#19: AI for Coders & Microsoft Power Apps Copilot with Christopher Robbins

Welcome back to another episode of the Junction.

We have another first here on the show.

We have an in studio guest.

I didn't know our studio was big enough

to have more than two people in it. Oh, no, it works.

Don't underestimate the studio.

We've done some big things in here.

We're gonna get, like four or five people in here next.

I mean, Scott's trying to bust

the wall out, like another 5ft. Oh, nice.

Just wait for season two. Right.

Get it in the budget. Love it.

I don't think we can automate that part. Yeah.

All right, Chris.

Chris Robbins from Lumis Technology.

Thank you so much for being on the show.

I know you go way back with Chase.

Tell us a little bit how we got connected up.

Give us your background.

Yeah, so my wife has been working with the venn

team for a long time, and we were in the

need of a salesforce implementation at my company, Lumis Technology.

And so we decided that we were going to bid it

out and Ven came in and they were great partner with

us, and we're in our second implementation with them now.

So it's been a great ride and

we're going to keep going there.

That's awesome.

I remember when I first met your wife and

she would tell us what you were doing.

And I know Bink will be listening to this, but

I'm like, wow, this guy is doing incredible stuff.

We really should talk to him.

And it took a while for us to get in contact

with one another, but now I'm so glad that we did,

because all the things that you're doing in your day to

day job are things that we're talking about on the show.

In real life, we just kind of conceptualize a

lot of the things and we're like, wouldn't it

be really cool if we went and did this?

Are there people out there

actually writing code using AI? Yeah.

Well, Chris, Chris is one of those guys that we're like,

this is who you should emulate, right, if you want to

do some of this AI stuff in your day to day

job or you want to implement in your business.

So maybe, Chris, for the audience, give us an

elevator pitch as to what it is around automation,

AI Rupin, some of the technologies that you're using.

What does that look like?

What does your day to day look like

from, I don't know, a 30,000 foot view? Yeah.

So I'm in charge of the modernization effort

at Lumis, where we're taking 270 legacy applications

and we're modernizing them into cloud platforms and

making it better for our user experience.

Internally, there's a lot of

internal work when doing that.

I'm having to translate legacy code, make upgrades,

and all those different type of work efforts.

Well, we don't have.

I'm a chemical engineer by background.

I have an undergrad in chemical engineering, a master's

in mechanical engineering, and I got into development because

I wanted to do my job better.

So I wanted to find a way to automate myself

almost out of my current role and into something else.

And that's essentially what I've been able to do.

I've been able to learn c sharp in a weekend.

I wrote an API in c sharp with having

no background in c sharp that's now integrated into

the Microsoft Power platform and working with all of

our power apps that we're working with right now.

So AI is just powered everything.

GitHub copilot is my best friend.

I call it my intern because I just ask it

with a comment, hey, how do I write this?

And it pumps out the code 80% correct.

Run a little checking, ask it why is this the error?

Just keep working with it.

That prompting of how I work with it just

makes my life better, makes everything move faster.

Is it kind of like that intern

concept that we've talked about before?

It's 100% the intern concept.

If I give it a really bad comment, yeah,

it's going to give me some pretty bad code.

But if I give it a good comment and I explain what

I'm trying to do, it'll get me over that line much faster.

There's talk of the ten X developer and AI

really is driving towards that ten X developer where

a single developer can be that much more efficient.

They don't have to try and

remember because most developers don't.

They don't remember the nuance of the code

on how to do every little thing.

So this just takes that away where you

don't have to overly worry about the syntax.

You want to worry about how do I write the

most efficient and the most getting to the point of

the code first versus trying to remember syntax.

Yeah, no one likes syntax. Yeah. How are you?

I'm curious.

We've had a guest on the show that said he

asks it to justify, like he says now, justify why?

That's the answer.

So you pointed out why is there an error

here or you're kind of continually prompting it?

Are you asking it anything specific like that?

And why is this right?

I don't typically ask it that. I will have it.

Give me an answer and then ask it, hey,

can this be done in a more efficient manner?

Is there a way we can do this a little bit better

or that it'll take less compute time and things like that?

So I work with it on trying to

find the better way to do things.

We've talked about this a couple of times.

I feel like in our Zoom calls, and you

said this right before we hopped on the show,

that everybody's worried that it's going to take your

job, whatever your job might be.

But you mentioned that it'll never take your job, or

maybe not never, but you don't necessarily believe that.

So tell me what your thoughts are on that front.

It's going to take away the parts of

my job I don't like to begin with.

The parts, the monotonous parts, the parts where I'm

having to look something up and find the source.

I'm going to be able to just ask, hey,

what's the budget running on this current job?

And it'll be able to go grab me that information, bring

it back, and then it'll give me a link to, hey,

here's the source material where I got this from.

So if I don't trust it, I go look.

If it doesn't pass that smell test, I go look, and

then I can ask it better questions if it's not there.

So it's going to help everybody just

be more efficient at what they're doing.

Everybody has those mundane tasks.

I mean, you guys aren't to automate here.

So with that automation idea, it's

just another form of automation.

Yeah, I see a lot of that in talk of the agent piece.

Right?

Like the chat GBT can be your agent.

Well, the agent is going to specialize

on one thing, coding or something else.

And I think we said this in last episode, you're

going to need a human to dictate and direct and

move things around, even in your internship, right?

Somebody's going to have to be there to be the expert.

Because if you don't know how the code should be

written or the direction that you're needing to go, you'll

never get there, because the agent will just start spewing

stuff and you won't know if that's right or wrong.

There has to be a human in the loop.

I mean, it's all engineering

programs and everything like that.

There has to be a human in the loop

that's there to give it that sanity check.

It can guide you down that path, but it's never

going to give you the perfect answer the first time.

It might get lucky, but you're going to want to

make sure there's a human in the loop that understands

the problem that you're trying to solve and can help

coach the agent get down that path with you.

It's an 80% thing.

It's a copilot.

It's not an autopilot.

There's a webinar that we accept that

we're going to watch later today.

It's Ava, I forget the name of it.

The idea behind this platform, brand new,

they're pre money, haven't really done anything

but that they can write.

They've written, supposedly a sales bot agent that

can do all of the sales things.

You give it an ICP, you give it your product

or your list of services, and it will supposedly go

email them, find a list, write emails, email everybody, and

then the chat functionality that we've all seen, right.

You're like communicating with the agent,

hey, how'd you do this week?

And the agent's like, I sent twelve

emails and I got one response.

But they have a thought that they can automate

that entire portion of the sales cycle, right?

Do you see that, like individual, maybe smaller portions

being fully automated, or do you see like.

Well, of course.

Like you said, chase, you'll have to

chat with the chat bot, right? Yeah.

So where I see that, yeah, it might be able to

send those cold call emails that no one likes to send.

Again, it can help you draft those emails, it can help

you research for those emails, but you still may want to

put your personal touch on it, or you're not going to

have to work with it to make it sound like you. Right.

The big disconnect that I would see is when

you finally get on the phone with me, you're

not going to have any of that relationship.

I'm going to have that relationship with the bot, and you're

either going to have to work with the bot to understand

who I am, to talk to me, or the bot is

going to have to be on the call too.

And I mean, that's coming with the ability to just plug

the bot right into being able to talk to you.

But that's still not necessarily the relationship I would think

is going to be built in the sales world.

Especially when you're talking about huge contracts and trust

that the company is going to do what they

need for you and things like that.

Well, I can even just envision,

like, you have 100 avas.

Sure, you can keep up with one, but if

they're even mildly, even just a minuscule successful.

And now you have 100 people that are interested

in your product, but you can't read 100 set

of emails to catch up with everybody.

Like you're just going to lose continuity.

Well, and how are you going to deliver on all that?

Totally. So great.

You've gotten 100 new leads.

That may even be really hot leads, but

you don't have the team to support that. Great.

Your sales cycle now got improved by that much, but

do you have the AI to do the rest of

everything else that you need to deliver on?

It's in progress. Right.

We're going to write it as we sell it.

I think without the AI.

I've always sort of had this love

hate with the cold email anyway.

Unless you've really kind of like.

I think there's probably to an extent they can go

scrape my LinkedIn profile and Ava can be like, hey,

Mel saw you, or a know, prime time.

And then all of a sudden they're sort of

building a relationship with me over a couple emails.

To Chris's point, I have always wondered, now

how do you have that same context?

How are we going to pick up from where Ava teed up?

Right.

I'm going to get on the phone expecting,

I'm going to probably look at your LinkedIn,

see that you went to a m.

I'm going to raz you a little bit about that,

and then you're going to maybe go with the flow.

But where's my primetime shout? So I don't know.

I think that there is the.

Are they providing you some sort of

like one pager snapshot summary of.

I reached out to Mel three times.

Here's what her pain points are based

on the ICP that you gave me.

And how is it giving you the level

of context as the salesperson to seemingly pick

up right where you left off?

And even without the automation, I think that's

where a lot of cold outreach falls flat

because there's just not the continuity that the

end or the person receiving it expects.

Yeah, it's like a bait and switch at a different level.

Right?

We built a relationship, but I'm going to have you

talk with Mel and it's like now I'm not.

Just get Ava on the line. Where's Ava? Where's Ava?

I want to talk to Ava.

I don't want to talk to you, Chase. I'm sorry.

Hey, man, switching gears,

you mentioned power automate.

When I picked it up, when it

first came out, it felt rudimentary.

A lot like Zapier, maybe a little bit different,

but based on what I've looked at and what

you've been telling me, it's a lot different now.

So I'm curious, what are you doing in the platform or

what are they doing that makes it so much better?

Know three or four years ago, yeah.

So Microsoft has been investing heavily in all of the

embedding of AI into these platforms too, because they're really

trying to push down that whole low code barrier, get

it into the hand of the user versus into the

hand of the developer like me, which has benefits on

both sides of the court, more for me to have

to maintain on the back end of when somebody makes

an automation and then we have to make it go

enterprise, but it also makes it easier for everybody.

So they have an AI agent now built into power automate

where you can talk to and say hey, when I get

an email from my boss, do XYZ actions for me.

So there's all sorts of just AI agents that help

you build these power automates now built right into it.

On top of that, we're using the power apps as

the front end for all of our modern applications.

So the front end is based in Powerapps.

We're doing power automate for some of

our workflow load that we might need,

but we're doing engineering applications.

We're designing chemical plants with these apps.

So we're also using Azure services such as Azure

functions and Azure API manager to help us drive

down and get all that detailed calculation work that

we need integrated into the database.

So the amount of work going on in power

platform, it's hard to stay on top of.

So I'm on calls weekly with Microsoft hearing about

the latest and greatest things that they're working on

and it's amazing to see just the evolution and

the speed that things are coming to.

I remember when it first came out and

I was like, this is Zapier basically, right?

But I knew in the back of my head that you

put a logo like Microsoft on the back end of that

and it's going to be just a couple of months before

they start, really start adding tons of functionality.

You see that time and time again with Salesforce.

Salesforce will buy a company and then they'll put a bunch

of money behind it and now they've got a bunch of

AI tools that are just coming right off the press.

Honestly, I haven't touched a whole lot of

them just yet because they're so new.

But that's what got me really excited so

I'm curious to see a lot of what

you're know behind the scenes offline.

We'll be showing some of that off today to you guys.

That'd be awesome things that we're Microsoft.

So if you're a Microsoft shop, you're a company that uses

it, can you have access to it or have they released

it to an x number of beta or do you have

to go add it to your account like someone's listening right

now that doesn't know about it and doesn't have it.

How do they go learn more about it and

can they readily take advantage of the apps?

You said we're using a number of different power apps.

Yeah.

So Microsoft Power platform is readily available.

It's not in preview or anything like that.

It's a general availability thing.

What it comes down to is

it just depends on your contract.

I'm not going to speak to all the different

types of Microsoft contracts out there, but there's a

free tier of Powerapps and power automate on almost

every enterprise and business level of contract.

So you're able to leverage apps at a minimum with

SharePoint for I'm going to throw this in quotes free.

So it's included in the service there.

But when you look at what you can do

with it, that's just the tip of the iceberg,

I want to say there's over 400 premium connectors. Nice.

Which yeah, you have to pay more to have

access to, but that allows me to integrate it

over to Salesforce or make my own integrations and

things like that where maybe my business processes don't

all live in Salesforce, they live in Salesforce.

JD Edwards and all the different

possible organizational systems that are traditionally

been hard to integrate.

This just makes it so that you can pass

the data around and take action where it needs

to be and keep people working in their area.

So don't take a salesperson and make

them have to go into another system.

Keep them in their spot so that they can be familiar with

their layout and do the job that they need to do.

On the show we try to

educate people about AI, about automation.

So speaking to those folks, what is power automate?

I know the answer to that question, but for the

folks that don't, what do you use it for?

Is it an IPAs? I don't know.

Give some folks some context.

So where I've been using power automate a lot

is for approval workflows and it handles approvals beautifully.

They have a really good approvals engine behind it.

That's just built in and out

of the box available to you.

But you can also just do integration work

or any automation almost you can think of.

That's in the Microsoft realm of when I get an

email and then you can start putting filters on that.

When I get an email from a

bot, ignore it, things like that.

When I get an email from Ava, we're not going

to respond to Ava anymore, but you can build all

different types of automation that you want to log that

email into Salesforce, you could have it when I get

this email from this person, log it against this account

or go look up the account that it might be

associated with in Salesforce and log that over there.

Just again, power automate is a way

to just make your job easier.

So those actions that are repetitive that you have

to do every day, you don't have to.

You build a workflow one time

and then you can apply it.

I love the idea of not automating

people's jobs, but giving them access to

automate, the things that they don't like.

And we work a lot with workato in

VJ's book that he just came out with.

That's his mentality of, well, we're not going to

build a set of automations for the people.

We're going to put the toolbox in the people's

hands and let the people build whatever they need.

But you were talking about the email deal.

I'm building out my own version of this because

I'm overloaded with, I wouldn't call it spam, it's

b to b, like the cold email stuff.

And so you can, I guess, ban the whole domain. Right?

Yeah.

Which I mean, maybe that's the band hammer, right.

But some of them are genuinely interesting. Right.

And I think we've responded to some of them.

So I sit in the middle on that front.

But one of the things I've wanted to do, or one

of the things I'm building now is pick up every single

email out of my own box and run it through AI. Right.

On a scale of one to ten, how

likely is this a cold email from somebody? I don't know.

Right.

Well, it needs to check to see all my contacts and see

if I have access or if I know that person already.

Anyway, that's something that I'm already doing because I

need to clear out my inbox of stuff that

doesn't necessarily need to be spammed, but I don't

necessarily need to look at it right now.

Well, and it can do sentiment analysis and stuff

like that too, where it can be like, is

this a client who's mad at me?

And try and get an understanding of that

type of sentiment, which gets really interesting.

I mean, with the email stuff, you can easily just

automate it to take their emails and go look up

the company, have it go look at any of the

links and give me a digestible thing of am I

even going to be interested in following up with this?

Is this in line with these ten

principles that we're trying to capture?

And then you could even have it draft

a response for you if you wanted to.

I would never have it.

Send the response automatically. Send it.

We'll worry about it if it's wrong.

Right, they'll send another email back. Exactly.

They're agreeing to contracts for you.

Oh man.

Ava signing contracts and we don't

even know what it's about. Exactly.

I'm curious, you mentioned you're doing these

complex calculations inside of power automate.

Is it you doing it or is

it the engineer has access, like direct

access to basically automate their personal workflow.

So we're building the enterprise

apps right now, my team.

So we're building the app that then the engineer is

able to leverage those apps to do their design work.

And we're not using power automate

to do the heavy engineering calculations.

It's not an engineering calculator.

We're writing those as azure functions

that we can then expose.

Now, the really cool thing is we've built

a custom connector to those azure functions.

So if we wanted to allow our users to

build their own calculations on top of it. Nice.

They could access those calculations, pass the

right parameters in and get results back.

So we're putting more of that work in people's hands.

Yeah, if I'm sitting on the other side of the

show and I run a business, or I'm a manager,

like a director, but I haven't really touched AI.

Talking to the guy that touches all the

AI parts, like, where should I start?

Should I just open up

a power automate free subscription?

Or should I make sure my systems are set up

and ready to roll chat, GPT, start having conversations.

Treat it like an intern, just talk to it.

Maybe there's something you want to make for dinner.

I think you talked about a whiskey smash on one of

these previous episodes, but yeah, it can give you a whiskey

smash recipe, but maybe you want a whiskey smash that's more

sour than sweet, and see if you can get it to

kind of work with you to be more creative.

AI can be creative to a sense but it's creative to

the prompt, so it's going to deliver an answer to your

prompt, or maybe you have ten things in your fridge and

you need to figure out what to make for dinner.

Put it into your home life versus your work life first.

It's so interesting that he says that because

I have been the opposite this whole time.

I keep telling Chase, I'm like, man, I totally

adopted it for all of my work things, but

it hasn't transitioned into my personal life.

I'm not quite there to where instead of Google

I'm in chat GBT and I'm asking it things.

And I don't know if it's because since I've adopted it

in the like, my brain hasn't hit that switch yet.

Curious, though, is that where you started?

Did you start with Chat GPT and you

started with recipes or you're giving us kind

of a simple example for the layman?

So I started with Chat GPT.

I didn't necessarily start with recipes, but

we had a problem to solve.

We had to figure out a way to

write a C sharp plugin for the dataverse.

Had no background in it, had no background in C

sharp, was able to talk to it, talk about, okay,

what do we need for our development environment?

Okay, cool, I have that.

What do I need for?

And is that in the last twelve months?

Yes, this was February. Okay.

And that was my first little into working

with Chat GPT and then with AI.

I had one of the junior engineers say,

hey, what do you know about Chat GPT?

And I'm like, I don't really know anything

about it, but I'm going to learn.

So that's what I do.

If somebody comes up with an idea, I'm going to

go look at it and try and figure out what

path it could actually be applied for what we're doing.

And now everybody in my organization is like, okay,

how can we point it at our documentation?

How can we help our people internally

by being able to find information better.

There's a lot of work that can go into that.

There's a lot of different paths to take.

I was just about to say that one of

the biggest issues that every organization is going to

have is how do you have clean data?

You don't want to solve that problem.

Everybody will be very happy, I'll be a billionaire.

Solve that problem easily.

But it's one of those situations, figuring out clean

data and how to get your clean data to

be what you're coaching, that's the biggest challenge.

Then you can totally. Start coaching.

But you don't want to point

your agent at raw data necessarily.

I mean, it might be able to help you,

but you definitely have to take it with more

of a grain of salt at that point.

I love that.

Going back to the thought about the intern, right?

Don't fully trust it, but ask it.

One of the things that I've been doing on the

recipes, and Rachel, my wife, probably won't listen this far

into the show, but she's like, hey, can you get

the turkey ready to roll for Thanksgiving?

Or whatever the ask is?

I'm like, yeah, what's the temperature for

turkey that it needs to be at?

And I'm just trusting that it gives me

the right temperature because if it's hot enough

and it's not pink, then it's probably good. Great.

Mel's giving me this thing.

Good luck.

It all turned out pretty good.

But I love the idea of, like, you just chat with

it and just have like, hey, how do I do this?

And then how do I do that?

And then get into more complicated examples.

One of the things I love to do is just paste

in code that I know where the bug is and see

if it can pick out where the problem is.

And it's, at least in my experience, not as

good as that because it will pick out other

areas where there's potential, like, syntax issues.

Right.

Or, hey, you could have written this better.

And it doesn't necessarily pick out that one thing

that I was looking at, but it also clues

me into other areas when I'm writing code of

like, oh, well, yeah, maybe I should look at

this other suggestion and not just this one spot.

So I really love the idea of the super smart intern

that doesn't really know how to put it all together.

Well, and that's the hard part.

You're giving it a chunk of code, right?

You're not giving it the context, the whole full code.

So it may be pointing out an error

that it thinks is syntaxual, but it's actually

okay in your book, and it's missing that.

It's not seeing the force through the trees, right.

So there is some of that, of being able to

give it all the context that it can have.

And that's where I like using GitHub copilot because

it's in line with me inside the code.

So I'm able to have.

And now they have GitHub copilot chat where I

can talk about, hey, I'm getting this error.

Why am I getting this error?

I'm not sure why I'm getting this nullable error. Right.

I get more nullable errors than I know what

to do with because I just not paying attention.

Sometimes when I write that code, the

first pass through and go through the

different reasons why that could happen. Oh, okay. Yeah.

I didn't substantiate this variable. Right.

Sorry if that got way too technical

there for a second for people.

No, you're good.

So in copilot, it's looking at your

code, asking for the folks outside. Right.

It's looking at your code, but it's not just

looking at that one chunk that you're writing.

It's looking at the entire code base, at

least that you gave it access to. Yeah.

So it'll be plugged in to

my actual program that I'm writing.

So all the modules that it has, everything

that's in that folder that I'm working on,

it's looking at all of that.

So what gets really cool with that type of

work is, hey, I need to make 20 variables

and I have them all listed somewhere else.

I can basically just start writing and it'll realize after I

do one or two, oh, you want these other 17?

And it'll just punch through those other 17. For me.

It's a time saver. Right.

It just allows me not to have

to think about it and keep moving.

It reminds me, I don't think you're in Google sheets

a whole lot, but in Google sheets when you start

typing and then it suggests the rest of the formula.

Yeah, it's a lot like that.

Excel has very similar functionality. I love it.

They're all very similar functionality.

At this point we were talking about predictions

and some episodes that we've got going on.

I'm curious what you think 2024 holds.

Where do you see all this going?

My mind is going to be blown.

That's all I know.

Every week there's something new that's coming

out or I'm finding another use case

where I can use it for 2024.

I think there's going to be a greater adoption of it,

but at the same time I think there's going to be

legislation that comes out across the globe that's going to, I

don't want to say hamper it, but it's going to make

you take a different look at it. Right.

You're going to have to look at your AI and your usage

and how you're able to point it at the right data.

And you have to start thinking about that

now as you're starting to use it.

You don't want to point it at your Pii.

You want to make sure that your

AI is not necessarily looking at Pii.

Or if it is, you have to make

sure you do all these certain things.

Have all the things in place for Pii,

like you have to for Pii these days.

So it's going to be a crazy year in 2024,

and I can't even make a prediction, I'll be honest,

just because there's too much that could roll out what

seems like more like a fad that won't stick around

that you've seen over the past twelve months.

I don't know if I've seen

anything that won't stick around.

I think there's going to be a lot more

adoption of graphic development inside of using these things.

You dally two and all those different things.

But again, I think you made a point on a previous episode

where you can tell it to put a name on a boat,

but it doesn't know where to put that name on that boat.

It doesn't have that context.

And that's where you still are going to.

I think you're going to see everyone get more

efficient using these tools because a graphic designer can

do a lot of their work in one shot.

It's that first shot, right?

The first shot gets you 80%, and then you use

your skill set and you refine your skill set to

be that last 20%, that differentiator that really closes that

gap for winning clients or for performing your job and

just your day to day work.

Yeah, I really like that.

It's a big nugget.

Like, a lot of what you do 80% makes up a

majority of the things that we don't really love doing, maybe.

Right?

Just general statement, it's that 20%

that we really love doing.

And if you can find a way for something else

to tackle the 80%, generate the template, generate the outline,

write the sow, write the cold email for me. Right.

A lot of that makes up that 80%, so

we can focus on what we're really good at.

I'm curious, you mentioned a previous episode where I said,

I think I said it can't write the text. Right?

It can't type it because it doesn't understand.

And I don't know when that episode came out.

I forget the exact timing, but between then and now, OpenAI

came out with a model that now can write text.

Like, how crazy is that?

I was like, oh, I'm sure

I said, this will never happen.

It'll never be able to write text, and then all

of a sudden it can, like three months later.

That's what happens when we record like

a bucket of episodes and start to

release them and you're like, dang it.

Well, even though it can do that text, but is

it putting it in the right place on a boat? Right.

Like you were saying, you got to put it in the spot.

That makes sense for where it would be on a boat.

Totally. And it'll learn.

I mean, there's how many pictures out there are boats.

It'll figure out, oh, yeah, somebody wants

to put a name on a boat.

This is where it needs to go. Yeah.

I think it will further elevate

all of that original content.

We've talked about that a lot, and there's

going to be stuff that we're talking a

little bit more like work setting, but it'll

be interesting to see how much adoption of

these generated images or generated really any content.

We talked about video games a lot.

We've talked about just written content.

And even to the extent that there's essentially

like an AI recording, they have their own

YouTube channel or whatever it is.

And I think maybe our younger generations are going to

have more of a, less of an aversion to it.

And I'm also, like I said, I'm curious from

a, I try to think about the people in

my inner personal life of people who use AI,

and I honestly cannot tell you of anyone.

And maybe it's because I haven't

asked them, are you using it?

But I've got friends and family that I know hands down

have not opened or signed up for a Chachi BT account.

And now I'm feeling kind of guilty

for not, like, being this force of.

I keep telling everyone we need to

coexist with these tools like you did.

You're like, I've been using it since February.

I wanted to understand it.

I started learning it.

Next thing you know, you're on stage with

a VP at Microsoft talking about these things.

And I'm seeing this gap just widen between the folks

of us, the people like ourselves who have stayed curious,

the people who are listening to the podcast, the people

who have signed up for accounts, and then this mass

of people that I'm like, no, we're leaving them behind.

What are you doing?

Is your wife using it?

Do you have friends that are using it?

Do you have conversations?

I think in most of my friend

groups, we're definitely talking about it.

Rachel uses it through me because she just

steals my iPad and we'll type something in,

but she's not using it regularly.

That's her using it.

Well, yeah, she's using it on your device.

What about you, Chris?

So, yeah, Bink has been using it. My wife, Bink.

So shout out to my wife, Bink,

but she's been using it for just

asking different questions and things like that.

And I do encourage all my friends.

I mean, we live out just outside University of Notre Dame,

so I have a lot of friends who are professors.

And it's a very interesting view on

how does it work in academia, too. For sure.

Original work.

But again, it's an 80% draft.

It can help you get over that blank sheet of paper.

My personal opinion.

I totally agree over that blank sheet of paper. Yeah.

It's caused me to think about things on

the outer edges, especially when we talk about

the 80% that maybe we don't enjoy doing.

You spend so much brain power on

that, even if it's not overly creative. Brain power.

And so then by the time you sit down to generate

that original campaign idea, or at least for me in my

world, I've only got like 20% battery life left.

And so that's all it's getting, and I'll

like it and we'll come up with something.

But, man, when I've utilized it and put stuff

in there, you're right, it gets you past that

blank page or, that's a cool idea.

Have you thought about these other five and conversations

like this where it's just like creating, brainstorming around

other ways we can be using it?

That's probably one of the things that I've underutilized

the most, is when you ask it for a

creative answer instead of just saying, hey, what do

you think this, what's one answer?

I start saying like, hey, give me

five different thoughts on this, right?

Or give me ten.

Ask me questions.

Yeah, I want to talk to you about azure functions.

I want you to ask me ten questions about

what I want to do with an azure function.

And let's go down that path.

Let's have that conversation.

And I talk to it again like it's an intern or

like it's a colleague, and I just have that conversation.

I mean, you can do the same thing.

You could actually, with your kids, you can

set it up to do a storybook, right?

You could have it build a storybook and then have it generate

the dally two images next to it that go with it.

And you can have your kids get

asked the question and give the response,

and then they're building their own story.

So that's another way just to, again, put it

into the lives of people outside of work because

that's the easier place to adopt it.

Work is tough.

Trying to get an understanding of how will this really

impact me at work can be tough, but when you

have different things, everybody has to eat right.

Make a recipe book.

A lot of people have kids

or have family that have kids.

Make Christmas around the corner. Yeah.

Make a little custom book. Inspired.

I'm going to hit everybody.

I'm going out of town next week for Christmas and

no one's going to interact with me and not hear

about how passionate I am that they sign up for

their chat GP or actually, I don't know.

I've become a recent superfan of Claude.

I don't know if you've used

anthropic or anything of their tools.

We just got access to their API. API? Yeah.

Nice, man.

I have been having so much fun and I find

it to be a bit more cordial and personable, polite.

I don't know, there's something about it.

I am also very polite.

Back in my responses, I try to

be very kind and thank you.

Say please and thank you to the AI.

Please help me, please.

To your point about questions,

it's funny, I've actually never.

That's another key takeaway from this conversation,

to have it ask me questions.

But I recently, Claude did ask me,

well, I need to understand these things,

so can you please answer the following?

And, I mean, that might seem kind of basic, but I

thought I had given it enough context and I'd pumped in

a bunch of different previous examples and it was like, well,

without understanding exactly what your ICP is and that challenge or

that pain point, and then it listed like an additional seven

or eight bullets that I knew.

But again, that agent, that intern, didn't know,

and I didn't provide that level of detail.

So that's a good takeaway for me. Yeah.

Chris, I think we could talk about

this for probably another three plus, I

don't know, three months, three days.

We could make a whole summit about it for.

Yeah, but we don't have time for that.

So I think that, you know, obviously

you're a subject matter expert on this.

Chase has said this a lot before.

It's a level playing field, and

now's the time to use know.

We all have our own interesting

use cases or takes on it.

For the folks who want to learn more.

How can they get in touch with you?

How can they keep up with the things that you're doing?

Yeah, I'm going to put it out there.

I'm going to be doing this more on LinkedIn now soon.

So follow me on LinkedIn.

Reach out to me on LinkedIn. Ask me questions.

I'm more than happy to talk about anything

around the automation, the power apps, the AI.

I love it, I'm passionate about it.

So please reach out. Cool.

Well, man, thanks for joining us today.

It's been a pleasure.

I'm excited to hear about on LinkedIn, everything that

you're doing to get you to write about it.

But yeah, guys, reach out if you have any

questions here at the junction at ven technology, shoot

us an email, fan mail, write us chat, GPTs.

I don't know if that's not really a thing.

You have to go build it.

Oh, I got to build it first, right? Oh, man.

Until next time, guys, keep it automated.

Yeah.