easily just give someone the link to my website. I imagine that would be great for a client if they needed some sort of tool. You could just give them a website, link that that they could go to. so yeah, I think like you were saying, like just working through something very new with it, that and accepting that it’s gonna say terms that I don’t know and it’s gonna be weird, but I can always ask it. to explain it better if I get stuck on a step or ask, is there a way you can do this part for me? And it will tell you if it can or not. and then once the light bulb goes off, then it’s like I can make anything like that. and then it’s like apps with a login, a sign in, you know, a real like entering this, you know, app that has all these things you can do inside of it. Like that’s, you know, next level.

But it could start with you just giving it a spreadsheet that you don’t like and asking it to make it better and make it not need Excel anymore.

Charlie (35:02): Yeah, it’s a great point. I mean, you could start with your existing, if you’re an Accountants, you could start with your existing work papers. And you don’t even have to use client data. Like just give it a completely blank work paper and say, Hey, can you make fictitious synthetic data and use this work paper that I have created and basically tell me how can it be improved? Right. Like how could this be better? and it’ll do that. I mean, it’ll create completely fictitious synthetic data and it’ll run it through. work paper and it could it could point out errors. this doesn’t tie or this this is a bad formula. This one isn’t flowing through to this one. might be better if it was presented this way or you added something here or you removed something there. and it’s really great at at kind of doing that. And like you said, I mean computer use has come such a long way from whenever I, you know, kind of started going down this path and yeah.

I’m still a big screenshotter. I mean I screenshot everything and give it to the AI and and tell it whatever feedback it is about screenshot. but it’s amazing at what it can do in the sense of like, it can go and set up GitHub, it can create a new repository, it can push the code there, it can, you know, merge the code or merge the changes, it can go change your DNS records. it can do all of this stuff. And so you’re right, if it says it can’t do something, typically I ask it well, why? What do you need in order to do it? And I’m like, well, I don’t have authorization. I’m not logged in to your GitHub account. Okay, well, let’s log into GitHub.

Rebecca (36:37): Yeah.

Charlie (36:38): you know, I’m not logged in to Cloudflare. Okay, well, let’s log into Cloudflare. let’s get you access to all these different applications that you use that do allow either MCP or API or whatever it may be to where they can just access it directly from Claude and yeah the alternative would be you know just computer use.

Rebecca (36:56): Sure. Yeah, it’s it’s great that we can get curious with it and be like this is kind of fun. I want to sort of learn how to do this myself with it, or you can just say, Look, I don’t want to know anything about this. Like, this is the end goal, these are the things you’re allowed to do. you know, I will authorize what you’re doing, but like you always joke that I ship a lot of stuff or whatever you say. I don’t even know the right terminology. But I can’t remember the last time I logged into GitHub. I mean I think like once when we were setting up like the ALFA GitHub, community GitHub, but I mean I’m shipping stuff all day. It’s just like I have this this thing that does everything for me, you know, from my phone usually.

Charlie (37:39): Yeah. Yeah, it is funny. I mean, and that’s the beauty though, is like there’s no right or wrong way to do things. I definitely get more into the weeds and am more involved in the process. and you do so many things and ship so many changes or improvements or whatever it may be, and you’re not like super technical. I mean, I’m not super technical either. But I mean that’s the beauty again, is that you don’t have to be super technical to actually produce something, build something. now again, if you’re gonna build something that is in front of other users or in front of clients, like you have to approach that completely differently. Right. If you’re gonna build an app that you want to bring your clients on or you want to have paying customers, like that’s a completely different conversation. I’m not saying you approach that haphazardly. But what I am saying, if you’re going to build an app for yourself and it’s going to be kind of a low stakes app, if you’re going to build a work paper for yourself, or even if you’re going to build a work paper using your firm, you can start out with just using basically demo data and saying, okay, I want to create this playground, this sandbox environment where I’m not going to use client data. I’m not going to use, any real data. Let’s just use completely fictitious data. Claude, can you build it? Codex, can you build This fake data. I want to make a completely fake profile of a business owner that is an S-corp that has X, Y, and Z. and it’ll build. I mean, it will. And then you can say, well, that’s don’t like it. It needs to have this, it needs have this, whatever it may be. Spend the time. I mean, I have this. I have now just numerous different demo clients, synthetic clients that are just all. just fictitious synthetic data that I use to test with on things that I’ve built. and it was all built, you know, by AI. and it’ll go and you can, I mean all these tax forms are available online. I mean the IRS even has designated social security numbers that can be used for demo purposes. and so you can just kind of get all this information that’s readily available to you and kind of build out This data set that you need. because I mean we’ve worked with clients for so long. Like we know that some of the typical things that we see, and we can kind of build out this data set that we can use to do this testing with, to try and improve our processes without even using any client data at all.

Rebecca (39:58): Yeah, and you can ask your team right now too, if you have a team of like what what really slowed you down this tax season or what’s like part of the review process that feels so tedious that you hate having to sit and like what what what do you have to tie out that feels like hard to tie out or you know, like asking those questions to too, from the whole team to kind of compile all of that into what can we improve before next tax season? kind of plan and what I I feel like as soon as you just dive in and make one little thing with it, it becomes exponential, like your comfort level and the the things that you can start making. you just have to start and you just have to keep pushing it and say, I don’t understand what you’re talking about. Can you do that for me? Is there a free way to do this? I don’t want to pay for that or you know just say whatever comes to mind.

And as soon as you do build that thing or solve that problem, then your mind’s gonna start saying, what else? you know, and and then the you have the confidence and then it just like snowballs from there.

Charlie (41:02): It really does. I feel like it just opens up your mind to all of these other, issues that you may have overlooked because you start solving these various different problems and things and areas in your life to where like, okay, you really think about this one up here, kind of forgot about this issue or this area. Let me try and tackle that and see what that looks like, see how that goes. And then Take the learnings from there and just continue to, like you said, snowball and compound into all these, areas of your life both professionally and personally. And that’s when I think you just the light bulbs start popping off. And now anytime I see anything, I’m just, you know, well, let me try tackle that with with Claude or Chat GPT and work through it. And then, a lot of times it ends up in a you know, a good result. And I’m like, wow, that was this is this is good. This is nice.

Rebecca (41:55): Yeah, I think the still the best part about all of this is how custom it is, you know, like when have we ever been able to have something work exactly how we want it to work and exactly how it makes sense for our unique situation. You know, this is a time where even if it’s not like an app or anything, it’s just an org chart or a workflow, like a flow chart showing how things are gonna work in your firm, like it can just be made in an instant without you having to drag and drop little like shapes and arrows and all of that. It can just be made for you. And then you can say, no, I don’t like that, or I want like show me this workflow this way. and I think that’s great. Like that we everything can just be super customized and updated in real time extremely quickly whenever we want anything changed.

Charlie (42:43): I do think it’s important to note that like these models are not perfect, you know. I do not rely on them

Rebecca (42:48): Yeah.

Charlie (42:49): fully. it’s not like I can’t, you know, do a task without AI. But again,

Rebecca (42:55): Mm-hmm.

Charlie (42:55): I treat them as a thought partner. Like, hey, I have this idea, I have this issue, let’s work through it. And I push back on it all the time. Or I’m like, no, I don’t want to do that. Or like, no, that doesn’t make sense. I don’t like that, like all the time.

I’ve rarely ever take things at face value when it does. I haven’t seen that too much with people that I’ve talked with. They don’t treat it as like a sparring partner. They treat it as more like a task worker. Like, do this task, you didn’t do it. These models are not good. And I’m like, that’s

Rebecca (43:23): Yes.

Charlie (43:23): not how they work. there’s a lot more setup to it. There’s it’s a lot more nuance, there’s a lot more back and forth. and a lot of people may think like Okay, well why is that worth it? Like you’re spending more time, it sounds like and I’m like, That’s not it either. Like that that’s that’s ignorant to say.

Rebecca (43:36): It’s not necessarily time savings might not be your goal.

Charlie (43:40): Right. I mean, you can’t quantify a lot of the things that AI does for you. I mean, there’s qualitative metrics at hand. and there is quantitative metrics as well. But I do think again, if you do something and it’s not great output or you don’t like it, if you just end there, then yeah, that’s probably a waste of time. But if you sit there and say, No, this is why Let’s get it right. And if it’s a repeatable thing, say, okay, now that we got this right, we need to save this as a template. update the instructions, update the memory, update the claude.md update, update, so that the next time I come and do this maybe same task or one very similarly, I want to have basically a pre-built template. I want to have maybe it’s a skill that you run, or like whatever it may be, If you do that, then that’s when you save time on the next one. you maybe spent two hours, three hours going through something that usually that could probably take you 30 minutes to do if you did it yourself. Right. But if you

Rebecca (44:39): Yeah, right.

Charlie (44:41): if the next time you do it, it takes you five minutes or ten minutes, whatever. And the next time you do it, it’s just gonna compound. Like you’re gonna save time in the future. I mean, it’s just delayed gratification.

Rebecca (44:51): And you don’t have to know what a claude.md is or what a skill is, none of that. You can just talk to this AI thing and say, you know, this is perfect. I love what we did here. We finally have a version that I love and I’m going to have to do this again. What can we do together right now to make it so that the end goal that I’m looking for, which is to be able to do this, my my end goal is to be able to have this be a repetitive thing that I can do, you know.

What should we do right now to make sure that that can happen? And then it might say, well, let’s create a skill and here’s how and you could say, I don’t know what a skill is. And then it could say, Well, you’re gonna click this each time. And I recommend that we put it in a project or I recommend that we update this file or you know, like just ask it, you know. and I think you brought up a good point of like a thought partner. you know, AI can be used to to do work or to review work, right? to make things, but it can also be used to help us think. and I don’t think we talk about that enough. I think you and I use it a ton for that. And I think I probably maybe use it too much for that because I already had a personality before AI of like I’m that girl who if I have to make a decision, I ask like 500 people like before I make the decision. So it’s like you know, having these voice, like two-way voice, like Chat GPT, having this person I can constantly say, like, what’s your take on this? What do you think? Is like probably all my friends are happy because now I don’t ask them as much. But it’s been great for me because I really like bouncing ideas off of someone else. so just having even if you never built a single thing or never had it take over a single workflow. let’s say your IT department says like you can’t even use it on client work and so you’re like, well let’s just say you’re not going to use it to create work or review work or anything. There’s still this whole other lane of having it be a thought partner, where it helps you flesh out, like you say, the details of, you know, help you make sense of bigger problems, help you just navigate things and help you look at it from a different lens or play devil’s advocate or, you know, a lot of times it’ll just it’ll s solve something, but then it won’t say on its own, like, but we could make it even better by doing XYZ. But if you just ask it, it like immediately has the answers. Like it was holding out on you of like all these other features that could be added or other ways that you could make tax season better. You know, so just just having it be someone that you think with and ask it any question that ever comes to mind about anything and go back and forth with it, I mean that would be so like itself without even worrying about the work part.

Charlie (47:29): Yeah, and I think the important part there is think with you, not think for you. Like

Rebecca (47:33): Mm.

Charlie (47:34): I do not let these things think for me, right? Like I

Rebecca (47:37): What can you give me an example of what would think for look like and what would think with on something?

Charlie (47:43): Well, it’s just like this example of like, hey, let’s do this process of us chatting with it about what I should do for the next three months. Like if you

Rebecca (47:49): Mm-hmm.

Charlie (47:50): just said, Hey, what should I do the next three months? And it says you should do this and this, and you say, Okay, I’m gonna go do this and this, that’s basically th thinking for you. Right. And like

Rebecca (47:57): Mm-hmm.

Charlie (47:57): for everything I do, I will basically tell how I’m thinking or what I’m thinking about. my perspective, right? Like here’s my thought on this. can we just chat about it, basically? give me your perspective on it. Is there something I might be missing? help me play devil’s advocate here. And then basically taking what it says and sparring with it, right? Be like, okay, that point’s interesting. I didn’t really think about that. Let’s talk a bit about that more.

And then you can just kind of expand on it and then be like, okay. some of that’s valid, some of that I don’t agree with, you know, and so I’m not gonna just agree with it because it says it. again, you just gotta constantly push back on it and spar with it. And then I mean, there’s been several times where it’s told me that’s not a good idea. Like, why would you build this app? There a million of them already exist. Why don’t you just download this one? Well, because I want to.

Right, because I want to learn the process of building one, not because I just want one to use it. and I’m I’ve told it that a million times of it it says, Well, you know, yeah.

Rebecca (49:01): Or the opposite, right? It it just agrees with everything and you’re like, you know, okay, like you’re getting me excited about all these things, but like I’m realizing that you just say yes to everything. So like not jumping on it immediately and and talking it out more and

Charlie (49:17): Yeah, and that’s a good point. if your AI does do that, then I would change the instructions to say, I don’t want you to be an attaboy. Whenever you’re I mean, that’s what my instructions say. I don’t want you to just agree with everything that I tell you. Like I want to use you as a thought partner, you know, sparring partner.

Rebecca (49:34): Mm-hmm.

Charlie (49:36): I want you to, not just immediately argue with every point. And not immediately agree with everything I say. if I ask you a question like that, that means I want to have basically an intellectual conversation about it and try to

Rebecca (49:46): Mm-hmm.

Charlie (49:47): consider all the different sides, maybe some things that I’m missing. but ultimately like I’m making the end decision And so hopefully the whole goal is that I’m now in a better place to make the decision Maybe I have better information now.

Because I’ve had this conversation, we’ve done this back and forth. I’ve ruled out some things and some things have been brought in that I wasn’t thinking about. And so that’s

Rebecca (50:10): Yeah.

Charlie (50:11): how I approach it and that’s how I use it. and I think that’s once you kinda like me again, going back to shifting of the mindset, like I think that’s how people should kind of have that mindset with with AI.

Rebecca (50:22): And I think what you just said, if you notice an annoying pattern of like it always does this, you could just tell it, hey, I don’t like that you always do this. How can we fix that? And it will say, like, let me put that in my instructions so that I won’t do that anymore. Or, you know, you see people say like, I asked it to do this and it did it totally wrong, it’s trash, right? But you could say, Hey, you did this wrong.

And I don’t know if I should use you to do this anymore. Like is there a way we could make it so that you do it this way instead of that way? there’s some maintenance involved here of, I’m not satisfied with the product that you delivered to me. I’d like it to be different. Or the I don’t I’m not satisfied with the way you respond to me on things. Like, how can we make sure you don’t do that again?

Charlie (51:08): I mean, I think it’s super important to have that approach to where if you again, if you get bad output, you should try and figure out why. You know, it’s

Rebecca (51:17): Mm-hmm.

Charlie (51:17): it’s it’s unlikely that it’s the model’s fault. I mean, sometimes yes, they hallucinate, even though those have decreased as the models continue to get better. And sometimes they make mistakes, they’ll flat out tell you, hey, I made a mistake, right? Those do happen. You know, I’m not saying they don’t, but other times it could just be like a setup problem.

I can’t remember what I was using the other day. but I had it drafted an email for me and it said hi, H I. And I was like, I have not seen that. I don’t say hi, I say hey. I say hey first name usually is how I start every email. And I was like, why did you just say hi in this beginning? And I was using something I think I was on a different machine. that I don’t use that often. I don’t think I had my my global Claude MD. I didn’t have something set up right. But I was just like, okay, well, I say hey first name, update your the Claude MD, update whatever you need to update so that next time I tell you to draft an email, it says that. Or maybe I was using Muse, using something that I hadn’t like fully set up. And it was like, okay, you know, instructions are updated, memory’s been updated, going forward, I will say hey first name.

Rebecca (52:29): And I don’t think I’ve ever really done stuff like that. Like I’m not using these things as official. I mean, I think now I’m starting to, but for the first many, many months of using these models, I would just start from scratch every time and be frustrated every time, or just edit it myself every time. but what you’re investing in is like this long term relationship. Like, look, we’re gonna be together forever. And I want to make sure that you know exactly what I like and what I don’t like. And when I change my mind, I’ll let you know so you can write that down too. I think that that probably a lot of beginners are not doing that. I I certainly wasn’t. I I am not at the beginning telling it, you know, I like to start with hey not high. I I didn’t tell it that I just would edit it each time myself, you know.

So that’s another thing is as you start seeing things you don’t like or that you’re having to manually fix, just tell it that so that it can make your life easier and automate that for you in the future.

Charlie (53:29): I think it’s a good way to put it of investing in the long term. yeah, I mean, I have put a lot of time into to my setup and how I’m using these models. And because again, I just, for whatever reason, being intuitive or whatever it may be, but I was just like, hey, I was always fascinated whenever I’d get a bad output. because I think I had seen like the capabilities early on.

And I’m like, how do you go from like doing this, which is amazing? Like, yeah,

Rebecca (53:56): Curing cancer to you know.

Charlie (53:58): I mean, just creating this incredible, you know, spreadsheet until you did this. Like how how?

Rebecca (54:04): Mm-hmm.

Charlie (54:05): What is that? Where’d that discrepancy come from? Why is that spread so large? And maybe it’s just

Rebecca (54:10): Yeah. Get curious.

Charlie (54:12): the yeah, maybe it’s the auditor in me. I would always dig into the things that

Rebecca (54:14): Mm. Yes.

Charlie (54:17): were wrong and be like, where’s the source of that? And I would just ask it. I’d be like, where did you pull this number from? this number’s not right. And it would be like, well, I pulled it from this document, which, now looking at it, it’s a stale document. And I’m like, okay, well, how do we prevent that from happening again? Well, we should do X, Y, and Z. I’m like, okay, well, let’s do that. Let’s get that set up so that this doesn’t happen again. I mean, my whole

Rebecca (54:42): Mm-hmm.

Charlie (54:42): goal when I’m working with AI is if there’s a bad output, it’s like, how does this not happen again? And I think now I’m at the point. fortunately have done that over a period of time to where I usually get really solid output from these models. and it just makes the experience just so much better.

Rebecca (55:00): And it’s not because you have like a computer science degree. It’s because you you asked

Charlie (55:04): I don’t. I do not.

Rebecca (55:06): it like, hey, I want to set you up well. How can I how do we do the setup? It wasn’t that you already knew how a proper setup should go. You just asked it or you made this mistake. How do I make it not happen again? It’s just asking it. Like it’s the genius, not necessarily us, you know. and this could be great for clients too. Like if they get a bad output, they send you some ridiculous email that’s like bad output, and we say, like, well, it’s all in the prompt, you know, you didn’t prompt it right. tell the client to tell their, you know, model or agent or whatever, this is not right. How can we make this so that you Are right in the future, to help you not get like garbage from your clients, essentially. or you know, I think a lot of accounting firms are going to start doing AI consulting for business owners. You know, if they want to help businesses implement a lot of this, I think you know, accountants know so much about how a business works and you know, the ops and the accounting and you know the how money moves through a company.

We know a we’re forced to know a lot about technology because of all the tech stacks we’ve had to, you know, figure out and all the different software that our clients have that we have to log into. So I think, you know, we us investing in understanding how to think with it and use it, then maybe it’s even another service we can add for our business clients where we help them, you know, learn how to do some of this or set some things up for them as well.

Charlie (56:33): I really think there’s gonna be a lot more synergy when it comes to client and client AI and Accountants and Accountants AI.

Rebecca (56:41): Mm-hmm.

Charlie (56:42): I’m just optimistic and I’ve even seen that in my own work that I do now. and I do think you have to like kind of foster this environment of like, hey, what’s our goal? Or I mean, you talking to a client, like what’s our goal? It’s to service the client. Like he’s paying me to. provide a service to him. and he has needs and desires and things that he expects, you know, from working with me and then vice versa from my end. And so we’re both working towards the same goal. Right. and so let’s figure out how we can get there together. You like AI, I like AI, let’s let’s figure this out and then work through with the client as of what that looks like. And just have that conversation with them.

And I think that naturally goes into okay, well, you gotta change your setup, right? You gotta do some things so that your AI has the access to the information it needs so that when we’re having these conversations, they’re intellectual conversations as well between us because it has the information it needs, I have the information I need, we’re on the same page, there’s no missing information. we’re all kind of looking at the same thing here. And the great thing is.

Is that it they’re from different kind of lenses, right? Like I’m looking at it from my lens, business owners looking at it from a different lens. And so now we’re kind of putting these two kind of hats together with AI. And you know, I now just assume that whatever I send my client, it they’re gonna put it into AI and send it back to me. And I think, you know, a lot of people don’t like that thought. But for me, I love it because I’m like, I hope their AI finds something. I hope it I mean, obviously I hope it doesn’t find anything, but it’s like a challenge almost like, okay, how great can I make my work so that there are no holes in this? I have really done my due diligence with this. And I’m I’m proud to to give this to the client because I’m like, look how great it is.

Rebecca (58:35): Mm-hmm.

Charlie (58:36): and the client’s like, yeah, I fed it to AI and I don’t see anything wrong. But I mean, look, we all make mistakes. Like

Rebecca (58:42): Yeah.

Charlie (58:42): I mean, I make mistakes all the time. I mean, I’m a human.

Rebecca (58:44): Mm-hmm.

Charlie (58:45): I make mistakes all the time. The key is catching them, catching those mistakes, right? And hopefully not making, material mistakes. But things happen. and I think the key again is getting into a position, creating this environment to where it’s set up in a way to minimize the likelihood of those mistakes and obviously learning. any mistakes that do happen and working to implement whatever is needed, any safeguards or whatever it may be, so that, the mistakes just start, going down and down over time.

Rebecca (59:17): And you can just ask Claude, how do we create this environment? How do we make sure that mistakes are always going down? Like, just ask it and it will help you figure that out. You know, like sometimes I feel intimidated that I don’t know how to set these things up right or I don’t you know, but if if you could just tell Claude, I feel intimidated about the setup, like what what could you do to help me with this? And I think if you invest, kind of bringing it back to October 15th and we have a few months here until we start again. If you can invest the time now into getting from you know where I was a few months ago to where I am now of you know feeling pretty comfortable with with using it, if you can get there, you might start seeing opportunities to help your clients get there too. And they may pay you for that service. And that doesn’t mean you have to be like an AI expert, but if you figured out a great tool with AI or a great way to talk to it or just a great way to set things up that feels very organized, you might identify opportunities within either their personal life or their business, like their personal financial life or their business that you know you can add value by doing a little consult with them and teaching them some of this stuff. So You know, this can really help you add more services to your firm, which we need because some things are gonna go away. It’s inevitable as we evolve with AI. So we need new things. And you know, we can’t we can’t start offering AI accounting services or you know, with an accounting lens until we figure out AI ourselves. So to figure it out ourselves, we just need to spend a few months talking to it and thinking with it and Giving it something that you don’t like and asking how it can make it better. And through a few months of practice, I think some light bulbs are gonna go off and things are gonna click into place. that’s that’s the story we hear from every ALFA member that is really comfortable using AI now is like there’s just that moment where it clicks and and the only way to get there is by using it and practicing with it. So that’s our homework, you know, that’s what we got a few months. And that’s the best thing we think you can do with AI, you know, to get closer to that AI native firm for next tax season. and let us know how it goes.

Charlie (1:01:34): Yep. All right. Well, we’ll end it here and we’ll see ’em in the next one.

Rebecca (1:01:39): Yep, see you next time.