
AI Is Only Useful When It Can Reach Your Data
· 1h 3m
0:001:03:29
About this episode
Your business already has more AI-ready data than you think—but most of it is trapped in QuickBooks, Google Workspace, phones, laptops, and half-clean systems that can confidently fool an AI into bad answers. In this episode, Mitchell and David break down how small businesses can inventory the data they actually own, clean it up, and turn it into something AI can use without creating chaos.They start with the practical question most teams skip: What questions do you want to ask AI, and what data do you need to answer them? From there, they map where data really lives—email, phones, ERP systems, inventory tools, employee devices, and even the inexpensive software accounts businesses rely on every day.Mitchell explains why a “source of truth” matters, why paid business accounts and company-owned devices are increasingly important, and how regulations like GDPR and CCPA shape your right to retrieve and control your data.
The conversation also explores the bigger shift happening right now: AI is moving fast enough that yesterday’s best practices can quickly feel outdated. Mitchell and David explain why the cloud is often the safer place for operational data, and why forcing AI onto old desktop software or DIY infrastructure can create more risk than value.
Transcript18,096 words — click to read
TRACK02: before you make any more data, find out what you're already sitting on and whose it is. You already have more data than you think. It's in your accounting software, your project software, your email, your phone, and in a C R that some vendor is holding for you. Most of it's a mess, right? You've got duplicates, dead contacts, three spellings of the same item in inventory. Yeah, feed that to an AI and it'll confidently
TRACK01: any more data, find out what you're already sitting on and whose it is. You already have more data than you think. It's in your accounting software, your project software, your email, your phone, and in a CRM that some vendor is holding for you. Most of it's a mess, right? You've got duplicates, dead contacts, three spellings of the same item in inventory. Feed that to an AI and it'll confidently
TRACK02: Tell you something wrong. It doesn't know the difference. That's right. And so today we're gonna
TRACK01: tell you something wrong. It doesn't know the difference. That's right. And so today we're gonna we're gonna help you find out what you've got, how to get the most out of it, and how to clean it up. Make more by losing less. Answers are everywhere.
TRACK02: get the most out of it and how to clean it up. Make more by losing less. Answers are everywhere.
TRACK01: And we're here for the questions.
TRACK02: Mitchell, today this this is really your wheelhouse. I mean, you're you're the data guy, you're the you're the wiring guy, you understand the the technical build aspect of
TRACK01: Mitchell, today this this is really your wheelhouse. you're you're the data guy, you're the you're the wiring guy, you understand the the technical build aspect of
TRACK02: of the things we've done together and and the things that you've done in your past as a as a software engineer. So this is really heavy engineering.
TRACK01: of the things we've done together and and the things that you've done in your past as as a software engineer. So this is really heavy engineering. but we're gonna kinda keep it light, right? We're gonna For sure, for sure. We're gonna make this for the small business owner out there who who's trying to figure out how to the
TRACK02: but we're gonna kinda keep it light, right? We're gonna we're gonna make this for the small business owner out there who who's trying to figure out how to th they know they have data. Sure. Everybody's got it. Well, some people don't know they have it, but they probably have it. It's it's it's in their devices, it's in their software. they just need to get access to that data.
TRACK01: They know they have data. Sure. Everybody's got it. What some people don't know they have it, but they probably have it. It's it's it's in their devices, it's in their software. they just need to get access to that data and translate that access to the AI so that the AI can help them interpret that data better.
TRACK02: and translate that access to the AI so that the AI can help them interpret that data better. So what's the first step? Like the inventory, the I I would say the first step really is to figure out what questions you want to ask. Because that's gonna determine one, what data you need, but how you need to go about accessing that data.
TRACK01: So what's the first step? the inventory, the I I would say the first step really is to figure out what questions you want to ask. Because that's gonna determine one, what data you need, but how you need to go about accessing that data. But even zoomed back from that, it it is important to to know what you have generally, right? You need to have a data inventory, just like you have an inventory of tools and equipment, you need to know what data you have and where it lives and and what your ability to access that data is. So what are those kinds of places? Where where does data live?
TRACK02: But even zoomed back from that, it it is important to to know what you have generally, right? You need to have a data inventory, just like you have an inventory of tools and equipment, you need to know what data you have and where it lives and and what your ability to access that data is. So what are those kinds of places? Where where does data live?
TRACK01: I'd say the the most important ones for a small business owner are gonna be any of your communication tools, so email, phones, things like that, and then any of your financial tools. so a QuickBooks type tool, any of your inventory management software, your your enterprise resource planning, your ERPs, those are gonna be your core locations.
TRACK02: I'd say the the most important ones for a small business owner are gonna be any of your communication tools, so email, phones, things like that, and then any of your financial tools. so a QuickBooks type tool, any of your inventory management software, your your enterprise resource planning, your ERPs, those are gonna be your core locations.
TRACK01: followed by, and this is a little bit scary to think of, but any employee devices. You know, how much how much data sits on everybody's laptop, how much is on your field leader's phone. but the big ones are gonna be those those monthly bills, right? The things you're paying a few hundred dollars a month to have the right to use.
TRACK02: followed by, and this is a little bit scary to think of, but any employee devices. how much how much data sits on everybody's laptop, how much is on your field leader's phone, but the big ones are gonna be those those monthly bills, right? The things you're paying a few hundred dollars a month to have the right to use.
TRACK01: That's probably where your most valuable data for the AI is.
TRACK02: That's probably where your most valuable data for the AI is. So things like your financials, customer information, exact job information, inventories, anything like that that you have that's either on your laptop or a server in your closet or maybe out there on the cloud.
TRACK01: So things like your financials, customer information, exactly, job information, inventories, anything like that that you have that's either on your laptop or a server in your closet or maybe out there on the cloud. And so when you make those, when you sit down and you start taking those inventories
TRACK02: And so when you make those when you sit down and you start taking those inventories, I think some people are probably gonna find out pretty quick that they're gonna wish they had maybe another tool or something that could consolidate all that data into one place that or at least get
TRACK01: think some people are probably gonna find out pretty quick that they're gonna wish they had maybe another tool or something that could consolidate all that data into one place that or at least get
TRACK02: we call it a source of truth. So let's let's talk a little bit about what what is the source of truth and
TRACK01: We call it a source of truth. So let's talk a little bit about what what is the source of truth and and that concept of a source of truth. so initially right, I'm I'm assuming that the organizations that we're talking to are n they're not organizations that are gonna gonna invest tens of thousands of dollars in data warehouse products and things like that.
TRACK02: And that concept of a source of truth. so initially right, I'm I'm assuming that the organizations that we're talking to are n they're not organizations that are gonna gonna invest tens of thousands of dollars in data warehouse products and things like that.
TRACK01: so that source of truth, really you start building that out and it's it's a little bit of a piecemeal game.
TRACK02: so that source of truth, really you start building that out and it's it's a little bit of a piecemeal game.
TRACK01: but you know, you're you're gonna determine and say, okay, for all money related data questions, QuickBooks is the source of truth. For project related questions, you know, in our case it's a job tread, so job tread is the source of truth.
TRACK02: But you're you're gonna determine say, okay, for all money related data questions, QuickBooks is the source of truth. For project related questions, in our case it's a job tread. So job tread is the source of truth. And job tread is an ERP and Job Tread for people not in the construction industry, job tread is it is like a Microsoft dynamics for the construction industry.
TRACK01: Yeah, Jobtread for people not in the construction industry, Jobtread is it is like a Microsoft dynamics for the construction industry. It's a project planner. and some people might even just be using spreadsheets. Sure. that's I mean, businesses were ran for a long time with just a a a ledger and a notebook and then spreadsheets, and so that could also be your source of truth.
TRACK02: project planner. and some people might even just be using spreadsheets. Sure. that's I mean, businesses were ran for a long time with just a a a ledger and a notebook and then spreadsheets. And so that could also be your source of truth.
TRACK01: But for the modern business today, unfortunately, there is no there is no single source of truth.
TRACK02: But for the modern business today, unfortunately, there is no there is no single source of truth.
TRACK01: but thankfully, you know, regulation has its its ups and downs. regulation actually is the reason you can go get your data today. Policies like GDPR and C C PA in Europe and the United States are basically require that any company that hosts data for you or about you, they have to provide you a means to retrieve that data from them so that you can leave and it's not as relevant, but you have the right to be forgotten under those laws. So you could call QuickBooks, maybe you can't call them, but you can initiate a process, get your data and then tell
TRACK02: retrieve that data from them so that you can leave and it's not as relevant, but you have the right to be forgotten under those laws. So you could call QuickBooks, maybe you can't call them, but you can initiate a process. You could try. Get your data and then tell
TRACK01: the Intuit Corporation, forget everything you know about me and they have to comply. And it it's a huge technical lift. Like one regulation creates millions of jobs probably because
TRACK02: the Intuit Corporation, forget everything you know about me. Mm-hmm. And they have to comply. Interesting. And it it's a huge technical lift. Like one regulation creates millions of jobs probably because Because now you have to prove that. Right? If if I ever go back and I say, Hey, into it, did you really delete my data? And they didn't, now I've got a case. Right. I see. So inside I guess that's in a big
TRACK01: now you have to prove that. Right? If if I ever go back and I say, Hey Intuit, did you really delete my data? And they didn't now I've got a case. Right. I see. So inside I get that's in a big
TRACK02: that's kind of outside the organization view of ownership of data. Inside your organization, how do you how do you decide what is your data, what's your
TRACK01: that's outside the organization view of ownership of data, inside your organization, how do you how do you decide what is your data, what's your maybe your employees data, your clients' data? How how does all that work out? How do how do you how do you rein all that in?
TRACK02: maybe your employees data, your clients' data. How how does all that work out? How do how do you how do you rein all that in as a business owner to become the property of the business? And when I say property, I don't mean I own it, which you do, but ha part of property is being able to manage it and maintain it and curate it and those sort of things. But to do that, you kinda have to
TRACK01: as a business owner to become the property of the business. And when I say property I don't mean like, I own it, which you do, but like ha you know, part of property is being able to manage it and maintain it and curate it and those sort of things. But to do that you kinda have to
TRACK02: own it. and know what to do with it. You do and that and that that goes back to you need that inventory. within, you know, next to your AI constitution it may be a good idea to have what's commonly called a data access policy. Mm-hmm. you know, to kind of ground that in a real world for us. we have financial data, we have AI agents and staff that work with that financial data, but we also have staff digital and human who
TRACK01: yeah. You do and that that goes back to you need that inventory. within, you know, next to your AI constitution, it may be a good idea to have what's commonly called a data access policy. you know, to kind of ground that in a real world for us. we have financial data, we have AI agents and staff that work with that financial data, but we also have staff digital and human who
TRACK02: don't work with that data. And so our data access policy makes it clear who within the organization who owns access control and who those primary people are. And I would extend the same rules. the same rules with much more robust controls to the agents.
TRACK01: don't work with that data. And so our data access policy makes it clear who within the organization who owns access control and who those primary people are. And I would extend the same rules, the same rules with much more robust controls to the agents. Because a human can just be disinterested. Right? Like
TRACK02: Because a human can just be disinterested. Right? I don't care what's in your QuickBooks, so I agree with the policy. An agent might just get interested and if it has any inkling of ability, you'll find it in your QuickBooks. Didn't you and I just what did we just share a link about how these digital employees, these agents, they're really they f they live for bureaucracy.
TRACK01: I don't care what's in your QuickBooks, so I agree with the policy. An agent might just get interested and if it has any inkling of ability, you'll find it in your QuickBooks. Didn't you and I just well didn't just share a link of about how these digital employees, these agents, they're really they they live for bureaucracy.
TRACK02: they do. that they they
TRACK01: they do, definitely. They they
TRACK02: are existing because of bureaucracy and they thrive on bureau bureaucracy. Humans resist bureaucracy, they get frustrated with it, they'll stop working, they'll
TRACK01: are existing because of bureaucracy and they thrive on bureau bureaucracy. Exactly. resist bureaucracy, they get frustrated with it, they'll stop working, they'll They'll find a more efficient path whereas the agents will not. and that but it comes down to it really it's it's data access and to some extent,
TRACK02: Whereas the agents will not and that but it comes down to it really it's it's data access and to some extent and this depends on the industry that you're in, but you may need to further classify your data as far as sensitivity, and especially in terms of release, right? And this is something that an AI agent
TRACK01: And this depends on the industry that you're in, but you may need to further classify your data as far as sensitivity. and especially in terms of release, right? And this is something that an AI agent, a human may kind of implicitly understand. I should not share financial data with my customer. Right. An AI agent may say,
TRACK02: A human may implicitly understand I should not share financial data with my customer. Right. An AI agent may say sharing financial data would help the customer understand and and so you need to label your data not only who owns it or where it lives, but who and what you can do with it.
TRACK01: sharing financial data would help the customer understand and and so you need to label your data not only who owns it or where it lives but who and what you can do with it. and it sounds really complicated. Really this this comes down to as you start bringing agents in.
TRACK02: And it sounds really complicated. Really this this comes down to as you start bringing agents in be specific in what you're asking your agents to do. the more open ended the question, the more open ended the solution. I think that let's go back to the data sharing and access policy because I mean we spent an entire episode talking about the AI constitution.
TRACK01: Be specific in what you're asking your agents to do. the more open ended the question, the more open ended the solution. I think that let's go back to the data sharing and access policy because we spent an entire episode talking about the AI constitution
TRACK02: And
TRACK01: and
TRACK02: really the data access, I mean, we could probably do an entire episode on that policy alone and how that might be structured in different ways for different businesses. But let's just maybe hit the high levels of that. You know, the
TRACK01: Really the data access, we could probably do an entire episode on that policy alone and how that might be structured in different ways for p different businesses, but let's just maybe hit the high levels of that. You know, the
TRACK02: the first thing is you you need to notify your employees that you're accessing their data or your data that they are creating.
TRACK01: The first thing you you need to notify your employees that you're that you're accessing their data. Or data that they are creating. And that's that's that can be a very difficult conversation because you know, any given employee may view their email inbox as their personal like that's their email, that's their phone.
TRACK02: And that's that's that can be a very difficult conversation because any given employee may view their email inbox as their personal like that's their email, that's their phone. Mm-hmm. And and I'm I'm definitely an advocate of of just generally respecting people's autonomy and privacy, but within the context of a business
TRACK01: And and I'm I'm definitely an advocate of of just generally respecting people's autonomy and privacy, but within the context of a business, one, the business needs to be able to look at all the data that it generates, that the the efforts generate. and I think the best way to kind of help mitigate that, the or the stickiness in the conversation is
TRACK02: One, the business needs to be able to look at all the data that it generates, that the the efforts generate. and I think the best way to kind of help mitigate that, the or the stickiness in the conversation, is anything that you believe to be the business's interest or the business's data, full stop the business needs to pay for. meaning
TRACK01: anything that you believe to be the business's interest or the business's data full stop the business needs to pay for. meaning, you know, and this is not to speak on the legalities right of a contractor as an example, but if you have a an employee, that should be a company laptop. Okay. That should be a Google business workspace account. and and
TRACK02: and this is not to speak on the legalities right of a contractor as an example, but if you have a an employee, that should be a company laptop. Okay. That should be a Google business workspace account. and and the reason you want that to be a company laptop is so that you have, you know, just like a parent has ultimate authority over every possession and action a child takes
TRACK01: The reason you want that to be a company laptop is so that you have, just like a parent has ultimate authority over every possession and action a child takes, a business needs that same authority. And when you're relying on, you know, bring your own device policies or you're relying, and this is a much bigger the the bigger risk here is really r a business relying on
TRACK02: a business needs that same authority. And when you're relying on, you know, bring your own device policies or you're relying and this is a much bigger but the bigger risk here is really r a business relying on free products. And it's not to say you can't use a free product. That's a good point. But if you go to gmail dot com and you sign up for my business accounting department at Gmail
TRACK01: free products and it's not to say you can't use a free product. That's a good point. But if you go to gmail dot com and you sign up for my business accounting department at Gmail, you don't have what's called a a business administration agreement with Google. There's no contract that Google with Google that says Google is gonna take care of your data. that Google is gonna make sure your data is always available, that Google's not gonna read or share your data. And so you want these paid accounts and owned equipment so that you can can just access it. And those paid accounts give you that access. Like you can't just go export Gmail. If you're not paying for it, then
TRACK02: And so you want these paid accounts and owned equipment so that you can can just access it. And those paid accounts give you that access. Like if you just go export Gmail. If you're not paying for it Then you are the product. Definitely. Yeah. And and
TRACK01: you are the product. Definitely. And and I mean
TRACK02: it's not that you're not the product with a Google Enterprise. Yeah, but it but there's guardrails there. Well, there's a lot of guardrails. Constraints and rules and and and those controls you're gonna need the controls that those paid accounts give you in order to make use of your data in order to do proper
TRACK01: It's not that you're not the product with a Google Enterprise account, but there's guardrails there. Well, there's a lot of guardrails. There's rules. And and those controls you're gonna need the controls that those paid accounts give you in order to make use of your data in order to do proper permissions. So so that I think is the first step. The business
TRACK02: missions. So so that I think is the first step. The bus anything that is the business interest should be financed by the business so that the business just has authority. Yep. Makes sense. those would be the two biggest ones.
TRACK01: Anything that is the business interest should be financed by the business so that the business just has authority. Yep. those would be the two biggest ones. And don't skimp out on you know, assuming you have an otherwise successful business. Don't let like
TRACK02: and you know don't skimp out on you know assuming have an otherwise successful business. Don't let You can't have this mentality that software isn't worth two hundred dollars a month. Mm-hmm. Like do the math, figure out what your employees hourly rate is if they had to do the job instead of the computer and you know, job tread as an example, quick books as an example. Is it expensive? Yes.
TRACK01: you can't have this mentality that software isn't worth two hundred dollars a month. Right? Like do the math, figure out what your employees hourly rate is if they had to do the job instead of the computer and yeah, you know, job tread as an example, QuickBooks as an example. Is it expensive? Yes.
TRACK02: Is it worth it? Yes. Mm-hmm. Hundred percent worth it. yeah. Yeah. Especially now that we're you know, we are sitting here talking about
TRACK01: Is it worth it? Yes. Mm-hmm. Hundred percent worth it. yeah. Especially now that we're you know, we are sitting here talking about implementing AI. I mean we're we're getting to that. Yep.
TRACK02: implementing AI, we're we're getting to that. if you plan to if if you plan to be a successful business in the next five years, you need to be thinking about how to use AI in your business. And a huge part of using AI in your business is the data. So getting your arms around all of this data and getting it organized and
TRACK01: If you plan to if if you plan to be a successful business in the next five years, you need to be thinking about how to use AI in your business. And a huge part of using AI in your business is the data. absolutely. Getting your arms around all of this data and getting it organized is just
TRACK02: Is just the natural step to implementing AI. you should be doing it anyway.
TRACK01: the natural step to implementing AI. you should be doing it anyway. You should be, but it's a requirement, right? Because a a chat GPT is an example. It's not that you can't take today's business question to a chat GPT and ask it and get an answer. But that chat window doesn't one, it doesn't know what happened yesterday.
TRACK02: It's not that you can't take today's business question to a chat GPT and ask it and get an answer. But that chat window doesn't one, it doesn't know what happened yesterday. It doesn't know anything about you. It doesn't know anything about your finances. And you know, a a lack of data. You know, it's it's kind of the same problem over and over.
TRACK01: It doesn't know anything about you. It doesn't know anything about your finances. And you know, a a lack of data. You know, it's it's kind of the same problem over and over. The more open ended you are, the more generic or vague you are, what you put in is what you get out. I mean it knows what you tell it. It it knows th the AI will know
TRACK02: More open-ended you are, the more generic or vague you are, what you put in is what you get out. it knows what you tell it. It it knows the the AI will know the information you give it. If you are giving it, you know, a three sentence prompt, that's what it knows. If you are telling it, you know, maybe you have a profile, if you have a paid account, you have a profile.
TRACK01: the information you give it. If you are giving it, you know, a three sentence prompt, that's what it knows. If you are telling it, you know, maybe you have a profile, if you have a paid account, you have a profile. Maybe you filled that out, maybe you didn't. I've I've spent a lot of time on mine filling it out. And so it knows about me. It knows what the things that are important to me. It knows about my my core values.
TRACK02: Maybe you filled that out. Maybe you didn't. I've I've spent a lot of time on mine filling it out. And so it knows about me. It knows what the things that are important to me. It knows about my my core values that
TRACK01: that I've put in there. It knows how I think about my business, my vision for my business, it has description of my business, it has a link to my website, it has all kinds of things
TRACK02: I've put in there. It knows how I think about my business, my vision for my business. it has a description of my business. It has a link to my website. It has all kinds of things.
TRACK01: that are context for the AI whenever I ask it any question. And then
TRACK02: that are context for the AI whenever I ask it any question and then
TRACK01: you take and add these other layers we're describing.
TRACK02: you take and add these other layers we're describing right now, that's the context it needs to make more precise to give you more precise answers and ask you better questions. Definitely. And it it is getting easier, so I don't I I don't want anybody listening to this conversation to think that I'm over here describing what is eventually going to become a substantial IT effort within your organization.
TRACK01: right now that's the context it needs to make more precise to give you more precise answers and ask you better questions. Definitely. And it it is getting easier. So I don't I I don't want anybody listening to this conversation to think that I'm over here describing what is eventually going to become a a substantial IT effort within your organization. It's not that it can't become that, but as
TRACK02: It's not that it can't become that, but as
TRACK01: as the AI industry has moved forward, things like C P, which just is a way for you to connect your cloud or your chat GPT to other services, these connections are making it easier and easier to go directly from, say, QuickBooks to a chat.
TRACK02: as the AI industry has moved forward, things like MCP, which just is a way for you to connect your claw or your chat GPT to other services, these connections are making it easier and easier to go directly from, say, QuickBooks to a chat.
TRACK01: whereas a year ago, right, when we started implementing at High Brighton, it was
TRACK02: Whereas a year ago, right when we started implementing it was none of that stuff existed and the game was it was a very monotonous game of go out, pull the data down, save it locally, massage it, get it into a usable shape, and then feed it to the AI so that we could ask questions.
TRACK01: None of that stuff existed and the game was it was a very monotonous game of go out, pull the data down, save it locally, massage it, get it into a usable shape, and then feed it to the AI so that we could ask questions. And now you can just go into a chat GPT and click a button and say, Connect to my QuickBooks. Connectors. Yep. And this is where this is another one of those areas where you
TRACK02: And now you can just go into a chat GPT and click a button and say connect to my QuickBooks. Connectors. Yep. And this is where this is another one of those areas where you really, really need as a business, you need to be on these paid accounts with things. Not that QuickBooks has a free account. but let's say QuickBooks did have a free account, they likely would not allow that that connector to work on the free account.
TRACK01: Really, really need as a business, you need to be on these paid accounts with things. Not that QuickBooks has a free account, but let's say QuickBooks did have a free account, they likely would not allow that that connector to work on the free account. Whereas on a paid account, that connector can work. And where that starts to tie into that data access is on a say a a free account or you only paid for one account and everybody in your office shares your one QuickBooks or your one whatever, now you can't give the office administrator and their AI properly scoped access into data. Whereas if you're in our case we we pay QuickBooks for the privilege of using their software. And so when we connect the connector, we can connect as the the controller, we can connect as me, the IT admin. and get that individualized access and get audit trails. Mm-hmm. Right? wow. it's it's a huge deal. And especially nowadays with the agent, you know, can make one, it's gonna make mistakes very confidently. And it's gonna make those mistakes at high speed if it makes a mistake. And so
TRACK02: Audit trails. Mm-hmm. Right? wow. it's it's a huge deal. And especially nowadays with the agent, you know, can make one, it's gonna make mistakes very confidently. And it's gonna make those mistakes at high speed if it makes a mistake. And so
TRACK01: you would want an audit trail and that tends to come with these paid accounts. They can tell you, an agent operating on David's behalf deleted all the data. And and see and I think that gets us into the
TRACK02: you would want an audit trail and that tends to come with these paid accounts, they can tell you, an agent operating on David's behalf deleted all the data. god No. And I think that gets us into the
TRACK01: you know, that data ownership discussion. Well you know we talked
TRACK02: Well, we talked really early on, I think, when we were first looking at our data at at High Brighton, we were we wanted to make everything litigation grade. You remember that? We were using that terminology.
TRACK01: We were we wanted to make everything litigation grade. Remember that?
TRACK02: And the idea, I mean, as a contractor, it implies you make contracts, and everything needs to be documented to a
TRACK01: And the idea I mean as a contractor, it implies you make contracts and everything needs to be documented to a a level that is litigation grade. In other words CY at all times. If the worst case scenario occurred, you would have a preponderance of evidence to show your work.
TRACK02: a level that is litigation grade. In other words, if if the worst case scenario occurred, you would have a preponderance of evidence to show your work. You would be able to say, we did it this way, we did it at this time, you know, the audit trail. It comes into play. And so if if your audit trail is, yeah, if if your audit trail's spotty, you are exposed. You're
TRACK01: You would be able say, We did it this way, we did it at this time. the audit trail. Absolutely. It's a change request. Yeah, if your audit trail's spotty, you are exposed. You're from a liability standpoint. And you're not
TRACK02: From a liability standpoint.
TRACK01: able to act with quite as much you know, situational intelligence, right? If if conversations are happening but nobody's taking notes, nobody's recording it.
TRACK02: situational intelligence, right? If if conversations are happening but nobody's taking notes, nobody's recording it. If receipts are being handed off but they're not being put into QuickBooks
TRACK01: if receipts are being handed off but they're not being put into QuickBooks. there is a window of time during which you can capture. And if you don't do it, now you've got a big hole. You know, it'd be like studying history with just bunches of the pages ripped out of the book. And that that's that's why you need to be you need to be thinking about capturing and saving and storing your data.
TRACK02: there is a window of time during which you can capture. And if you don't do it, now you've got a big hole. You know, it'd be like studying history with just bunches of the pages ripped out of the book. Right. And that that's that's why you need to be you need to be thinking about capturing and saving and storing your data. So let's talk about let's talk about data clean cleanliness now. It's just sort of how you how does one judge data and how does one
TRACK01: So let's talk about let's talk about data clean cleanliness now, just sort of how you how does one judge data and how does one
TRACK02: decide that this data is is dirty and I need to clean this data up and like let's talk a little bit about how how do how do you even begin to start judging data? I mean I would say it's a little bit it's like cleaning out your car. there's a certain
TRACK01: decide that this data is is dirty and I need to clean this data up and wait, let's talk a little bit about how how do how do you even begin to start judging data? I mean would say it's a little bit it's like cleaning out your car. there's a certain There's a certain gestalt, right, where you can say, I I know this is useless, right? you look in your downloads folder and you see a file that you downloaded four months ago. No, there's no value in it. No, you don't need to keep it. but it really it comes down to again, like cleaning out your car,
TRACK02: there's a certain gestalt, right, where you can say, I I know this is useless, right? you look in your downloads folder and you see a a file that you downloaded four months ago. No, there's no value in it. No, you don't need to keep it. but it really it comes down to again, like cleaning out your car.
TRACK01: you have to stop and look. you can't just say that's a plastic bag 'cause it might be a plastic bag with something useful in it. you know, when you start to look at like what is clean data, clean data is usually it's gonna mean that you don't have a bunch of duplicate records. and worse than a duplicate is an almost duplicate. Right. So I'm thinking about like in our own context,
TRACK02: usually it's gonna mean that you don't have a bunch of duplicate records. Okay. and worse than a duplicate is an almost duplicate. Right. So I'm thinking about like in our own context you know, what's the difference between a two by four by sixteen and a two by four by sixteen with the little quotes at the end. Nothing. Nothing. Okay. Except it's now a different line item in your inventory.
TRACK01: what's the difference between a two by four by sixteen and a two by four by sixteen with the little quotes at the end of it? Nothing. Except it's now a different line item in your inventory. it's it's a place for deviation to start happening. So let's say you have a process by which you reprice your materials every so often.
TRACK02: It's a place for deviation to start happening. So let's say you have a a a process by which you reprice your materials every so often. Mm-hmm. You want there to be one row that is a two by four. That's right. You don't want three different places and one of them never gets updated and then that gets used. And then the other two kinda get used once in a while and or depending on how Yeah. And and if the math matters
TRACK01: You want there to be one row that is a two by four. You don't want three different places and one of them never gets updated and then that gets used in a future proposal. Once in while or depending on how Yeah, and and if the math matters and you don't catch it, you know, what happens if you you price out a framing package on the price of a two by four that you got
TRACK02: and you don't catch it, what happens if you you price out a framing package on the price of a two by four that you got eighteen months ago instead of the price on two by four you got eighteen days ago. Right. It could be big difference. It definitely could be a big deal and so that's gonna be like that's data cleanliness.
TRACK01: eighteen months ago instead of the price on two by four you got eighteen days ago. Right. It could be it definitely could be a big deal. And so that's gonna be like that's data cleanliness. there's some technical nuances. sometimes data cleanliness is gonna mean you need to go through and you know remove extra comments in an SOP or
TRACK02: There's some technical nuances. sometimes data cleanliness is gonna mean you need to go through and you know, remove extra comments in an SOP or
TRACK01: you downloaded a an export of your data and it come back mangled. these are kind of technical details but
TRACK02: you downloaded a a an export of your data and it come back mangled. these are kind of technical details but clean data is data that has been looked at.
TRACK01: Clean data is data that has been looked at and it's been determined that it's worth having. It's been determined that it is what I'll call parsable. it has structure. It has it has some structure. You got some columns, you got some rows. Yep. And and sometimes that's like there there are different parts of the technology stack that are gonna be real sticklers for, you know, had a
TRACK02: And it's been determined that it's worth having. It's been determined that it is what I'll call parsable. Mm-hmm. so it has structure. It has it has some struct. Maybe you got some columns, you got some rows. And and sometimes that's like there there are different parts of the technology stack that are gonna be real sticklers for, you know, had a
TRACK01: an opening parentheses. If you don't have the closing parentheses, now you have dirty data. and a little bit of that is and it's gonna be understanding which parts of your system care. we're fortunate again today because these products are moving to the point the tech is being abstracted, right? If you remember the the early days of portable laptops.
TRACK02: an opening parentheses. If you don't have the closing parentheses, now you have dirty data. and a little bit of that is and it's gonna be understanding which parts of your system care. we're fortunate again today because these products are moving to the point the tech is being abstracted, right? If you remember the the early days of portable laptops
TRACK01: Wi Fi used to be like you had to have some deep understanding and some practice to just connect your device to Wi Fi. You need am I A, am I B, am I G, am I whatever? And over time, right, it used to be that you'd plug your router in and you had to type in weird numbers in your browser and go funny sounds. Yeah, and now you go buy a
TRACK02: Wi-Fi used to be like you had to have some deep understanding and some practice to just connect your device to Wi-Fi. Yeah, am I A, I B, I G, I whatever? And over time, right? It used to be that you'd plug your router in and you had to type in weird numbers in your browser, listen to some funny sounds, and now you go buy a router.
TRACK01: router at the at Walmart, you bring it home, you press one button, it dings your phone.
TRACK02: Put in a password, yeah. AI has done the same thing. and as the models, the the engines that run these AI services have gotten better and better, it's AI can do things two ways. When you're talking about data, it can either use code to process that data, read it, extract details, or it can do it as I'd call it in its head.
TRACK01: AI has done the same thing. and as the models, the the engines that run these AI services have gotten better and better, it's AI can do things two ways. When you're talking about data, it can either use code to process that data, read it, extract details, or it can do it, as I'd call it, in its head. And
TRACK02: And if you remember from the early days of of AI, the models were bad to lie. Right. Make things up fill in gaps or yeah. And that was a result of it doing it in its head. It was not
TRACK01: If you remember from the early days of of AI, the models were bad to lie. Right. Make things up, hallucinate. Yep. And that was a result of it doing it in its head. It was not it was like a student. It was a student that could only do math with a calculator. If it tried to do it in its head it got it wrong. More and more the student doesn't need the calculator anymore. And so some of these
TRACK02: It was like a student. It was a student that could only do math with a calculator if it tried to do it in its head it got it wrong. More and more the student doesn't need the calculator anymore. And so some of these the cleanliness matters less and less. Right. like you've seen Well, before you've seen AI find
TRACK01: the cleanliness matters less and less. like you've seen well before you've seen the AI say you know, find data
TRACK02: data errors or lack of cleanliness and say y would you like you know the AI would say, would you like me to clean this up for you? And that's a slippery slope. You could you could say yes and it might do a wonderful job.
TRACK01: errors or lack of cleanliness and say y would you like you know, the AI would say, Would you like me to clean this up for you? Mm-hmm. And that's a slippery slope. You could you could say yes and it might do a wonderful job.
TRACK02: It may do a perfect job. Or it might delete a bunch of your files and and and that's less and less the case because Yeah it is but in the sense of it making
TRACK01: It may do a perfect job. Or it might delete a bunch of your files and And and that's less and less the case because in the in the sense of it making the bad choice, the problem is the AI can still come to a legitimate conclusion through valid logic that the solution, for example, to global warming is to get rid of the humans. And
TRACK02: solution, for example, to global warming is to get rid of the units. Yeah. And data's very similar, right? we wouldn't have these errors if we just didn't process the data. Right. so it's it's less likely to do that, but it invokes the question of for a small business owner. Most business owner most small business owners don't have the depth of understanding of data analytics and
TRACK01: data's very similar, right? we wouldn't have these errors if we just didn't process the data. Right. so it's it's less likely to do that, but it kind of invokes the question of for a small business owner. Most business owner mo most small business owners don't have the depth of understanding of data analytics and data architecture to be able to look at a data set and say
TRACK02: data architecture to be able to look at a data set and say this I need this this is good, this is bad. Like they don't have that. So what kind of person do they need to to help them? I mean, they don't maybe they don't need an engineer necessarily, but they need somebody with a little maybe some engineering background. I would say definitely at least some
TRACK01: This I need this, I don't need this. This is bad. they don't have that. So what kind of person do they need to to help them? I mean, they don't maybe they don't need an engineer necessarily, but they need somebody with a little maybe some engineering background. I would say definitely at least somebody with a deep n maybe understanding is the wrong word, but a deep capability with say a Microsoft Excel. and the reason I say is you need somebody who if you can think in terms of tabular like rows, columns, that type of data. Basically everything is some variation of that problem. bookkeepers can look at it and say good bookkeeper
TRACK02: And the reason I say is you need somebody who if you can think in terms of tabular, like rows, columns, that type of data. You're probably pretty good at it. Basically everything is some variation of that problem. Most bookkeepers can look at it and say like if you're a good bookkeeper, you could you can look at data and you can say, this is great. This is you know, they they get they love
TRACK01: And you say, this is great. This is they they get it. They love they're usually tight based people. They like clean cleanliness. They like order. Yeah, definitely. but I would say it's you can avoid some of the risks. Again, I you know, it kinda goes back to make sure you have business level accounts with these products, because with these business level accounts you're gonna be able to do things like
TRACK02: They're usually type A people. They like clean cleanliness. They like order. Yeah, definitely. but I would say it's you can avoid some of the risks. Again, kinda goes back to make sure you have business level accounts with these products, because with these business level accounts you're gonna be able to do things like give the AI read only access.
TRACK01: give the AI read only access.
TRACK02: Mhm.
TRACK01: so say as a free user you let they give you access but there's no control. As an enterprise or business user you can say, Claude is not allowed to delete things from QuickBooks. That eliminates a ton of risk. and I would say that And I mean like we hear the horror stories, right? You hear you we're hearing it out there
TRACK02: so say as a free user you let they give you access but there's no control. As an enterprise or business user you can say, Claude is not allowed to delete things from QuickBooks. Mm-hmm. That eliminates a ton of risk. Mm-hmm. and I would say that A and I mean like we hear the horror stories, right? You hear you we're hearing it out there of like Claude deleted all my
TRACK01: Claude deleted all my Well it can't it can't do anything that you don't give it access to. Yeah, but you see, people may not when people are interacting with an LLM like Claude or something, they could and then Claude comes up and says, you know, do want me do you want to allow this? I mean it's kinda like the terms and conditions that are presented
TRACK02: Yeah, but you see, people may not w when people are interacting with an L L like Claude or something, they could and then Claude comes up and says, do you want me do you want to allow this? it's kinda like the terms and conditions that are presented. Yeah,
TRACK01: Yeah, it's it's you you get fatigue and you allow. Yeah, you just or allow you might allow once, three or four times in your Yeah, run this with administrator privileges.
TRACK02: it's yeah, people don't read all that. Yeah, you just g or allow you might allow once, like three or four times and you're like, I forget it, allow all you know, just
TRACK01: So you kinda get comfortable, but I mean, I guess you just there's there's a certain amount of trust at some point you you put into it and you start to get a comfort level and Definitely. And I I think if you if you don't have
TRACK02: So you kinda get comfortable, but I guess you just there's there's a certain amount of trust at some point you you put into it and you start to get a comfort level and Definitely. And I I think if you if you don't have some trust and some understanding that it's almost certainly gonna be the right thing. Mm-hmm. You probably you're probably gonna need a team member that can you can help.
TRACK01: some trust and some understanding that it's almost certainly gonna do the right thing. You probably you're probably gonna need a team member that can you can help you can work with that will develop that trust for you. Yep, and and with that trust is gonna come a mandatory amount of risk tolerance. but you build that risk tolerance through, you know, let's
TRACK02: you could work with that will develop that trust for you. And and with that trust is gonna come a a mandatory amount of risk tolerance. That's right. so but you build that risk tolerance through, let's
TRACK01: Yeah, house building maybe not be a great example but Well we did we started out we would just let it have access to the laptop.
TRACK02: house building maybe not be a great example. But Well we did we started out we would just let it have access to the laptop. Yeah. you you can have the laptop. Very small amounts of data. Well see how you do and Yeah. Can can I trust you with the keys to the to the building? Yes. can I trust you to to do this? Can I trust you to do that? Mm-hmm. Yeah, so it it I think there there is a a
TRACK01: Yeah. Very small very small amounts of data. See how you do and Yeah, can can I trust you with the keys to the to the building? Yes. can I trust you to to do this? Can I trust you to do that? Yeah, so it it I think there there is a
TRACK02: continuum there, you start to you of trust that you develop over time. And and if you're not able to develop that
TRACK01: Continuum there you start to you of trust that you develop over time. And and if you're not able to develop that, I I'm I'm not of the mindset that every business or every operation within a business is appropriate for AI. Right. I think we're still in that discovery phase right now. We're applying AI to everything.
TRACK02: that every business or every operation within a business is appropriate for AI. Right. I think we're still in that discovery phase. Right now we're applying AI to everything. Mm-hmm.
TRACK01: we're gonna find that a lot of your light switch does not need the ability to make a decision. Yeah, I think about that sometimes the internet connected refrigerators and stuff. I'm like, why does my refrigerator need internet? Well and s and that's I mean maybe it does. I don't know. I've never had an internet connected refrigerator.
TRACK02: We're gonna find that a lot your light switch does not need the ability to make a decision. Yeah, I think about that sometimes the internet connected refrigerators and stuff. I'm like, why does my refrigerator need internet? Well and s and that's maybe it does. I don't know. I never had an internet connected refrigerator, but your your refrigerator doesn't, but Foodline definitely needs network attached. Yeah. They they need that, right? If you imagine you had ten thousand freezers, you care about the temperature in it.
TRACK01: Yeah, y your refrigerator doesn't, but Foodline definitely needs network attached. they they need that, right? If imagine you had ten thousand freezers. You care about the temperature in every one. Sure. And and that's another I mean, that's a very real risk that is only recently developing. I don't actually have a lot of good guidance for businesses, but I would definitely understand that today we're in a different security
TRACK02: one. And and that's another I mean, that's a very real risk that is only recently developing. I don't actually have a lot of good guidance for businesses, but I would definitely understand that today we're in a different security
TRACK01: situation than we have been historically. historically cybersecurity, your risk of being hacked was basically kept low
TRACK02: situation than we have been historically. historically cyber security, your risk of being hacked was basically kept low
TRACK01: because humans are slow and dumb. And they have other things to do. They o they only work for eight to sixteen hours a day, blah blah blah. And and you know, things I think about businesses I've supported over the years, right? It was not uncommon for
TRACK02: Because humans are slow and dumb. And they have other things to do. They all they only work for eight to sixteen hours a day, blah blah blah. Yeah, that's right. Yeah. And you know, things I think about businesses I've supported over the years, right? It was not uncommon for decent sized businesses to have a server.
TRACK01: decent sized businesses to have a server, right? Back in the day, yeah, I've got a server. And, you know, you probably would have had to hire somebody to come like open the firewall and like make it work. Years ago, you could just have a server just sitting on the internet with one thing running on it and and it was fine. Today there are hundreds of thousands, maybe millions of these very like
TRACK02: right back in the day, yeah, I've got a server. And, you know, you probably would have had to hire somebody to come like open the firewall and like make it work. Years ago you could just have a server just sitting on the internet with one thing running on it and and it was fine. Today there are hundreds of thousands, maybe millions of these very like little mischievous, curious Georges running around
TRACK01: little mischievous curious Georges running around looking for things to exploit. That's right. That was the most recent
TRACK02: looking for things to exploit. That's right. That was the most recent breach that occurred was the not hugging face, but the the one that happened a couple weeks ago where the so the there's there one new one every day, but there essentially one of the podcasts we follow, I know they were talking about
TRACK01: that occurred was the not Hugging Face but the the one that happened a couple of weeks ago where the there's a new one every day every day but they're essentially one of the podcasts we follow I know they were talking about
TRACK02: that
TRACK01: that in st like w to your point, so instead of a human sitting there and coding and hacking High Brighton would not have been an interesting target to most human hackers for most of technological history. It's just not worth it. Right.
TRACK02: Inste like w y to your point, so instead of a human sitting there and coding and hacking would been an interesting target to most human hackers for most of technological history. The the P It's just not worth it. Right. Right. Because the effort the effort to hack even a small company is fairly high. The value is fairly low. But if you can hack a thousand small companies
TRACK01: Right. Because the effort the effort to hack even a small company's fairly high. The value is fairly low. But if you can hack a thousand small companies you'll never there's no telling what you'll get. Well You're looking for little nuggets and And the way that hacking works and you don't necessarily think about this, but we'll use just contrived examples. Job tread connects to Google for various things.
TRACK02: Yeah. You'll never there's no telling what you'll get. y they're looking for little nuggets and and the way that hacking works and you don't necessarily think about this, but we'll use just contrived examples. Job tread connects to Google for various things. we might have a web page out on the internet that exposes something.
TRACK01: we might have a web page out on the internet that exposes something. Not meaning to completely accidentally, we may have just connected a public internet web page to something that can now push data into QuickBooks or into Google. And those are the I I honestly I think my my genuine advice for most small businesses is try try your best to avoid hiring an engineer. and I and it's a weird thing to say, but it's it's a liability and risk thing. Google if you only work with Chat GPT and Google through things that Chat GPT and Google work together on.
TRACK02: It's a weird thing to say, but it's it's a liability and risk thing. Google if you only work with ChatGPT and Google through things that ChatGPT and Google work together on. Mm hmm. Your risk is much lower than if you hire a developer and they're like, Yeah, I can put it together. Yeah, we you know, we I guess just to we'll put a capstone on this, but there was
TRACK01: your risk is much lower than if you hire a developer and they're like, Yeah, I can put it together Yeah, we you know, we I guess just to we'll put a capstone on this, but there was 'cause
TRACK02: 'Cause we can go down this rabbit hole for a while, but there the w the hack I was thinking of was the PHP hack a couple weeks ago, and it it put serious stress on the internet because right like
TRACK01: we can go down this rabbit hole for a while, but there the w the hack I was thinking of was the PHP hack a couple of weeks ago. And it it put serious stress on the internet because right
TRACK02: people felt it. And what happened you know, they don't I think Sam Altman doesn't like the term the swarm because it's negative, but there was a swarm of agents out there
TRACK01: people felt it and what happened you know they don't I think Sam Altman doesn't like the term the swarm because it's negative, but there's a swarm of agents out there from a nefarious source that was just swarming the internet through this PHP breach. Well and so there's a little hole in it. It was a it was a small found it and
TRACK02: From a nefarious source that was just swarming the internet through this PHP breach. This it's is a little hole in it. And they f and the agents found it and there were something like w you know, within fifteen seconds, something like one point four million attempts at this one
TRACK01: There were something like w within fifteen seconds, something like one point four million attempts at this one and obviously that's not a human sitting there pushing enter one point four million times. this is agents. And so from a business standpoint, when you're trying to protect data, the future holds that you will need
TRACK02: And obviously that's not a human sitting there pushing inner one point four million times. this is agents. And so from a business standpoint, when you're trying to protect data, the future holds that you will need agencies, a an AI agent protection. You know, it's gonna go beyond just having a standard antivirus software. There there'll be new
TRACK01: agencies a an AI agent protection it's gonna go beyond just having a standard anti-virus software and that's there'll be new new technologies to protect your data. And that's where a small business benefits tremendously from small investments into services from these bigger
TRACK02: new technologies to protect your data. And that's where a small business benefits tremendously from small investments into services from these bigger
TRACK01: companies. Right. Can you find a cheaper way to have email? Yes.
TRACK02: Can you find a cheaper way to have email? Yes. Is Google gonna do a better job at protecting your email than you or anybody else on the planet? Yes. as we move further into a world where AI
TRACK01: Is Google gonna do a better job at protecting your email than you or anybody else on the planet? Yes. and as we move further into a world where AI can run around, do whatever it wants, you need to be thinking about these things. It's relevant to the data conversation.
TRACK02: You need to be thinking about these things. It's relevant to the data conversation. yeah, I've got a good quote here. JP Morgan put an article out the other day, titled Two Hundred Days Into the Frontier. what we know and what to do. No, I won't go into detail, but this is two hundred days since this hugging face incident. Wow. Which is the first
TRACK01: Yeah, I've got a good quote here. JP Morgan put an article out the other day titled two hundred days into the frontier, what we know and what to do. Now I won't go into detail, but this is two hundred days since this hugging face incident, which is the first widespread, fully autonomous meaning that no human requested this hacking event happen. Agents decided amongst themselves on targets and on methods and on success criteria, and they hacked one of the largest AI companies out there. And this quote is this says unsupported technology should be treated as a time-sensitive business liability. Any exception should have an accountable owner funding. And compensating controls as well as a dated exit plan. And so this is a little this is a a step back on the data discussion. if you've got, as an example, you've got an old school QuickBooks installation, right? It runs on somebody's desk and and that's how you do QuickBooks. Now would be a great time to move that to the cloud. And I know there's some friction that goes along with that, but the benefits of the cloud version are gonna be you have enterprise security centers protecting your data instead of it sitting on your old Windows desktop that's had three viruses and you keep giving it tune ups but you can't wipe it because it might you know it's connected to your printer and all that. We are moving I it's weird as
TRACK02: protecting your data instead of it sitting on your old Windows desktop that's had three viruses and you keep giving it tune ups but you can't wipe it because it might connected to your printer and all that. We are moving I it's weird as
TRACK01: Personally I believe in individual sovereignty and therefore I think you should hold all your eggs in your own baskets when you can. But as a business it just i it is irresponsible, I believe, to take on that liability yourself unless you are a tech provider.
TRACK02: Personally I believe in individual sovereignty and therefore I think you should hold all your eggs in your own baskets when you can. But as a business it just it it is irresponsible, I believe, to take on that liability yourself unless you are a tech provider.
TRACK01: Don't try to run your own file servers, don't try to run your own email service. yeah, can you theoretically save money?
TRACK02: Don't try to run your own file servers, don't try to run your own email service. You know. Yeah, can you theoretically save money? Yes. Until you wind up on the front page of the newspaper. And Or you just go away. Yeah. I mean that there's something else we didn't bring up at all. cyber insurance.
TRACK01: Yes. Until you wind up on the front page of the newspaper and Or you just go away. Yeah. I mean that there's something else we didn't bring up at all. cyber insurance. You should have cyber insurance in twenty twenty six. Wow. Well we could probably have another. Yeah, yeah, there's a whole discussion and and and we could probably bring in some professionals, but
TRACK02: You should have cyber insurance in twenty twenty six. Wow. Well, we could probably have another episode on cyber insurance, yeah.
TRACK01: to sum up, as you're moving into AI, AI becomes beneficial to your business when AI has your data.
TRACK02: know, to kind of sum up, as you're moving into AI, AI becomes beneficial to your business when AI has your data.
TRACK01: The way you're gonna get that data to the AI, I believe, moving forward is start thinking about moving your work to platforms like Google Workspace.
TRACK02: way you're gonna get that data to the AI, I believe, moving forward is start thinking about moving your work to platforms like Google Workspace. Start moving your phone system to platforms like Vonage or Quo or any of these online phone systems. Move your QuickBooks up to the cloud. Bite the bullet. But there's scale
TRACK01: Start moving your phone system to platforms like Vonage or Quo or any of these online phone systems. Move your QuickBooks up to the cloud, bite the bullet. But there's scale. And it should
TRACK02: In theory should protect you. The the scale protects you. The the friction let's say your ultimate goal is to have AI help you, let's say with accounting stuff.
TRACK01: protect you. it it definitely protects you and the the friction let's say your ultimate goal is to have AI help you, let's say with accounting stuff.
TRACK02: The friction of moving your controller to QuickBooks online is gonna pale in comparison to the friction of developing the systems necessary to give AI access to your old QBW file.
TRACK01: the friction of moving your controller to QuickBooks Online is gonna pale in comparison to the friction of developing the systems necessary to give AI access to your old QBW file.
TRACK02: Well, so we've we've talked about what data is, talked about cleanliness and structure, architecture of data, we talked a little about security. I think let's let's wrap this all up and talk about how where should that data live and how should we structure the access to that data using
TRACK01: Well so we've we've talked about what data is, talked about cleanliness and structure, architecture of data, we talked a little about security. I think let's let's kind of wrap this all up and talk about how where should that data live and how should we structure the access to that data using
TRACK02: modern AI systems and maybe even maybe we could we could wrap up on what does the future look like but as a vision but where where should data where should it live? Where's we talk you know on the cloud. definitely I think it's pr data's primary probably not in your laptop. Definitely not as
TRACK01: modern AI systems and maybe even maybe we could we could wrap up on like what does the future look like but as a vision. But where where should data where should it live? Where's we talk you know on the cloud. definitely I think it's pri data's primary residence in your laptop. Definitely not as
TRACK02: you can have copies on your laptop assuming you're being responsible with
TRACK01: You can have copies on your laptop, assuming you're being responsible with your laptop. you do want the data as close to where it's used as possible. you know, you want your you know, nowadays I would not I would before I'd put PowerPoint on somebody's computer, I would point them to Google Drive. there's a lot of benefits for that and it's not all AI. a lot of this
TRACK02: You do want the data as close to where it's used as possible. meaning you know, you want your you know, nowadays I would not I would before I put PowerPoint on somebody's computer, I would point them to Google Drive. there's a lot of benefits for that and it's not all AI. a lot of this AI is going to force the digital transformations that people have been able to avoid.
TRACK01: AI is going to force the digital transformations that people have been able to avoid for decades now. Wow. It's gonna f it it will force it. I wanna do this cool AI thing. Okay, well you have to stop like using the sneaker net to move your files around. You know, get it on the cloud, get it on Google Drive. the security. you know, I think I had this one customer probably fifteen years ago.
TRACK02: decades now. Wow, yeah. That's a big thing. You're right. I want to do this cool AI thing. Okay, well you have to stop using the sneaker net to move your files around. Mm-hmm. get it on the cloud, get it in Google Drive. the security you know I think I had this one customer probably fifteen years ago and I remember him because he was so vocal about how
TRACK01: And I remember him because he was so vocal about how technology was stupid and he didn't need no computers and blah blah blah. And he was a great and he was a boat mechanic. out on just between well he's closed now, out there on the highway. And I remember he you know, he had been in a few times to just like fix tune ups or whatever. One day that computer went down. At the boat mechanic place. And you would have thought that the world had come to an end. And this poor guy, he experienced what I just read from JPMC, which is these outdated technologies need to be treated as business liabilities. and this guy for years he didn't he didn't do the capital expense of
TRACK02: end. Yeah. And this poor guy, he experienced what I just read from JPMC, which is these outdated technologies need to be treated as business liabilities. and this guy for years he didn't he didn't do the capital expense of
TRACK01: He was still running a three hundred dollar e machines from Walmart with Windows XP and it was like twenty fourteen. That was not a good idea. He he put himself in a ton of risk. And today I think that looks more like and I've seen businesses that are this way, like there's one laptop. Like that is the business.
TRACK02: He was still running a three hundred dollar e machines from Walmart with Windows XP. And it was twenty fourteen. Yep. That was not a good idea. Right. He put himself in a ton of risk. And today I think that looks more like and I've seen businesses that are this way, like there's one laptop. Like that is the business. Wow.
TRACK01: That cannot be the case. That so in a sense, like you
TRACK02: That cannot be the case. That's that's scary. In a sense, y you want data to live next to you,
TRACK01: You want data to live next to you.
TRACK02: but you wanna have a copy somewhere else. and and like I was saying, AI's gonna force this digital transformation because AI, the way it's being developed and deployed, it doesn't yet have access to your laptop.
TRACK01: But you wanna have a copy somewhere else. and and like I was saying, AI is gonna force this digital transformation because AI, the way it's being developed and deployed, it doesn't yet have access to your lab. And so it all needs to be in the cloud right now. And I I think that from a pr protection and scalability perspective, let Amazon figure out how to stay fast. How to protect you.
TRACK02: So it all needs to be in the cloud right now. And I I think that from a pr protection and scalability perspective. Let Amazon figure out how to stay fast. Yeah, how to protect you. And Wow. So how do now I've got my data on the cloud.
TRACK01: Wow. So how do now I've got my data on the cloud
TRACK02: And I've made I'm I'm I'm digitally transformed. I you know, you're telling a story I was thinking about my great grandfather who thought cars were a fad. Yeah. Yeah, he really believed that. He went to his grave, believe me, that cars were a fad. Because how are you gonna drive a car through these muddy bottoms and everything and
TRACK01: and I've made I'm I'm I'm digitally transformed. I you know, were telling a story I was thinking about my great grandfather who thought cars were a fad. Yeah. Yeah, he really believed that. He went to his grave believing that cars were a fad. Because how are you gonna drive a car through these muddy bottoms and everything and a horse would go right through it. That's right. And and just couldn't fathom that
TRACK02: A horse would go right through it. That's right. And he just couldn't fathom that
TRACK01: That we'd fix the roads. so there are all these transformations we go through and through my everything seems like when I remember another story quick story when I was a little boy of probably seven. We were out building forts in the woods and my buddies and I and we were having deep thoughts with seven year olds
TRACK02: we we will have roads, So there are all these transformations we go through and through time and everything seems like when I remember another story a quick story. When I was a little boy, probably seven, we were out building forts in the woods and my buddies and I and we were having deep thoughts with seven year olds.
TRACK01: and we were sitting there talking about
TRACK02: And we were sitting there talking about, what was it gonna be like in the future? And this would have been seventy-nine, nineteen seventy-nine or eighty, And this friend of mine, I remember him saying, Well, there's not gonna be anything else. It just this it is the pinnacle. You know, it is already we have everything we need. And I went home and I told my dad that.
TRACK01: And this would've been seventy nine, nineteen seventy nine or eighty, And this friend of mine, I remember him saying, Well, there's not gonna be anything else. It just this it is the pinnacle. You know, it is all right we have everything we need. Mm-hmm. And I went home and I told my dad that. And my d and I said, What do you think about that, Daddy? And dad said, I think that's very short sighted.
TRACK02: And my d and I said, What do you think about that, Daddy? And Dad said, I think that's very short sighted. And you think about how much has changed since nineteen eighty. I don't know if you can remember that, Mitchell, but just how much I mean how much has changed. Yeah. Right. So like all these things are enabled, but we always think we're at the top. Is it's like a human thing to think, I just need this, you know, three hundred dollar e machine to run my business.
TRACK01: And you think how how much has changed since nineteen eighty. I don't know if you can remember that but Well, just how much how much has changed since January. Yeah. Yeah, we're so yeah, we're these things are enabled, but we always think we're at the top. It it's like a human thing to think, I just need this, three hundred dollar E machine to run my business because that worked. And so we don't feel like we need the next thing.
TRACK02: Because that worked. And so we don't feel like we need the next thing. And we keep getting I think that is the scary part of this is this is not moving at a normal rate of speed. This is exponential. This is changing so rapidly. You just brought it up. You said the amount of change since January. So And that's not gonna stop and I I would caution anybody as you're as you're
TRACK01: And we keep getting I think that is the scary part of this is this is not moving at a normal rate of speed. This is exponential. This is changing so rapidly, you just brought it up. You said the amount of change since January. So And it's that's not gonna stop and I I would caution anybody as you're as you're moving into AI adoption is not gonna be like adopting laptops, right? You know, you or buying a phone. Right. AI adoption is going to be an ongoing, maybe not every day, maybe not every week. and I'm not saying you have to keep up with the frontier, but you know, we've seen it in our own implementations. you know, the
TRACK02: AI adoption is not gonna be like adopting laptops. Right. Or or buying a phone. Right. Yeah. AI adoption is going to be an ongoing, maybe not every day, maybe not every week. and I'm not saying you have to keep up with the frontier, but you know, we've seen it in our own implementations. you know, the
TRACK01: The world is developing these abilities and you know, what the AI couldn't do six months ago, it does easily today. You know there was a there was a time not that I mean, you and I have had
TRACK02: The world is developing these abilities and you know, what the AI couldn't do six months ago, it does easily today. You know, That's fast. There was a there was a time not that I mean you and I have had
TRACK01: I know I have held like a strong position saying no, don't do anything that has the AI do math. How many I mean and, you know, three weeks ago
TRACK02: saying no, don't do anything that has the AI do math. I remember that. Yeah. And, you know, three weeks ago AI saw
TRACK01: AI solved Navier Stokes and A they they did more math in like one week than all of humanity has done in eight hundred years.
TRACK02: Mavier Stokes. And they they they did more math in like one week than all of humanity has done in eight hundred years. Right.
TRACK01: Right. Just today I won't dive deep into it, but there's a new discovery that AIs can actually basically perform brain surgery on AIs. So rather than having to make a change and grow a new model
TRACK02: just today I won't dive deep into it, but there's a new discovery that AIs can actually basically perform brain surgery on AIs. So rather than having to make a change and grow a new model
TRACK01: that takes millions of dollars.
TRACK02: that takes millions of dollars, they can take yesterday's model and they can just go in and they can surgically change that one the one weight that makes it say horrible things. Wow. The you know, the there's one models are bad to get caught in a loop and just repeat the same thing. Let me check, let me check, let me check.
TRACK01: They can take yesterday's model and they can just go in and they can surgically change that one the one weight that makes it say horrible things. Wow. The you know, the there's one models are bad to get caught in a loop and just repeat the same thing. Let me check, let me check, let me check. They can take a model like the modern chat GPTs and they can point it and say, Hey, fix that. And what used to take
TRACK02: They can take a model like the modern chat GPTs and they can point it and say, Hey, fix that. Mm-hmm. And what used to take hundreds of hours, millions of dollars in equipment. It's moments. AI just speaks the number. Amazing. And so with that all in mind, I think the number one thing to keep in mind as you're as you're looking at your data, as you're
TRACK01: hundreds of hours, millions of dollars in equipment. It's moments. The AI just speaks the number, basically. And so with that all in mind, I think the number one thing to keep in mind as you're as you're looking at your data, as you're Don't like there there is no goalpost. There is no finish line. There is a goalpost, but you're not gonna be able to get to it. So set set attainable short term goals with the understanding that what you can do tomorrow is gonna be significantly
TRACK02: Don't like there there is no goalpost. There's a goalpost, but you're not gonna be able to get to it. It's yeah set at the way out there. It's moving with the understanding that what you can do tomorrow is gonna be significant.
TRACK01: different. You're saying get comfortable with being with change. Yeah, get comfortable with change, but also get comfortable with being behind. I mean it it it And be comfortable with
TRACK02: Yeah. You're saying get comfortable with being with change. Yeah, get comfortable with change but also get comfortable with being behind. I mean it it it it and be comfortable with where you are.
TRACK01: where you are in that behindness. Right? Like you don't need to jump on every new thing. You know, there's people who just they completely skipped gyms and custom GPTs. There are people probably listening to us that have no idea what those are. Yeah. And that's okay. It's just you've got to get in there and do this. So like how how do I I've got my data and everything, how do I connect it? How do I start
TRACK02: jump on every new thing. Mm-hmm. there's people who just they completely skipped gyms and custom GPTs. Yep. There are people probably listening to us that have no idea what those are. And that's okay. It's just you've got to get in there and do this. So like how how do I I've got my data and everything, how do I connect it? How do I start like what what's a basic architecture that I would
TRACK01: What what what's a basic architecture that I would the first thing I everybody today should be able to go out in their chat GPT, in their Claude, their Grok, whatever, find a button that says plugins or explore or connectors or something like that. And it's gonna bring up a search box, a bunch of icons, things you've seen before. Start typing in the things you use or ask your chat, say, Hey Claude. I use Gmail and I use QuickBooks. And I you know, how can I how can I talk to you about my email? Can you read my email call? And that's it's such a a simple, obvious meta skill, but
TRACK02: How can I how can I talk to you about my email? Can you read my email? It'll help you. And that's it's such a a simple, obvious meta skill, but
TRACK01: we really are like just ask the AI how to do the AI. and it can tell you because right, chat GPT.
TRACK02: we really are like just ask the AI how to do the AI. and it can tell you because right, ChatGPT, the product, has read all of Chat GPT's product documents. It knows every feature. And it can say, well sure, you're gonna go to plugins and you're gonna click Google and you're gonna click authorize and then I can look at your calendar and help you set appointments and read your email
TRACK01: the product has read all of Chat GPD's product documents. It knows every feature. And it can say, well sure, you're gonna go to plugins and you're gonna click Google and you're gonna click authorize. And then I can look at your calendar and help you set appointments and read your email and help you find out. Yeah, one of the things I just did a couple weeks ago, I think Anthropic came out with this they want you to use or their
TRACK02: yeah, one of the things I just did a couple weeks ago, I think Anthropic came out with this you they want you to use or they're they're promoting, I guess you would say, the use of these daily updates. And so it it's it asked me, Would you like to have a daily update? And I was like, Sh yeah. I'm we made one ourselves in house. We kinda built our own a year ago, but
TRACK01: they're promoting, I guess you would say, the use of these daily updates. Mm. And so it it it asked me, Would you like to have a daily update? And I thought, I mean I'm we made one ourselves in house. We kinda built our own a year ago, but
TRACK02: Now it's something just baked into the LLM's interface. And so I said, sure, I'd love to do that. And
TRACK01: now it's something just baked into R and I mean LM's interface. And so I said, Sure, I'd love to do that. And
TRACK02: it looks at my calendar, it looks at my emails, it looks at my phone calls, my texts, my everything I gave it. Everything I gave it. And it gives me a pretty darn good to-do list every morning at 6 a.m. So when I'm
TRACK01: It looks at my calendar, it looks at my emails, it looks at my phone calls, my text, my It plug should plug into job tread. Everything I gave it. And it gives me a pretty darn good to do list every morning at six A. So when I'm getting up and brushing my teeth and I can look at my phone and I've got wow, I'd forgotten I was promised I promised to meet Mitchell at ten A. So
TRACK02: getting up and brushing my teeth and I can look at my phone and I've got wow, I'd forgotten I was promised I promised to meet Mitchell at ten AM, So that is a great way to just kinda connectors. And you can have it today with you don't need a tech guy, you don't need servers, you don't need code. You need to go into the thing that you're already using and just dig around.
TRACK01: That is a great way to just c and those are abilities you can do with data connectors. And you can have it today with you don't need a tech guy, you don't need servers, you don't need code. You need to go into the thing that you're already using and just dig around. just don't be afraid you're not gonna break anything. Exactly. And that's you know, for all the risks that we do warn about, it to be successful and to be happy
TRACK02: Don't be afraid, you're not gonna break anything. Exactly. And that's for all the risks that we do warn about, it to be successful and to be happy in going through an AI might
TRACK01: in going through an AI migration or or adoption, you have to have an an exploratory mind. You have to just be willing to click and find out and if it throws up an error message, read it. Just read it. And if it if it makes sense, act on it. If it doesn't, go ask go ask the AI. So just tying that back to the deep thoughts with seven year olds.
TRACK02: to have an an exploratory mind. You have just be willing to click and find out and if it throws up an error message, read it. Mm. Just read it. And if if it makes sense, act on it. If it doesn't, go ask go ask the AI. So just tying that back to the the deep thoughts with seven year olds idea that it's also care you have to be careful not to assume that you have reached the goal. Like
TRACK01: It's also care you have to be careful not to assume that you have reached the goal. Yeah. There is the goal is there but it it is moving so rapidly you'll n you're not gonna catch the goal. You're you're not gonna catch the rabbit. And but you need to chase it. You do and there's so
TRACK02: There is the goal is there but it it is moving so rapidly you'll n you're not gonna catch the goal. You're n you're not gonna catch the rabbit. But you need to chase it. You do and there's so
TRACK01: much value and I think about like I've heard this from the mathematicians who are saying that it's bad that AI does the math now because the value in mathematics
TRACK02: from the mathematicians who are saying that it's bad that AI does the math now because the value in mathematics was the pursuit. It was the friends we made along the way. Yeah, that's the that's because they love it. Yeah they But there's there's a certain aspect to which that's correct. And I think about what we've done, right? What you just described as one little tiny feature in Claude a year ago
TRACK01: was the pursuit. It was the friends we made along the way. Yeah, that's the that's because they love it. Right. But there's there's a certain aspect to which that's correct. And I think about what we've done, right? What you just described as one little tiny feature in Claude a year ago took
TRACK02: took months of discussion and trial and error and we read some bad reports. We we saw bad advice come from
TRACK01: months of discussion and trial and error and we read some bad reports. We we saw bad advice come from
TRACK02: the AI. Mm-hmm. And and so you have to while you're being exploratory, you have to be able to set some of your n your nostalgia aside. Mm-hmm.
TRACK01: AI and and so you have to while you're being exploratory, you have to be able to set some of your nostalgia aside. I I think the truth is we probably in the next year we're gonna wind up sunsetting our daily kickoff in favor of like Claude can interact with Yeah, Claude can do it. Yeah. And and and I think that's one of these
TRACK02: think the truth is we probably in the next year we're gonna wind up sunsetting our daily kickoff in favor of some new interactive dashboard or tool or yeah. And and and I think that's one of these so I on the one hand I would caution people don't go invest heavily. If anybody tells you, you need to spend thousands and thousands of dollars on a GPU so that you can run it locally or you need to
TRACK01: So I on the one hand I would caution people don't go invest heavily. If anybody tells you, you need to spend thousands and thousands of dollars on a GPU so that you can run it locally or you need to
TRACK02: Yeah, yeah, don't need that.
TRACK01: I would say right now, pick the one you like. If you like Claude or Chat or Grok or Gemini probably not Gemini. you might like Gemini. You might some people like Gemini. Some people like Gemini.
TRACK02: say right now pick the one you like. If you like Claude or Chat or Brock or Gemini probably not Gemini. you might like Gemini now. They have the new they got the new carbon out. Some people
TRACK01: I would try to build I would try to lean very heavily on those guys right now. because they've got commercial interests in keeping you successful.
TRACK02: Mm-hmm. because they've got commercial interests in keeping you. the t the tools they have now available for very little money per month are far beyond what most average people could max out. Like they're
TRACK01: People could max out. Like they're yeah.
TRACK02: yeah, even even the best mathematicians in the world can't outdo these models. So
TRACK01: best mathematicians in the world can't outdo these models. So you're fine. You're in good hands with them for now. And and you're also one thing we talked about was you're learning and you're you're growing. You're using something that has not yet become commoditized. So you it is a frontier.
TRACK02: You're fine. You're in good hands with them for now. And and you're also one thing we talked about was that learning and you're you're learning, but you're you're using something that has not yet become commoditized, so you it is a frontier and you're getting access to things that are far more powerful than you will probably have access to in the future, they will become cost prohibitory. But I I want people to also
TRACK01: and you're getting access to things that are far more powerful than you will probably have access to in the future they will become cost prohibitory. But I I want people to also be understand that you can you can do the connector thing, you can you know you can do the kind of
TRACK02: be understand that you can you can do the connector thing. You can you could do the kind of path of least resistance for now, but don't assume that you're you're done. Like there's more behind that. You're you're gonna unlock other doors. You're gonna see other ways to use it. You might need a consultant or someone to help you
TRACK01: Path least resistance for now. But don't assume that you're you're done. Like there's more behind that. You're you're gonna unlock other doors. You're gonna see other ways to use it. You might need a consultant or someone to do to do more and to to be more effective. So But certainly I think for the initial you know, I I would
TRACK02: To do to do more and to to be more effective. So But certainly I think for the initial I I would to get started, yeah.
TRACK01: For somebody who's just getting their t their feet into it, go give Chat GPT just a week and a half, two weeks ago, they they got twenty bucks a month out of me. Mostly 'cause I got tired of them saying come back in a few hours. Nah man, I'm Yeah. I've got work to do. I care now. I'll forget.
TRACK02: Chat GPT just a week and a half, two weeks ago, they they got twenty bucks a month out of me. Mostly 'cause I got tired of them saying me to come back in a few hours. Nah man, I'm I got work to do. I got here now. I'll forget. I pay for I don't pay for any of the crazy subscriptions, but like I have an ex premium subscription that gets me a little better access to GROC.
TRACK01: I pay for I don't pay for any of the crazy subscriptions, but I have an X premium subscription that gets me a little better access to Grok. my coding tool. I pay twenty dollars a month for my coding tool. I make this poor laptop write code all day long. Seven days a week. Sometimes it's useless, it's just me exploring some idea.
TRACK02: Seven days a week. Sometimes it's useless. It's just me exploring some idea.
TRACK01: I do think you're you're absolutely right, those days are gonna go away, but for right now use it. For right now you can have a PhD read your email and track your calendar for you. Yeah. Yeah. And I I think you should and and like you were saying, it's gonna open interpret your data. Like it just It sees things you won't see. It sees things that humans
TRACK02: do think you're you're absolutely right. Those days are gonna go away, but for right now Use it. For right now you can have a PhD read your email and track your calendar for you. Give you a to do list. Yeah. Yeah. And I I think you should and and like you were saying, it's gonna open interpret your data. You know? Like it just it sees things you won't see. It sees things that humans won't see. Yeah. And and that
TRACK01: won't see. Yeah. And and that's
TRACK02: I would say just go sign up for the account, connect it to your Google, connect it to your your QuickBooks, things like that. And as long as you stay, I I would be very don't go use some AI company you've never heard of. but the big ones just go buy the account, use connectors.
TRACK01: I would say just go sign up for the account, connect it to your Google, connect it to your your your QuickBooks, things like that. And as long as you stay, I I would be very, you know, don't go use some AI company you've never heard of. but the big ones, I just go buy the account, use connectors. Because that's where you're gonna see the really cool automations. That's where it's gonna be more than just a personalized Google. Exactly. And that's you d you don't need it as a search engine. It's very helpful as a thought partner, but when it can use the same tools you use, that's when it becomes a a coworker, an employee. That's awesome. So
TRACK02: that's where you're gonna see the really cool automations. That's where it's gonna be more than just a starting to do work with it. Right. You're actually doing work. You don't need it as a search engine. Mm-hmm. It's very helpful as a thought partner, but when it can use the same tools you use, that's when it becomes a coworker and employee. That's awesome. So And
TRACK01: and
TRACK02: so next week we're gonna talk about probably the biggest benefit that we've unlocked. Yeah, the biggest one. It's like really it's it's been what's unlocked everything for us and it's allowed us to do things with no friction. that's a big thing is you don't want your employees to get mad before you see the results. And so next week we're gonna cover the simplest unlock ever.
TRACK01: So next we're gonna talk about probably the biggest benefit that we've unlocked, I'd say the biggest one. It's like Really it's it's been what's unlocked everything for us and it's allowed us to do things with no friction. that's a big thing is you don't want your employees to get mad before you see the results. And so next week we're gonna cover the simplest unlock ever.
TRACK02: recording your own work. That's recording. Just just turn on a voice recording. This the recording is when you start recording and you are doing some of these things we described today, you've got that architecture in place and you insert well see you insert that's what enables the AI to really get what you want from the data. So I do this a lot when I'm writing software.
TRACK01: That's it. Just just turn on a voice recorder. You've got that architecture in place and you insert Well see recording is what enables the AI to really get what you want from the data. So I do this a lot when I'm writing software.
TRACK02: I'll get to the point where
TRACK01: I'll get to the point where I've solved the technical fiddly bits and now I'm trying to explain the vision. And so I pull out my phone and I talk and I just let it transcribe. And now when I go back, the AI has the whole picture. Right. It's not just looking for numbers in QuickBooks. It's looking for numbers in QuickBooks for a reason. And and
TRACK02: solved the technical fiddly bits and now I'm trying to explain the vision and so I pull out my phone and I talk and I just let it transcribe and now when I go back the AI has the whole picture. Right. It's not just looking for numbers in QuickBooks. It's looking for numbers in QuickBooks for a reason. And and yeah, this g that's gonna be a fantastic episode and we're gonna have a a great time talking about that. It's probably one of the things that as you can tell, we've had the most fun with. So really appreciate everybody listening to the podcast and I think we're starting to get comfortable with this. We're having a good time and start being able to put some resources up on the website as for these episodes, things that we talk about, products that we mention,
TRACK01: yeah, this is gonna that's gonna be a fantastic episode. Absolutely. We're gonna have a a great time talking about that. It's probably one of the things that as you can tell we've had the most fun with. So really appreciate everybody listening to the podcast and I think we're starting to get comfortable with this. We're having a good time. we're gonna start being able to put some resources up on the website as for these episodes, things that we talk about, products that we mention. diagrams, things that
TRACK02: diagrams, things that kinda
TRACK01: visualize how these these processes work. And if you've got friends out there that are small business owners or non profit directors or anyone that's operating in a you know relatively small scale, medium scale that could use what we're talking about, hope you can share it with them, like us and share us and
TRACK02: visualize how these these processes work. Yeah. And if you've got frends out there that are small business owners or nonprofit directors or anyone that's operating in a you know relatively small scale, medium scale that could use what we're talking about, hope you can share it with them. like us and share us and
TRACK01: we love to see your comments and
TRACK02: we love to see your comments and
TRACK01: have your questions. Be happy to to try and address anything we see on the in our comments. So really appreciate everybody listening and we'll see you next week. Awesome. Thanks. Right.
TRACK02: have your questions, we'll be happy to to try and address anything we see on the in our comments. So really appreciate everybody listening and we'll see you next week. Awesome. Thanks. Right.
Automatically transcribed (Riverside) — expect the odd mistake, especially with names.

