The (over)promise of AI

***AI generated, so likely some errors in this transcription.***

00:07.22

satvcpodcast

And ladies and gentlemen you're in the right place. It is strategy in the virtual controller we're just playing around with some of our intro music to see if we can find something new because change is as good as a holiday and since none of us are traveling anywhere. We may as well see what we can change so a bit of ah, a new music and we might be testing a few different intro and outro tunes in the coming episodes but ladies and gentlemen. My name is Damien Grayhead 1 of the co-hosts of strategy in the virtual controller delighted to be with you here today. Penny Breslin my co-host sitting in San Diego how are you penny.

00:41.85

penny

I'm fine Damien how's the weather air in Australia.

00:43.50

satvcpodcast

Good you know I think it's this lainia thing at the moment isn't it and that means that it's wet. So it's wet down here in Australia everything is green and lush but everything's also sodden so it's hot.

00:48.46

penny

So yeah.

01:00.15

satvcpodcast

And humid and wet. But no, it's good. Can't really complain. It's better than dry and and wildfires bushfires.

01:08.80

penny

That's for sure and we're getting the same effect here in San Diego but it's not necessarily hot right now. Although I wouldn't call it cold either. It's just you know, nice and pleasant and high sixty s and the sun finally comes out in the afternoon I almost feel like it's may gray and June gloom rather than.

01:24.50

satvcpodcast

Yeah, and and and also um, the good thing is at least the reservoirs are getting some water isn't it is that That's what I've heard.

01:25.54

penny

January but I'll take it.

01:31.79

penny

Yeah, that's good. That's good. It would be nice if there were collectors and cisterns for times like this where we could collect even more Unfortunately I think it's something that has to be looked at but that's ah, that's a total difference.

01:49.00

satvcpodcast

That's another episode. That's another podcast I think um penny you sent me an email um earlier about as we were discussing today's episode and.

01:50.38

penny

Yeah.

02:00.49

satvcpodcast

Ai and context was it was what you replied with and having ah an episode on Ai and context and it got me thinking um and very interested to to dig into why you want to discuss that but I'm going to put it out there as I was thinking about it and certainly from my receipt bank decks. Days. We've been banging on machine learning and ai for a long time and and I sort of had the realization this morning that it might not have come that far in the accounting industry. That's my sort of initial thought but hand it over to you? um. Why did you want to go with Ai in context for today's episode

02:39.27

penny

Well get on your end obviously and it's it's invariable that we're going to get comments from firms that outsource to us like you know what did you. Did you allocate this this way and part of it becomes Well you're allowing the ai to do the work in order to take you know cut down on data entry time and if you don't take the time to review the work that the Ai does. You can miss something so say what happens is something is being done over and over and over again the same way and then that transaction all of a sudden has a nuance to it and the nuance might be a difference In. And the size of the transaction and ah it may be that they were same vendor same amount of money but it's being paid through a different way and that doesn't get noticed until somebody goes in and reviews the file and then the question becomes why is this different. So. Let the eye. The Ai looks for the similarities and the human being has to look for the differences right? The a I is like my dog while now. Why Oh if you take them to the dog park.

04:01.88

satvcpodcast

Ah.

04:14.61

penny

He has his ball and you can throw that ball and he'll go for that ball and you can pick up another ball and throw it and he'll look at you like where the hell is my ball I want my ball and if he loses his ball and you say go find your ball. He'll look at me and then he'll look around and he'll start sniffing. Every single ball in the dog box until he finds his ball and then he goes back and he's all happy because it's his ball. He is driven. He is partdasi pot border calling he is a driven dog and that's what Ai is it's driven as a goal in mind it is. Totally logical in its goal. It does exactly what it's told and it will do what it's told until you stop it just like why it's like he'll fetch that volunteer go get the but leave me a what I want to just exactly.

05:03.87

satvcpodcast

Ah, and and do what it's trained to do isn't it and that's the same thing with machine machine learning at the end of the day ai is glorified machine learning the machine has learned how to code that transaction from Office depot.

05:10.41

penny

Yeah, it's.

05:16.54

penny

aybe something that was over $: 1000

05:34.10

satvcpodcast

Still office expense. Yep.

05:35.93

penny

And it just got out into an expense and there was no question as to um, do we have to put this as an asset nobody asked that question and then six months out it's like oh man, why didn't you say something well because the computer does it know the context of it. They don't understand that can. Context now there's a lot of ai out there that works extremely well. Um, but in this case, you can get really good Ai and you can have a really good gl application but you've got a small business owner who's going to spend their money any way they want. No matter what you say and they're going to use external and and non-linked cards and banks to spend money they have to do a business and then they're going to use business cards and business accounts to spend personal money. And that's you can you can talk to your blue in the face but you're always going to have some clients who are just never going to follow the rules but the computer is always going to follow the rules. The Ai is always going to follow the rules until you go in and train it to do something else.

06:40.60

satvcpodcast

Um, of the.

06:51.84

penny

So I could train this dog to go do something else, but um, be hard pressed to train that dog to make a decision on when it has to switch between the two linear items that I've told it to do either fetch the ball or go sit down or fetch the ball. And stop off fetch somebody else's ball I could probably try I could train it to fetch a different ball. But then now that's the ball it's going to go after it's not going to switch back to the other ball. The first ball easily so I can train Ai to do something and it will do it.

07:23.20

satvcpodcast

And and if.

07:29.90

penny

Until the cows come home and that's bad and the ugly of it.

07:32.99

satvcpodcast

and and I think that's but and I think that's what pract business owners practitioners need to recognize as they are trying to leverage machine learning and ai understand the limitations of it understand what it does really? well. And load it up on that type of work and those types of clients but understand that as your business clients as your business clients become a little bit more sophisticated or as they become a little bit more nuanced because of the type of business or the type of industry or or just who they are. That there's going to be limitations and so rather than trying to fight it and trying to fight the the provider because or trying to get the provider to do add the Xyz functionality to the product just recognize that this. Product is really good at coding transactions in this particular industry for this particular type of client and be okay with that.

08:29.90

penny

Yeah, and then also you have to you have to look at it as I'm just not gonna let this this system run and never go look at it. You do have to look at yeah so you have to provide the context you have to provide.

08:39.65

satvcpodcast

Yeah, yeah.

08:48.83

penny

Picture it can paint by numbers. No problem but somebody's got to actually put that picture in a shape and break it up into numbers and make the colors um in a particular pattern that makes it a a likable picture that anybody can understand.

09:05.95

satvcpodcast

And that's a big issue. Yeah and and.

09:08.30

penny

So that's your job Now your job is ah to type stuff in your job is to look at what a computer just did and go is this accurate appropriate and how does that affect the story of this particular file.

09:24.68

satvcpodcast

And and that's a big challenge as as firms are coming into tax season isn't it because they've got this massive compliance ah swear jar. Um.

09:29.26

penny

Fuck yeah, there's no oh sorry? yeah ah no one that drop out of the but you know what it's the last two weeks in January so I know exactly how they feel. Yeah.

09:41.30

satvcpodcast

And and and so that's the thing is is basically they're coming into massive compliance over the next three four five months and then recovering from massive compliance. Um, they don't necessarily have the capacity to go in and do that check and so. 3 hree and four months can go by with without that books review which um, which means that a lot of mistakes errors issues can can arise over the coming coming few months

10:09.99

penny

So when do they arise because you haven't looked at the books for several months they arise all in January. What are we most stressed about January. It's probably 1 of the most especiallyly for when it comes from bookkeeping. It's it's extremely stressful month because you've got year end. You've got ten ninety nine s you got w 2 wo's you got payroll. You've got a lot more sales tax reports to put out than you usually do because you got all your annuals as well as your monthlyes and quarterally everything comes to an a head in January and so you're stressed to begin with.

10:43.68

satvcpodcast

And you're probably signed on or added a new application in November December so you'll learn so you're learning that try to trade the t one that? yeah.

10:48.90

penny

Oh yeah, all you decide you decided to outsource in December. Yeah, let me oh how many of those happened this book this month. but but you, you it's like. If you could if you could just kind of load level your work and this is where workflows tasking workflows come in so handy as you can load level that work so you can take 5 minutes every month to just review what the computer did um. And take 5 minutes to review what the computer failed to do and and then then you can evaluate not only the client you can evaluate your ability or your employee's ability to deal with this. You can evaluate. The application that you're using for tasking. You can also evaluate the application that you're using to get that Ai and instead you're waiting until the last of the year and believe me I'm just as at fault with this as anybody else. Um. But you wait till the end of the year and it's like wow you know that really didn't work very well and then you want to change it which I have just just today alone told 2 clients. That's really interesting I'm going to fix that. But I am not going to evaluate or think about a new way to do it until the second week of february do you understand why and they'll go. Oh yeah, I'm sorry we shouldn't have asked you right now like fuck me right? You shouldn't have asked you right? Yeah, but.

12:20.27

satvcpodcast

Are.

12:26.80

satvcpodcast

But but building on the last episode we did um make a note in your tasking system put a deadline against it and then forget about it until until February when you said you were going to revisit it.

12:41.20

penny

Which is exactly which is which is exactly what I did I went into my tasking on my phone I was walking my dog because both of those calls came in while I was walking the dogs and I like okay pulled it up on my mobile phone set a task for myself two weeks from now. Here is what you need to research and come up with a solution for um, just don't intend.

13:00.60

satvcpodcast

Um, I think the other thing is I think I think it's also a tremendous training opportunity for juniors for new people coming in. They don't need to do all the key bashing to get the data in there. Let the technology do that but but having them do the monthly review. Um, is a wonderful training opportunity because I think they can have a look for those anomalies and and what I used to do back in my bookkeeping days and we were actually entering it manually. But the most important thing was you'll just do a month on month comparison. Has something gone up significantly is is something appeared in ah in a chart a chart of accounts line that that's never been there before or has the number jumped up by you know more than 5% and if if that has happened if we've got a new line item in a chart of accounts that. Probably means that I've coded it incorrectly. So let me go back and have another look there because I haven't had that type of expense in in the chart of accounts for the last three months and then alternatively if it's more than yeah, five ten percent on the last three months then it's probably means that. Ah. Might not mean anything but certainly worth me going back in and having another look at it and see if I've coded it correctly do I need to split it is it a capital versus an a capital expense versus an expense so that's what I did sort of manually once I'd entered all of the data to make sure I'd coded it correctly. So now that the tech the the ai gets it. Into the the gl. It's it's a a fairly simple review process as you said could be 5 minutes but you've got to do that review process to make sure that the training was correct and if the training wasn't correct. You can then correct it.

14:45.37

penny

Yeah, and and there's also I mean ah and Ai didn't do all of it. Yeah I mean you're go have to even as an account manager overseeing even bookkeepers you have to probably at least quarterly go in and check at minimum to see if the bookkeeper made the adjustments that should have been made and didn't miss it. You know it's it's interesting. We do a lot of different types of work in our office in Chinna we do tax returns. We do audit preps. We do internal administrative billing for firms. We do Quickbooks desktop online 0 nets sweet so we're doing a lot of different types of work that you would ordinarily see maybe specific to a particular type of firm or at least segmented into different silos in a firm and I have to tell you the most difficult. Job that we have is allocation. We could pull out a tax or every yeah and and my partners will tell this and everybody the worse for us. They'd all much rather be on the audit team or the tax prep team because there's no ambiguity there this is compliance driven. It's. Time driven by compliance. It's compliance driven by the rules and regulations of not only the government but the applications that are involved allocation at the bookkeeping level is like herding cats and um.

16:09.10

satvcpodcast

M.

16:20.32

penny

You can be correct on a transaction one month and incorrect on a transaction the next month So when a human being can do that a computer that is doggedly like why oh my dog looking for that ball hunt that ball but hunt the ball hunt the ball hunt the ball hunt The ball. It's always going to go in the same direction. Gonna make that same mistake over and over and over again till you correct it? Um, it's it's something that you have to look at how do you bring on new people. Um, how do you?? Um, teach them what to look for. Um, you know I It used to be that you would bring on a new person and they would like you said they would be typing the information in and they would do the best they can and then gave it to somebody to review Now. They're a traffic cop now they are looking at the technology is it sinking is it.

17:11.45

satvcpodcast

Um.

17:19.17

penny

If there's a mistake is it us or is I just got off of a conference call webinar with an app that we use and I found 2 mistakes that I went and I told them I said is this your system or is this the way we entered the data. Is this something that the client did because we're touching it. The client's touching it and the Ai's touching it and all 3 of those had some kind of um function within the the system. And there were 2 blatant mistakes and they were sitting there at at the top of this the view and I just I said I'm not I'm not ready to blame anybody on this because the computer touched it. Well the sopmore application touched it. We touched it and the client touched it.

18:09.21

satvcpodcast

Yeah.

18:16.00

penny

And so what's the best practices for dealing with that and they said well thank you for giving us the opportunity to look so they're gonna go look. They'll report back to me on what they think caused those 2 mine and ah and a slew of transactions why did were those 2 not dealt with properly. And we'll figure it out. Um, but nothing is perfect and you don't throw. Ah I'm not about to throw out this app just because I found 2 minor things and I'm not really ready to find out I'm I'm just interested to find out how it happened and how it can be corrected. Um, and I think that's something you have to take sit because a lot of times you'll see people go well I didn't like that application. It made this one mistake and I never tried it again. Ah you know and that's that happens and that's sad as.

19:04.17

satvcpodcast

Yeah, yeah, yeah.

19:11.74

penny

The other thing too is these apps. Also they grow up and as they grow up, they they take on more functionality and they they provide more changes based on the feedback that you give it just like you give feedback to the computer and it Lawrence you need to give feedback. To these app developers so that they learn what it is that you're really experiencing out there because they're writing their app in this nice little bubble and they know how it's going to integrate to the gl. But you're looking at the whole picture. Remember you're you're the one who did the paint by number for this client and you've got the whole context.

19:52.43

satvcpodcast

But also I think the other thing is we sit and and I think I'm a ah cause of this as well. But accountants and bookkeepers. They they sit in the conference room or on the virtual um on virtual Webinars, Etc, etc. And. And basically where we're jamming down their throats. The the potential of ai and machine learning and it'll do this and it'll do that and it'll do this this this and it'll make you a cup of tea and it'll spot this anomaly in that anomaly and I I think we I think we as app providers as developers probably paint a false. Or give a false narrative in terms of the full potential of it and so therefore accountants and bookkeepers would come up to us at the stand to be like well why can't it just do this this and this I mean you said it can do this so surely surely. You should be able to program it to do x y and z and and so I think. We we probably need to as developers and providers need to better manage expectations because the theory is yes, there's no reason why we couldn't develop the code to do x y and z but the the impact of doing x y and z on.

21:04.99

penny

Well.

21:05.24

satvcpodcast

A b c d e fg everything that's been developed beforehand. Um, that's that's what the great unknown is and and so I think there's I think we have to better manage expectations from a developer standpoint and and be honest about what Ai and machine learning can do now. Is likely to do in the next six to eighteen months and then just sort of better manage expectations because the number of phone calls I was on and the number of um customers or or prospective customers that would come up to us complaining that particularly like it doesn't recognize tax. It doesn't do this. It doesn't do that. Why not. Um, you said up on stage yesterday that it could do x y and z um, why can't it do why? can't it do that and so I don't know I think that's that's part of the the ai and machine learning conversation as well is probably better managing people's expectations because as you said the the. Capability and the feature set grows. Um, and unfortunately we do have limited resources and and that means just just like in yourself as as you add new skills to your skill set. You probably means you make a few more mistakes along the way. Um, with other areas as you're learning. So I don't know I think it's there's there's work to be done on the provider side to better manage expectations about what is actually achievable and certainly in the next six months the next eighteen months um so that I think accountants and bookkeepers. Ah, ah, better prepared. Ah, to implement this technology. Yeah.

22:39.43

penny

Yeah, cause it's not going away is it's well there was a um, a podcast that we both know David and Blake who talked about you know people that complained about desktop going away how they quickbooks desktop is going away in the U K and

22:51.13

satvcpodcast

Yep.

22:59.28

penny

Afraid it's gonna go away here. This stuff's not going to go away right? Um, it's gonna look different two or three years from now than it does right now. Um, you have to be adaptable. You have to be open to change. Um, and I think that that you have to also understand that as you're learning so is the robot and the robot is expecting you, whatever you want to call that robot ai whatever if it's if it's that little dog robot that walks across the floor. It's out even like even the rumba you know if it's you, you've got to teach it where to go and you've got to teach it how you wanted to navigate around your house. Um, it's the same thing with all of these um apps that we put on onto the computer and.

23:52.91

satvcpodcast

Um, yeah.

23:57.89

penny

And if you let it run its course for several months before you look at it don't be angry at the computer. Don't be angry at the app. Don't be angry at the client just recognize that everything has to be checked and when you're doing your flat rate. Um, billing. Whatever it is. You're doing put in there a cushion for the time that you're going to take to review then set a task in your workflow to do that review and then actually go in and do it and make the changes and adjustments. That will reeducate the ai and ah let you know whether or not that is a viable use of your time and money to use that ai because it's not cheap. It's not cheap. Once you add up a tech stack and you look at everything that's connected and at the end of the month how much are you paying? How much is your client paying.

24:47.24

satvcpodcast

Um, no.

24:55.50

penny

Um, it better be working for you and if it's not working for you. Get it rid of it but don't do it in December but yeah much and you yeah to call me up in January ago so what do you think we we think about switching like what are you crazy.

25:00.88

satvcpodcast

Oh January um.

25:11.33

satvcpodcast

We're thinking about switch it. No no just bear with it. Yeah, call it call me by month and and and most importantly, let's let's in in four or five months time let's actually look at the app. Let's look at the training. Let's look at what you want it to do and and do it properly and and.

25:13.91

penny

Um, me and 5 know fail with it.

25:27.99

satvcpodcast

Probably 7 times out of 10 it'll be the case that this app does exactly what you need and want it to do but um, maybe a couple of steps were skipped in the ah in the training and implementation process. Maybe maybe a bit more understanding of what it does and doesn't do. Um, yeah, but 7 times out of 10 it's it's the right application just probably not implemented correctly. so so fear not it's it's not about starting from scratch on a lot of those cases penny 1 question I have and and 1 thing for listeners. We've talked about niching or niching on a couple of episodes. If you're sort of looking at this controller type service and you're looking to scale that um, are there industries or business types that you've seen that do much better on the ai side of things so that. That the the machines do a good job of the transactional processing are there any industries that you see better than others are better than others.

26:27.60

penny

Um, is service Industries professional service. Industries is pretty good for scalability. Um I would say real estate development um is investment. No real estate investment.

26:32.99

satvcpodcast

Okay.

26:43.57

satvcpodcast

Those inside development.

26:45.84

penny

Good yeah real estate investment. It's a good vertical e-commerce is huge um, you have to pick the right apps for that because and you have to make a decision on what you want to go into the g o. There is a new product out there that I've been looking at that I think would make ecommerce much more of a scalable niche to somebody because you got to think about with like scaling scaling is huge. If you are going to do any kind of control it type services. You definitely want to niche it. You cannot be in an e-commerce business, a professional services business and a retail business and have 4 or 5 clients in each of those and think that you're going to not have a nervous breakdown. By the end of the first quarter. Um, so you want to scale it you want you want to scale it then you want to niche it if you want to niche it. You need to learn all of the best products for that niche and focus on that. Don't let don't let another niche distract you at this time until you get to the point where. On a learning curve. You went from um, learning stuff to wow I'm rocking and rolling this is really sweet to damn this is boring I've got it down now. It's time to get niche and find somebody else to take over that thing that you just got bored with because it was just rolling along now it's i'm.

28:07.46

satvcpodcast

Are.

28:17.32

penny

So you scale, but the only way you can do that is if you stick with 1 and the other thing you have to learn I find and this is just me I you know take it what you want, but if I'm in a niche I don't just learn the applications in accounting that help me in that niche. I need to learn my clients applications that they're using in their everyday business. How are they using technology. How are they communicating in their everyday business. So it's a much more holistic approach. Um, and.

28:40.85

satvcpodcast

Um.

28:55.95

penny

Um, I think that that has to be taken into consideration. So you're charging more because you're focused and you're niched and you're scaling a little bit slower but here's the thing is when I talk to clients. Very rarely asking me about an individual transaction in a gl they want to have a conversation that is completely consultive based and it happens if it happens to me as little as I do. Compared to a lot of bookkeeping and accounting firms out there if it's happening to me as often as it does I know it's going to be happening to them and if you don't take the time and put down everything and focus on what they're saying and think outside of accounting for the issues that they're having you're leaving a lot on the table. You have to be aware of the industry that you have niched in you have to be aware of their external presses and what's what's going on in that industry. You really have to spend a lot of time learning about their applications outside of accounting and. Be well roundunded and well versed in that niche and that takes time It's not something is gonna you know you're not gonna have the proverbial 2 or 300 clients and you're doing the monthly bookkeeping and you're calling yourself controller. No, you know you're doing often. Fact, right up, you know it I know it.

30:09.46

satvcpodcast

Um, yeah.

30:24.70

satvcpodcast

And and and and also 2 and 300 clients you've got a pretty impressive team to to be handling that type of volume of work whether it's onshore or offshore I don't care because that's that's a lot of client management. That's a lot of client relationship management.

30:25.00

penny

So shut up good.

30:39.51

penny

Oh got here.

30:42.99

satvcpodcast

and and and I sort of think we we sort of say that the term scale your accounting practice and and I don't mean that in terms of 2 and 300 clients I mean that in terms of scaling it using the technology. So it's less reliant on you so you could you could. Add 2 and 3 and 4 clients if you wanted to if you needed to but but you you got the flexibility to do that when you want to do it rather than oh shit I've got to go out and hire another person before I can actually bring on a new client. That's what I mean by scale. It puts you in control of that growth.

31:14.89

penny

Yeah, yeah.

31:19.30

satvcpodcast

So that I know a couple of firms that I've worked with they started. They picked up a couple of microbreweries in the area. Those microbreweries talked to other microbreweries so they became customers as well. Um, and slowly but surely grew. Ah, a nice.

31:27.97

penny

Yes, they do yes.

31:36.32

satvcpodcast

Niche of microbreweries down the Pacific Coast as they got to a point that and then then they consolidated and then they were like okay, let's just focus on this.. What do we want to let's make sure we know this inside and out then they sort of reset and said Okay, what's next and they went. Um, adjacent so they went to sort of the distilleries and some of the startup distilleries and and that's that's where they're focused, but but they grew at the rate they wanted to and they were in control of that and it wasn't based on trying to find another body to to bash the keys.

32:09.63

penny

Yeah, exactly I mean you gotta you? it's you know it's great. What people do about selling the new way of doing business but businesses grow the same way they might grow like. Incrementally quicker or they might die incrementally quicker but you still the the process of development of learning and of growing hasn't changed since human beings first crawled out of a cave so that process is still there. How. We can speed it up or slow it down. You have to be in control of and when you have multiple different types of businesses with multiple different types of needs. That's really hard to scale at a controller in a controller ship and a virtual controller ship and. Fractional controllership never mind just you know going into those offices. You know what it would take back in the day when you did go into the client's office and do a couple hours here and then go to the next client and do a couple hours there. Well you doing the same thing. Just doing it using technology and using the cloud to have access stuff and um, you still have to think about the learning curve. You still have to think about the context you still have to direct the traffic. Um, whether that traffic is being done by another person in your office. Or somebody offshore or by a robot Ai you still have to be directing it and looking at it and give yourself that time every month to review it. Otherwise you never gonna scale because you're always gonna be fighting a crisis somewhere.

33:55.23

satvcpodcast

Yeah, and it somewhere and and it took it took me a ah while to realize that when I was at Receipt bank as well because my feeling was the artificial intelligence and machine Learning. It was the great savior for the for the accounting industry. Um, it took me a while but ah, but I slowly became very comfortable with saying actually receipt Bank's not a good fit for that that client all of these anomalies and all of these exceptions that you're talking about Receipt Bank's not a good fit for that. How many of those types of clients. Do you have are 2 or 3.

34:28.82

penny

Yes.

34:34.29

satvcpodcast

you know a nice pocket of of: 1520

35:24.61

penny

Being comfortable with saying no when a client comes through the door. A potential client comes through the door and it's the end of the year and they're just saying hey listen I need a bookkeeper I need somebody to get all my books clean so I can turn it over to my tax account and could do it by tomorrow and.

35:37.41

satvcpodcast

Yeah, yeah.

35:42.64

penny

You know? and and unfortunately I I see this and I'm kind of surprised but I do see accountants doing that and I know that they're being empathetic but you're really not being. You're being kind of disingenuous to yourself and to that Person. Who's coming in I think you're better off saying look at I could do this while you but I cannot do it in the time frame that you want nor in the and the price range that you want So This is what I can do think about it and come back to me and then move on because don't try to fight I've had that happen.

36:15.47

satvcpodcast

Yeah, and.

36:20.50

penny

Twice this month where I just said you know? Yeah well this is what it will take and this is when we can get started. So if that doesn't work for you I'm sorry. But yeah.

36:29.36

satvcpodcast

And have have a think about it and and let me call you in February once I've got January behind me and I've got a bit of space to think about it and then set a task reminder follow up with prospect and forget about it.

36:39.33

penny

Um, wall. Yeah, yeah, yeah, because yeah, it's it. This is not at this time of year. It's not worth it to to to to try to be a super person think world. You've got to just.

36:52.47

satvcpodcast

Yeah, absolutely.

36:57.61

penny

At this point we're all just saving ourselves. But um, you know context is everything and don't blame the Ai don't blame the bookkeeper. Don't blame the business owner just understand that.

36:58.12

satvcpodcast

Um.

37:16.40

penny

Shit happens and as one is as 1 gentleman and and El Paso Once told me when I was there at his firm he was sitting down having a glass of wine after a 2 day training course that we did down there and he said you know nobody ever died because their books weren't balanced one day.

37:33.71

satvcpodcast

Um.

37:35.81

penny

And ah, you know what? That's a good way to look at it. He was very relaxed about it and um and he had a good company had a good firm and he had good. He had what he really had was good loyal employees too and you could tell us because he had an attitude as we'll get to it. It's going to get done. This will be Fine. Don't worry about it and and you could tell in his employees that they were very relaxed and they were very happy to be working there and they were been longtime employees. They were very they were gonna stick with him That's for Sure. So Just you know context is everything. Ai is a help. It's just a tool.. It's just a tool and it's a dog that needs to go do its job and you have to redirect it once a while once long.

38:21.27

satvcpodcast

Once in a while and I really liked what you said there in terms of you got on the phone with the app partner and said look we've been in there. The client's been in there. The technology's been in there. There's a mistake and we just want to get to the bottom of it and I think that's a really um, a really good way of approaching. It is is. Ah, couple of hands have been in here and and let's get to the bottom of it so we can fix it so that we can retrain it so that we don't have to have this conversation again I think that was really really powerful.

38:48.92

penny

Yeah, well, it's a good app I'm not going to throw the baby out with the bath one. Yeah, that's.

38:52.94

satvcpodcast

Yeah, right? Penny I think that's a really good spot to leave today's episode Ai in context, um, fascinating conversation. Thank you very much ladies and gentlemen if you have enjoyed what you're listening to do us a favor jump on Apple Podcasts leave a review for us. We'll share that review. Ah, in our next episode and then also if you do have any questions for us or if you'd like us to help out with a particular situation feel free to reach out reach out on Linkedin and drop us a line connect with us there but penny always a pleasure. And until the next episode of strategy in the virtual controller we will see you next time.

39:29.39

penny

All right - Thank you have a good one.