I don't. I'm not chaotic for the sake of being chaotic, and I'm not in

the. And I'm not chaotic for the sake of breaking something or causing

harm to others. Right. We're chaotic because our brains work differently. And this

is now the age of the chaotic brain. We can now take these cool,

crazy ideas, get angry at it for. For a half an hour like

you said, and then have Claude code come in here for 40

minutes and then fix it, and then you're on to the next thing. If that

sounds like your brain on a good day, then you're in the right

place. Welcome to Data Driven.

Hello and welcome back to Data Driven, the podcast where we explore the emerging industry

and field that is AI data and of course, all the

stuff that underpins it. Normally, my most favorite

data engineer in the world would be with me, but however, today he is

not. However, I did do have a different Andy.

Welcome to the show, Andy. Is it Butcher or Betcher? It is

Betcher. Betcher. Okay, well, welcome to the show.

You are a Chief innovation officer at.

double track. So tell me a little bit. What is

a. What does this Chief Innovation Officer do?

Well, I will tell you. My favorite way to describe a Chief

Innovation Officer is the one person you call

when you're stuck and you need to move your top

or bottom line and don't quite know how. Got it right.

So I will come into an organization and ask some

of the crazy questions, propose some of the crazy

approaches, throw a bunch at the wall, see

what sticks, and then the old

overused trope of go fast, fail fast.

We will try a bunch of things as quickly as we can with all of

the new awesome technologies that we've never had before in our

careers as you, and probably twice as

better as this Andy, your normal Andy that you have on your podcasts.

As you guys have talked about, we have not had this type of

technology available to us ever in our careers. And, Frank, I've been doing this for

32 years, right? I have done a lot of fun

stuff in my day, but not like I have in the last couple years.

I mean, heck, Frank, much less than last six months. So taking

all of this wonderful crazy brain that we all have and applying it

with some really cool technology, that's how I see a Chief Innovation

Officer is. Is just. That's the best way I can

describe it, man. I think that's a good way to put it because you see

a lot of companies, they really struggle with. They know they want to use AI,

particularly in the software space. Right. They will know they want to use AI,

they know they want to do all these things, but they don't really know how.

Right. And it's a different mindset, I think, you know, in the

virtual green room, you said, you know, you, you, you, you don't

want to be responsible for running things. Right. You want to be responsible for trying

things, is basically what you said. Yep. And I think that's a. Put it.

Because we live in such a. No one, no one

knows how this is going to play out. Like, honestly, like, everyone thinks they know

how it'll play out or, or whatever. But, I mean, I was able to,

you know, put together,

literally I had, I had a car accident in December. Literally that day I

started a new project on GitHub and just started Claude code, just

chewing away at something, and Now I'm like 80,000 lines of code later,

and it's only been maybe three months. Yeah. Right.

And, you know, 80,000 lines of code is not a trivial amount

of work. Right. And it's an

idea that I would not have. Yeah. I mean, theoretically

I could have, you know, raised the money, found the money, paid

people to do it, but I wasn't, I wasn't going to do it.

Like, realistically, I wasn't going to do it. But now, now I am the precipice

of having this product, you know, that's out

there, that helps podcasters. I built it for the

needs that Andy, Candace and I have for the different shows that we have. And,

and, you know, basically product, you know,

production line for podcasts is basically what it is.

And recently had to change the domain name because somebody

else had something very similar. So, but, but,

but again, that was not a lot of work comparatively. I just told,

you know, Claude, like, hey, look, this is too similar. This is the new branding.

And it went, it did it in about, you know, I think I was salty

about it for like an hour. Yep. And then,

you know, Claude had it fixed in 40 minutes. So, you know, I

was, I was mad longer than it took to implement the fix, which

if that's not, if that is not a metaphor for our time,

because, you know, maybe a part of me was stuck in the old ways, like,

oh, my God, I have to change the domain. Oh, my God, I have to

change the code base. I have to do this. The AI doesn't really care that

much. Right. To it. It's just finding a replace. Yeah.

It's crazy. So I will

say, what was it? Probably

2006, 2007

was my last kind of career pivot, right? And

I was moving from a primarily Microsoft driven developer

and then vb, C net, SQL Data, like all the

things as I was pivoting that into what would

turn into a 13 year Salesforce career when I stopped, when I stepped in

Salesforce and I was looking, comparing, contrasting, walk in saying, well, heck, out

of the 100% of time. Or you know, Frank, as you and I know as

nerds, we've got about 120, 130% of time.

You know, families might disagree with that, but that's what we do.

If you look at the 100% of time, you used to spend 100% of the

time standing things up and dealing with domain name security and stuff. And when I

looked at Salesforce, it was like, well, heck, I can take 80% of that too

it off the plate because the platform handles it and the rest 20%. I can

use my crazy brain to actually like do something cool. And that went for

a bunch of years. And now we are in the world of AI and the

same thing's taking place again right now. It's with all of the other

technologies plus all the new ones, right? To your point

of making that application to help you and your cohorts there with, with the

podcasting, you can now walk in with the idea,

excuse me, post production, and edit that one out there or leave it in for

comedic effect, whatever, right? So you can look at this stuff

and like really stand some things up. But my wife pointed out,

you get you Frank. You've seen on the Internet the stupid little like, you know,

nine box D and D role character matrices,

right? And my wife sticks me in the center column all the way

to the right, which is the chaotic neutral,

right? I, I, I, I don't, I'm not chaotic for the sake of being

chaotic. And I'm not in the in, I'm not the chaotic for the sake

of breaking something or causing harm to others, right? We're chaotic because our

brains work differently. And this is now the age of the chaotic brain.

We can now take these cool, crazy ideas, get angry at it for it for

a half an hour like you said, and then have Claude code

come in here for 40 minutes and then fix it and then you're on to

the next thing, which both addresses our wonderful chaotic nature

and our crazy ADHD brains which are jumping around like a pair of, you know,

like a whole bunch of popcorn kernels and a popcorn popper. So, you know,

I similarly, but with my clients and also with all of Our

side projects that we all have, right? You know, to try to apply this

technology and bake it into our brains. You know, same type

of thing, 80, 90, 100,000 lines of code. You're typing stuff in. You're putting

what would have been six to eight months of a team

doing test frameworks, which is

one thing, by the way, Frank, if you have not tied a test framework

into your work with your applications or anybody listening to this

podcast, the. The power and the resiliency

that you're getting through Claude code and, you know, pick your tool. I'm just

picking on Claude because, I mean, my opinion, about four months ago, they. They went

into a different ballpark. They're not even in the same ballpark as everybody else right

now. So you take all the crazy, awesome stuff you've got going with

cloud code right now and how it helps you through things and all this, tell

it to go pick on what you're not thinking of or put

a test framework in, or have it do security auditing.

Have it. Go find, you know, find. And I can give you a list here,

Frank, of the different security organizations that put out really good

white papers on the. On the methodologies they go through that all apps should

do. And you point Claude at it, you say, hey, either it's Claude code by

itself or you pick up co worker dispatch. Now, that will tie into the browsers

and browses you to go through and look through stuff and say, hey, I need

you to compile all. All of this and then apply that good logic in here.

So it's not just ideation, it's also a

level of fortification. Now, I will also say

this for all of the old grizzled gray hairs

that are listening to the podcast, because, Frank, you and I both, I

think we've been around the block a couple times, and I know one of the

biggest things I get talking around this stuff is, well, you know, I've been

doing this stuff for a while. There's a lot of things that aren't spoken. There's

a lot of experience we bring into play. Well, absolutely there is.

We have to teach our tools to help us go through that. That's why we

use our crazy brains and do it. Like I mentioned that testing framework, right?

So, man, it is an exciting time. What was

the craziest thing, Frank, when you were dinking through your app, like,

what was the one, like, bang, aha. Moment that came up that you could

not have done otherwise? Oh, God, there's so many. But the

first one was really I think testing framework, no one really

enjoys testing. No necessary evil,

but no it's necessary evil. For me it was

Planning mode, right? Because for me, because it would be,

you know, it would go through and be like have you thought about this? Like

I think it just said that and we can go into and it basically suggests

we go into planning mode and discuss it. And I, I find myself having my,

this discussion with an AI that you know is

approximates a pretty reasonable conversation one would have

with a junior to to mid level

architect, right? You talking through these problems. I love

Planning mode, right. I what I'll do in my projects because

now with my, my ada I like to say I have Schrodinger's

adhd, right. Because it's both, you know, I don't have it

diagnosed, right. So it's adhd ish. So I can have

it when I need it and I don't have it, I don't need it

is I have like four different project ideas going on. Maybe, maybe

five. Right. And it's kind of like I also have

my co host on Impact Quantum is also very, very

neurodiverse and she leans into that and it really is kind of the

superpower if you don't, you know, especially in the age we live in now, right.

Because you can have these ideas and

you know, as long as you can prioritize them. And I find, you know I

basically created out of a Claude project a project

manager, right. So have a VP of project engineering. Each one of the

project ideas have kind of their PM that manages that project

and I kind of talk to them and it sounds weird but I mean

I converse with them one way or the other and they come up with ideas.

And you know, part of it was ideating on the name change, right.

Originally I called it Podsy because it was going to call it

Podzi McPaderson, right. But I had to change it

so it kind of like had the whole list of things and you know,

it helped me ideate the ideas and I gave it the name of the other.

So ultimately I landed on Show Dog. Okay,

but, but it basically kind of helped me walk through

it, walk through the branding chain. So you can tell it to act like a

marketing manager, act like this and it will kind of, for lack of better term

switch hats and it'll do that. And for me that was amazing, right? And

I can kind of have them all meet together and then I basically

one thing I discovered is you just output your conversation, the ideas that you have

into report and markdown Right. Which you can read at your

leisure and other bots can read.

Yep. So you go through and you kind of have this, like, very productive,

you know, session of like an hour or two, and they basically,

you. You plan out with the different bots and different Personas. You,

you, you write everything down. It's like the olden days, right, where, you know, people

would come up with a Project Action Memo and things like that. And,

you know, everybody acronym TPS report type of thing. Right.

Except useful. And, you know, you can basically put

Claude on dangerous mode. Right. And

go out, Go out to lunch, do something else. Do something

like, you know, hang out with the family. Right. And a few hours later you

come back to it and it's done, right? It's checked in, it's done. I

also always have it kind of do like a. Like a change log and like,

write a daily report, like, what'd you do today? And I can look back and

like, when I feel like I'm not making progress, I can look. Well, you know,

10,000 lines got written today, right? I mean, this is just you.

Basically everyone has effectively, like

a 10,000, you know, team of

developers, right? Because, I mean, how long would it take to write 10,000 lines

of code? Right? It would take, you know, if you needed to do them a

day, you would need to have at least a thousand developers.

And that's being generous, but then being able to, on a dime,

say, that's not what I meant. Pivot.

Yes. With no attitude or very little attitude. With very

little attitude. Right, Right. Yeah. It's that. That perhaps

was one of the. One of the biggest things, like when. When the

Anthrotic. The Anthropic app came out, and then all of a sudden they turned on

Claude code. And then I was able to hook my IDE into the

cloud versions of it, so I can be sitting, you know, in a restaurant

and, you know, and on my way to the bathroom and back. Not

in the bathroom because it's weird, right? But on the way to the bathroom and

back. You can just type something into your phone now and. And

just get the idea out of your head, because I know, Frank, I know about

you, but a lot of my thoughts don't happen when I'm just sitting here. Right.

When I'm sitting here, I've got. I. I don't know if you can see off

camera over here, but I got like 10 monitors in front of me and three

computers. Right. When I'm here, I'm locked into productivity mode and

I'm doing things And I'm being distracted by IMs and all this

other stuff when I'm out walking around the block or I'm driving my kids

somewhere or driving my grandkids somewhere or my wife and I are out.

That's where the ideas happen, when your brain is free. Right. I do a lot

of off roading. If you read anything, read anything about my bio. Anybody who's

listening, right? Like there's, oh yeah, he knows data and AI stuff. Oh

yeah, and don't get him started on jeeps because he'll monopolize the conversation with off

roading and jeeps all day and forget about data and AI, you know, save for

the fun space of when all of those intersections, right. When I can

finally figure out how to make money making that intersection happen, then

come talk to me too. Right? But that's the time. All that fun

time is when the cr. The cool thoughts are coming out and

what do you have in front of you? Oh, I need to remember to do

this. Or you take a voice memo and forgot you made the voice memo.

Right. So I can just go quick, go pop open, you know, a claw

dispatch a cloud dispatch and say, hey, go do these Google searches, put this crazy

hair braid idea back together. Notify me here when you're done at the end of

the day and then ask me if you have any questions along the way. And

then I put my phone back down and boom. You know, Bob's your

uncle, goes for three hours and does stuff and ask me some questions.

Like it's, it's like a digital assistant, an

offshore team that, you know, I, you know, again,

I, I pay the high end for, for Claude because I heard burning out

limits, right. And it costs more to buy more than just to buy the top.

Right? Right. Incredibly good marketing strategy and product strategy

from Anthropic. Right? Right. Like

when have we ever had this ability, especially as

neurospicy individuals, being able to go dive in and just have all

this crazy insanity. 80% of it drops off. But the

20 that sticks, Frank, is pretty damn cool.

I mean it is amazing. It is. I think it's a time. This really is

a great time to be alive. Everyone's talking about, worried about,

you know, we're recording this day after Easter, right. And we had

a friend over for Easter who is very much an old school developer. And

I'm like, you know, like you don't understand understand. He's, you know, not really dived

into AI and I'm like, you know, you really need to take a

Look at this, right? And he's in between jobs right now. And I'm like, you

really need to look into this, right? Because there's no, you know, there's no

avoiding it now, right? Everybody and their cousin

is. Is doing something with AI. Everybody. And their dog

is an expert in AI now. Right?

But, you know, one, I was talking to someone

who was a former Microsoft evangelist and teammate and

coworker and friend. Still a friend, not a former friend, but, you

know, she's like, AI has basically made every developer a manager now,

right? Because you have. Everybody has a team

if they so choose to treat it like a team. Yep. Right. It's a threat

and an opportunity. It's a threat for those who don't really seize on it,

but it's an opportunity for those that do. Right? Like, I mean,

you know, you mentioned Salesforce. Salesforce is probably the poster child

for SaaS success, right?

They have a. A building in San Francisco. You know, they have a tower,

right? You go back to, like, New York, right? There was the Woolworth Tower, right.

They changed retail, right? And there was the Empire State Building, the Chrysler Building. All

these things. You know, to think that a SAS company could have a tower

named after them, right. Is phenomenal. And, you know,

is every idea I have gonna be like, you know, is there going to be

like a, you know, a show dog tower in

Baltimore? Well, probably not, but, you know, maybe.

Maybe, right? Like, you know,

I mean, but, you know, the fact that I have kind of these five active

projects in my mind right now and they wouldn't. They would

stay in the back of their cocktail napkin in the past. Now they went

on the whiteboard and I kind of basically type into the. Into Claude

what was on the whiteboard or, you know, whatever the voice notes have.

I don't think people realize, like, just. Just how much power is at your fingertips

or your voice at this point, right. I drive my kids around

all the time, right? I'm taxi dad. And

I'll, you know, I'll see like a new research paper dropped. And,

you know, I don't. I don't have time to read them all, right? But I

do have time to drop them in a notebook lm and then have a

podcast made out of it that I can listen to. So that way, the first

time I sit through and read, it's not my first

time exposed to the material, right? So, like, I noticed I get a lot more

out of actually reading the. Not the physical, but like the electronic, you know, reading

it. I get a lot More out of it because I've kind of been prepped

and I heard lecture notes and things like that, that, that the AI created.

It's just, it's a phenomenal time to learn, for sure. And, and I

think now, particularly in the last six months, it's a phenomenal time to build.

Yes. Although I will say I've got two

phrases I'm going to pop here. Right? So the first one

is more serious. The second one is a little silly, right? So the

first one here is as I talk to individuals, you mentioned your friend

who's between jobs right now, right? And you did the exact right

thing. Hey, you know what? You got time. Go look at this.

Make your own decisions, use your own critical thinking. Like if this is something

you could use or something you can reference or something speak to, you

know, all that kind of thing, right? So the first one is I talk to

individuals and I'm like, you know, AI is not coming for anybody's job at

this point. Not yet. Right? That is a. Even,

even the craziest super thinkers, right, are like, you know,

six months, 12 months. No, there's no AI coming for your job. Right?

However, people that use AI are coming for the jobs that

people that don't, Right? And that same thing goes for

companies. I've got myself a nice little sweet

battery of customers. I've got a bigger network that I talk to. Like.

Like, that's the one thing. And anybo. Anybody in my network who's listening to

this is probably going to snicker a little bit and say, oh, Andy says that

to me a bunch. It's like, you know, companies. I mean, the AI is not

coming for companies, but companies who are leveraging AI are going to come

for the companies that don't. And that's actually, you know, Frank, to an

earlier point you made, and I'll get to my silly thing in a second here,

but a point you made earlier was that, you know, all companies are like,

we need to use AI. You know, that was like three years ago when I,

when I went out on my own and, and really took

all of my data backbone and my data history and brought it in and then

said, I know these tools are coming. I need to get brain

centered around to be able to talk to people about this. You know,

company after company after company are coming in and saying, we need to use AI.

The board's pushing us to use AI. My CEO is pushing, we need to use

this stuff. I have no idea what to do. I bought a bunch of stuff

and it didn't actually work and it's sitting on the shelf. And Now I've spent

$15,000 a month on AI stuff sitting on the shelf

and now they get a bad taste for it in their mouths. Like, no, you're

doing the right thing, but you need to think

differently. These are tools for ideation. There's no

easy button, right? You don't. I just put a post on LinkedIn

a little bit ago. But you don't treat an AI agent. An AI agent you

should treat like an employee, right? You need to onboard it. You need to feed

it information. You need to work with it. You need to understand that it's not

perfect. You need to be able to put it in situations where it can fail

and support. Support it and bring it back and work together, you know, so

that's what I talk with companies a lot about. I've got a whole methodology

and framework which, Frank, I'm not going to bother you guys with today, right? That's

a whole different podcast for a different day. But,

but really that's how I talk to individuals and companies today. You have to play

with this stuff. And then. So the silly one, Frank here is, is the Spider

man reference, right? It's the. With great power comes great responsibility,

right? You mentioned it. There's no better time right now to be able to learn

something, be able to, to, to have Notebook

LLM, go tear apart a large, a large document and share it to

you. This is one piece I am

very, very, very, very fervent about here.

And this is. I want to. I'm going to preface this with everybody

listening, that this is not a political statement, this is not a religious statement, this

is not an anything statement. But, but critical thinking is

now more important than ever because these tools are

so fast that they will summarize information

in the way that you. That if you worked with it enough, it'll summarize in

the way that, the way that you think. But you still have to use your

critical thinking skills to be able to pull this stuff apart. You still need to

use your critical thinking skills to question it and then go back and look for

more information. But I'll tell you, even that, you know, slightly

funny, slightly weird statement I just made there, Frank, still, I

have never been more excited in my career other than, you know, when

I was 20 and making cool things, right? Like, you know, staying up for four

days in a row making cool things. Now I stay up for four days in

a row with a lot of coffee and make cool things, but that's right. There's

a really so good. You're right, Frank. It's a cool time to be alive. These

technologies have never been here. I really thank you for inviting me on and

talking about this stuff. My, my, my, my

biggest, nastiest appliance is my Jeep Wrangler. That's where all of my

cash and time goes. So I actually have used AI to actually

help try to troubleshoot that sucker too. It's amazing, really. Yeah,

Tell me about that. Because I've had mixed results with like, AI troubleshooting. Right. So

like, we had a, you know, we're on well and septic, right. So we had

like a well, tank leak and, and

AI kind of got that completely wrong. But the other

day, actually for the, for the clothes dryer, like, I, I

basically said, I need this type of replacement screw. What does it use?

And it was like, well, tell me the model and make a model. I told

him, make a model. And it was like, well, try this. And I bought a,

I was at Home Depot and I bought a package of those screws and it

was magically it worked. Right. So it does seem to be kind of a,

I think, I think goes back to what you said, right. Like think about critical

thinking. Right. Had I followed its advice with the

plumbing, I probably would have done a disastrous work.

But with the, you know, for, you know, like $5 for a package of

screws. Right. It was the, the risk reward was, was pre, like, you know, if

it's the wrong one, I go back and I return it. Yeah. Right. These days

I'll spend more in gasoline, probably to drive back to the store.

Unfortunately, yes, right now that is the case. Yes. Then,

yeah, Then, then it would actually just be just, well, I have a box screws

I'm not going to use. But I do, I do think though,

like, so, like, how do you find it? Help you with,

with, with your Jeep Wringer? Because I am, I am also, I'm kind of,

I wouldn't say disgruntled. I would say a, I, I,

I'm a car guy, right. I like, I like big Chevys and I cannot

lie. And you know, I,

about 10 years ago I had this, I guess you could call it a midlife

crisis, I suppose. But I bought a 76 El Dorado

and convertible and that thing. Car was beautiful. But I

really overestimated my abilities and underestimated

the cost of owning and was kind of like, well, I had my

fun with a classic car, so. But it's, and it was funny, right? Like

when you have the money, you don't have the time. When you have the time,

you don't really have the money. Right. So I found myself,

you know, when I got laid off. When I, When I got laid off, I

found myself with time and I was like, you know,

oh, God, but this is expensive. Yeah.

My. My exploits into a. So there's kind of two ways I've done in car,

right? Yeah. There is the it's broken, I'm

3,000 miles from home and I have to figure this out moment. Right? Right.

So I've got a really good battery of friends who. We

all have different levels of mechanic. None of us are

mechanics. Right? Right. We. We are all weekend

warrior mechanics at best. And we've only

been. We've only honed our skills over years of being stranded in

places or looking at spending $4,000

in labor on something or giving it a whirl first.

Right. You know, I look at how I spin over to

AI with it is, you know, so the Chrysler,

Chrysler can bus is behind the Jeep Wrangler, right. And it's got a pretty standard

thing of codes and there's a lot of readers out there that read them. And,

you know, it's. It's like you look at regular like corporate AI stuff. You

gotta ground it, right. You gotta ground it. You gotta ground

the silly thing before you do anything, whether it's cars and you go

give all of the codes that are possibly there. And you pointed at a bunch

of good troubleshooting websites and forums and boards you

use again, coworker dispatch to go have it dig through a couple things and

say, hey, you know, I found some stuff over here. Go figure out what the

patterns are. You know, when you get down to things, there's always a bit of

artist or artistic interpretation and

good old physics that defeats an AI every day.

Right. But at least gives you an idea to be able to go

make that and I'll spin it over to your dryer comment to go make you

buy three different bags of screws for a low cost of

trial and error. And you find one works and you put the other two

bags in the drawer that you'll probably never touch again except for 10 years

down the road when you need that one screw. Right? Right. So, you know, there

are places it works. There's places it's not if I super

pivot for a minute. Right. If I go take a. Take that same thing

we talked about with how we ground and look at cars and use good critical

thinking and spin it over into like how we talk with corporations and how

to use AI and data stuff.

When I left my previous organization,

I did not have a job, right? I know I needed to

do something in here, and I was waiting. I'm a big guy that

believes in fate. I'm a big guy that believes in, and in. You know,

when you put good juju out in the universe, like, like

good things come back. So that's why, that's how I usually try to live things.

And I took three months off and I'm like, all right, how the heck do

I pivot this stuff around? You know, even three years ago,

we're talking 20, 23. We all

knew that AI wasn't brand new. It's been, it's been an

academia for decades, right? We all know about

grounding, we all know about all the things. Now it just became commercial.

So it's not brand spanking. I was thinking, well, how do I ground

my conversations with organizations

to be able to take what we just talked about with the car and apply

it toward how they want to move or move or change their top or bottom

line? Or I'm talking to a friend of mine and they're trying to troubleshoot a

lawnmower, right? You know, you gotta

ground the stuff. So I came up with, you know, every organization has eight types

of data. So we feed, you know, I've got a whole head, I got like

a massive prompt that feeds that in. And then every, every one of those domains,

every field to be relevant, reliable, revealing and reusable. Like every

single data piece has to be that. And we feed that with a prompt over

there. Then we feed some schemas and talk about stuff. But you know, again, with

great power, with great power comes great responsibility. You have to be

responsible and be critical in what you do. And then you can fix your lawnmower,

fix your Jeep Wrangler, or fix the fact that your support

department can't figure out X, Y and Z and

communicate it to the customers. Right? Like, that's the whole chaotic brain

side of things. You know, chaotic doesn't mean

messy. Chaotic means different or unpredictable. Right?

Or unpredictable. Right. I. One of

my superpowers is I can see commonalities across many

different, distinctly different areas that

are not normally tieable. Right? So

I don't know. That's my, kind of my personal walk in with this stuff. The

tools are fantastic. But if you've got a,

if you've got a crazy brain, Frank, man, you can do

some cool stuff right now. Oh, exactly. All the things that were

Would have been a headwind before now. Tailwinds,

right? Like, in terms of what,

you know, just. It would be impractical

to. To build out something like this. Like, you know, whether

it's. Whether it's show dog, whether it's another tool. I have a command

line tool I wrote like three, four years ago called Dingo that helps me

write blogs, and it was a command line tool, right? So

I'm not really, really the friendliest UI in the world, right?

But I basically pointed Claude and I was like, here's the code base.

I want this to be a web interface. And that's going to be another, you

know, another SaaS. Maybe it'll be the Dingo Tower in downtown Baltimore, who knows?

But the,

The. The solving problems

is what people pay for, right? And if I look at it this way, as

a podcaster, as a blogger, as a content creator, as a marketer,

as a technologist, right? I encounter a problem,

I'll be like, you know, I should probably have AI see what it can do

about this, right? Like, and that's how each one of these things was built, right?

As, you know, as we were kind of building out something and trying to

organize all the content from the different blog posts and

podcast episodes and video clips that we make from them,

we really was like, there really should be a tool to do this. And in

the olden days would be like, yeah, let me get something out on the whiteboard

or notebook. And then now I'm like, wait a minute, I can just tell it

to Claude. And Claude will build it, right? And that's. I mean, it's fascinating.

But you're right, you do have to exercise some common sense, right?

And I don't think. I don't think everyone's kind of figured this out yet. I

think very few people, to your point, like, the people that are going to be

the next Mark Benioffs are going to be the ones who figure this out first.

And I don't think people. I think people are still stuck in that mode of,

oh, my God, AI is going to take my job. Funny story, last.

Last springtime, actually, there

was a baby bird fell out of a nest, and my. I live in kind

of a rural area now, and I'm like, I'm a city boy, right? Like, I

have no idea what this. What to do. So I basically,

you know, basically asked Chat GPT, like, what do I do? Like, what type of

bird is this? Like, you know, and it told me and it

gave me a list of resources. I called around and you know, the,

the baby bird was saved because, you know,

Chat GPT recognized what species was. And I said they're like, how did you know

to contact us? And I was like, oh, Chat GPT told me. Yeah,

I was like. They were like, what? You know, it was a first for them.

Which is pretty funny. Yeah, I mean, I mean, Frank, look,

I don't know at this point, what, 15, 20 years ago we had the

same, I mean, to be overly simplistic about this, right. 15, 20

years ago it was Google, right? How'd you know to call us?

Well, I Googled it. Googled, yeah, yeah. Now, I mean

the, the tongue in cheek thing is my Google fu is strong. Wrong. Right.

I can find things faster, better, quicker on the first page than

other people because we don't, we, we, you and I know how to structure queries.

It's going to be the same thing with LLMs and, and

I'll even just say, you know, AI is a very broad term. Right, Frank, you

and all of our listeners know that too, right? AI is a very broad term.

We're talking specifically around like the, the LLM generative

AI side of ChatGPT and Claude and all that kind of stuff. Right?

Like there's going to be a point where prompt

generation and prompt engineering become second nature to the human.

Just like Googling has become second nature for years.

For the last 15 or 20 years. We're going to get to that point and

then there's going to be the next generation or evolution beyond that.

You and I are able to build cool stuff with

Claude code right now because we can at least

half rear end prompt generate on the

fly and can learn and see how it's reacting and go twist it

and turn it and restart and give it queued messages to go to

refine and tune it and respond to it. Right. We're

learning in real time there. I mean, heck, there's college courses out about

prompt generation. You know, I saw that two and a half years ago. There were,

there were college courses out there and it's a good base to get.

But that's, that's, it's, that's, you know, a long way around the bush

talking around Frank about, you know, prop generations and Googling. Right?

So when you're talking about the baby bird thing, it doesn't surprise me.

It makes me smile. It's like, yeah, it's because Frank knows how to ask a

question. I know how to pick a picture and send

it to the AI. Well, it's true,

like you mentioned 2023. Right. 2023. OpenAI

looked untouchable. Oh, yeah. And look at the

tool that we've spoken most about. Drops. Right, right. Chat TPT

drops commercially in December of 22. And everyone stares

at it like the baby bird that has fallen out of the tree sitting in

your yard. And you're like, what do

I do with this? Right. And

not to get really queer, weird meta on your, on your reference there,

but. But 23 was a wild year. 24 was

wilder. 25 started to be able to give us like,

actual stories about people making businesses out of

this. Hell, I've made a business out of it.

26 has no signs of slowing down. Right.

Well, and, you know, I also think too, like, the dynamics of it.

Right. You know, OpenAI looked untouchable for most of 2023.

And here we are. The tool we've spoken most about has been

Anthropics, you know, flagship Claude. Right.

Now, the other ones kind of also are roughly peers. But I

mean, in terms of adoption and

preference, like, I think it's. People don't realize, like,

this is still no 1.

No 1 has really dominated the space just yet. Right. This is a lot, like,

reminds me, you're old enough to remember the browser wars, right? Like, you know,

Internet Explorer was the underdog. This is before Internet Explorer became the

punchline to a joke. Right? So just

time can change a lot of things. It's. I don't know,

like, it'll be interesting to see how this pans out. I'm also excited about

agentic stuff too. I don't. I have an open claw instance, but

I kind of see the value, but I don't see what the rah rah hype

is about. Oh, it's, it's, it's just starting,

Garten. Right again. Yeah. The concept has been in academia,

right. It's the, the minds are spinning it

to commercial purposes and there's a lot of trial and error.

I think I saw there was a. I, I could pull it up. I got

it in one of my favorites here, there's a Gartner article about like, 80% of

all agentic projects are doomed to fail and sit on the shelf because they didn't

have the right outcomes in mind when they did it or the wrong technology

or, or, or, or, or. Right. If I pick back to something

you had just said ago a bit ago, you're, you're right. We're

talking about Claude because you and I have,

and people like us have really tied to Claude.

And you know, this Isn't the, you know, after the Department of Defense,

you know, the whole thing and the social before that. It's not that. Right.

It's like you and I are using it because it speaks to us and we

can speak to it and we're getting some cool stuff done. Right. This could be,

you know, Gemini might come back out in a massive leap forward in six months.

Right. Or pick your next company that we don't know about out right now

as a, as a, as a strategist and implementer in a

space that is so rapidly changing and I always miss this

up. Is it, is it Boyle's law or Moore's law? I

don't remember the name of the law, but it's the one that measured the progressive

technology over years by saying the number of transistors would double on a chip every

two years. You know, I, this is not an official statement. This is something

I just say. But you know, basically it's running on a monthly cycle right

now. What was taking a two year cycle is now taking a one month

if, if even that long cycle to jump. And one

of the biggest challenges that I have been,

I have, I have always got tripped up a little bit on and then

always think about as a strategist, implementer is I

cannot bind anything that I'm doing to one specific

company, technology or model. Right.

You have to be able to abstract those concepts far enough to be able to

say, I need an LLM here that does X, Y and Z

and build it in a modular way where, you know, it might be GPT

5.2 right now. And then I'm going to swap in

Sonnet for this and I'm going to swap in Llama for that. And

I'm going to do this and continue to grow. Being able to effectively

articulate that as a business owner, Frank, to do stuff for yourself

or, or as a customer, being able to say, hey, you know,

you're gonna, I'm gonna charge you a bunch of money to go do this thing

and we're gonna go play with this stuff and we're gonna build something really cool

and it's gonna bring ROI and it's gonna move our. It's gonna, there's our

defined outcomes, here's how we're gonna get there. Oh, and it's on technology that's

gonna literally shift every month. How do we plan for that?

Like that, to me is one of the biggest challenges right now

to be able to snap

based on the technology we have. We're going to have about a 60 month. 60.

Not 60 month, no 60 day build cycle before this

thing comes out all the way into your production environment. But

on day 61, while it runs off and

meets the outcomes that we had, we're going to be talking about version 2 already.

Not because something is completely outdated, not

working. It's because there's that much new that has come out in 60 days that

could make our outcomes even better. Better. Right. It's an odd

balance and it's an, Honestly, it's a, it's a mind shift

with companies that have been dealt with Salesforce for years or

dealt with Microsoft Dynamics for years that they'll, they, they talk about,

you know, quarterly or annual value improvement

cycles. Frank, the stuff we're talking about is monthly. It's

insane. Right? I mean that's, I can easily see it going even more

granular than monthly. Right? Yeah. I don't know when, but like it, you know,

it. I, you know, sales software cycles used to be what,

three years, then 18 months, then 12 months, then quarterly.

I mean there's no, really no, I'm sure

there's a practical upper limit, but I, I think we still got

a lot more headroom. Oh, Frank. Frankly, the, the, the

ultimate limiter is going to be the human ability to adapt to change.

Yeah. Yeah. That's what it is. Right. I mean, you're right. I mean I

just put out a response to a massive RFP that I've

got more change management people than I have

technical people. I've never done that before, Frank. But

that's where, right, it's, it's. I need to be able to

change hearts and minds and the way people think and the way that people see

outcomes rapidly enough in a corporate

setting to be able to accommodate the change, not just put it

in and let you know, internal change management or governance. Right. It. Or business

leaders take it like we actually need a fleet of change managers.

That sounds absurd, but the more I think about it, the more I think that's

smart. Right. At first blush you're like, well, that sounds kind of. Oh

yeah, I see why I couldn't even finish the thought. The

sentence in my head was like, oh yeah, that makes a lot of sense.

It's a crazy time, man. Yeah, it's like the old

playbook, while not completely like useless, is definitely, definitely need some

updates. You know, history is always

the best teacher. Right, right. You know, now you just have

to compress your history time scales. You don't

compress your history lessons. You know,

we both, I guarantee you, we've Both been through

36 month SAP implementations at different organizations. Right,

right. And even the change management. But once you intrude,

entrench people in that, you know, I mean you.

They're not changing. Lord, they're not changing. Right.

But now you're taking things like I just built

an account, an accounting, I'm not going to say I built accounting system because

that's dumb, but built an accounting helper to be able to do

things and trying to introduce that into a regulated or

any type of industry that has hardcore rules that they

have to answer to, like you have to change hearts and minds. The

tech is easy. The tech's the easy part. How change management is the

hardest part. Yeah, it's a good way to put it because like, you know,

one of the things I think that has been exposed through the use of,

you know, AI generated code is that generating

code isn't the only thing that software developers do. They solve

problems, they have to talk through problems. Right. I spend probably

most of my time interacting with any of these AI tools

trying to solve a problem. Right. Or when you do solve a problem

that, you know the engineering discipline of, when you solve a

problem, you have certain trade offs which trade off matters to you.

Right. As kind of like some of them are hard, some of those are not

hard, difficult, but like kind of just physical limitations versus some

of those are mental limitations. Right. In terms of how you design a ui. Well,

you know, if you call it this, then you. That's at the

expense of that. And then it's just a lot of trade offs in

conversations, which is what I find myself looking at. Most

of the interactions I've had with these tools have been

mostly talking through the trade offs. Yeah,

we used to, we did the same thing for years, Frank, just at a slower

pace. Yeah, yeah, yeah. I mean this would be like a meeting you would have

every week or two now. Now you're having it in the half

an hour that you're working with Claude and, and a,

a business representative to be able to ground you on outcomes.

Oh, shoot. Okay, now I got to think of this and this and this and

now we've got identity and access we have to worry about. We got to worry

about this UI piece and this ux and we have this ux. We have to

worry about this and that and this and that. You're literally making

changes as fast as that into your, into your plan mode and

Claude to put out something that somebody can

poke with a stick. Right. And then they can poke it

enough with A stick and it works just fine. Then like talking about your friend

the, the, I'll call it the again,

older and haggard software architect who's built things

very well as a living. Like eventually

Claude is never going to replace those individuals

at scale, right? All we're doing is getting the

next alpha or the next beta or the next

stable piece out there. And then if somebody needs to harden that, then you need

to put real process around. How do I properly secure this?

Because I mean Frank, the

spoken and whispered in corners is

data, right? We're making these applications

and you're a smart guy, I'm a pretty smart guy.

Together we're smarter together, right? We don't

automatically cover everything. Claude leaves a hole here and there,

right. I was talking with this

absolutely brilliant woman who was trying to pull an application to market

and I eventually I talked with her for a good like six sixty days about

ideating through this thing. And it got to a point where we were collecting so

much private information about an individual for a good

reason, right. But it automatically puts a big old target on your

back. And I said, you know, we're gonna have to alpha through this

thing, get a workable, a workable prototype, get a couple people to

trust us enough to put stuff in here and have it secured

and disconnected in a way where I'm not gonna risk their stuff. But

then I'm going to need to put a security architect on this where to put

these layers. We're going to spend a quarter million dollars to get sock to a

compliance to make sure that we hold this thing right. You know, I've

never gone from idea to a Sock2 conversation

in 60 days and I ended up walking away from that

because of the data. Chance was, was so big,

but even, I mean, and that, that was, that was like gathering

will information and bills and things like that. Like, like really

like potentially harmful information if somebody got that in the wrong hands. But

you think about your company CRM, right? We talked about

Salesforce a minute ago. We can go put up, we can go stand up an

agentix CRM in a week

to be able to have somebody log into it through a, you know, through a,

you know, login through Google, Microsoft, sso, all that kind of stuff.

But, and I know I'll stop, I'll get off my soapbox here in a moment

because I know we're coming to the close of our time here, Frank, but the

one of the, you know, we talked about change managers. A big thing I want

our listeners to walk away with here is to think about that piece. The second

is data security. You know, just because you can doesn't mean

you should. Right? There

is. There is such a market in people's data

and the tools that we have are so awesome, but they. We

in our, in our, in our absolute happy path. Exuberance. To be able to

solve a problem, you always should have assigned

tag to your wall that says, what about the data? Because the last

thing you want to do is to have this awesome thing that solves a

perfect problem and then somebody gets wiggled

through an unsecured API and downloads your data.

I don't care if it's your company's CRM, which is frankly, a CRM is pretty.

I mean, you're assembling it for public information, for God's sakes. Right,

Right. You know, much less anything that has financial data data, or much less anything

that has personal identifiable data or health data or health data.

Dear God. Right? Like there is a whole bunch of stuff that we are barreling

toward. And like, like I've got right at my eyeline

up here, I have five pieces of paper that have immutable

concerns that I always have because I can glance up over my

monitors and I can see them and I can see. Well, how do I manage

my Outlook inbox? Right. I'm an inbox zero guide, Frank.

Like, here's my inbox zero stuff. Then I've got what about the data? And then

I've got a couple other things up here as well. Topics for a future thing.

Andy's weird things on the wall. But, you know, it's

just one more piece I wanted to talk about. Here is. Is, you know, again,

with great power comes great responsibility. Just stop and think about what you're making.

Right. Right. Yeah. All right, Frank. I'm off my box, bud.

No, no, I think it's important. It's. It's important. Sometimes soapboxes,

you have to have those soapbox moments, particularly around data security, data

quality, data provenance. I guess that's what the cool kids

are calling it now. There needs to be. Sorry. I have

docs and puppies who are wrestling in the background.

It's that kind of day. Kids are off from school and all that.

Appliances breaking down. Dogs and cats living together. Mass hysteria. Mass

hysteria.

Where can folks find out? I'd love to have you back on the show

because I think that we barely scratched the surface. Because an

interesting perspective on, you know, not just how, because

billions of YouTube videos, you know, tell you how to do it, but

like the why, the what and the constraints and the issues that you'll come up

with. I think this has been a very enlightening conversation.

Love to have you back on the show. And we'll have two Andy's on the

show and you know, that'll be, that'll be interesting experience.

And where can folks find out about, more about you and kind of what you're

up to these days? I mean, frankly, the two biggest places is

of course, my LinkedIn profile. Like I'll spout out a random thing here

or there. You know, even a blind squirrel finds another every now and then on

LinkedIn. Right. Frank, what I'm really trying to do is my website,

doubletrack.com I'm working with my marketing people, so I

try not to do anything in a vacuum because they get angry at that. But

I'll work with my marketing people to put out good,

impactful tools, utilities,

questionnaires, things like that. Like you said, Frank,

there's, there's millions of people right now that have YouTube videos out there.

Everyone's telling you how and how cool it is and how fast, but

no one's going to come to my website for that. They're going to come for

the, the alternate point or the part that says, oh, nobody actually asked

me that question before. That's what I'm trying to put out on my website.

DoubleTrack.com is a bunch of those type of things. So,

you know, if you want to go and just kind of challenge your own brain

or, you know, even if you're talking to Claude, say, what am I missing

missing? Ever done that before, Frank? Have you ever asked it like, what am I

missing? What am I not thinking of? What are the holes? What is the anti

design pattern that I need to do? That's kind of my

mode when I come into this is I'm always challenging myself and

others about what are you not thinking about? What should

you be thinking about? That's. So to answer your question, Frank looked at my

LinkedIn, weird things come out there. Look at my website, more structured things

come out there. I would absolutely love to come back on, on, on your show

and talk about crazy stuff. I, you know, I

promise I won't get too crazy or too weird. I read

a lot of weird books. So. Hey, man, when, when the going gets tough,

when the going gets weird, the weird turn pro. Yes, they do.

All right. And with that, we'll end the show. Okay.