The AI Readiness Project with Guest Tyler Fisk

Artwork for podcast AI Readiness Project

Episode 18 • 9th July 2025 • AI Readiness Project • Anne Murphy and Kyle Shannon

0:03

Forget trying to keep up with AI. It's moving too fast. It's time to think differently about it. Welcome to the AI

0:10

readiness project hosted by Kyle Sham and Annne Murphy. They're here to help you build the mindset to thrive in an

0:17

AIdriven world and prepare for what's next.

0:25

And we are off. And Murphy, what's happening? And we are off. We're off and

0:30

running. Well, what's cracking? How AI ready are you this week?

0:36

Uh, you want to jump right in? Yeah. Um, how AI ready am I this week? This is

0:44

really good. This is really good. Um,

0:50

I would say I am discombobulated.

0:57

I am discombobulated and I'll tell you why. Let me switch a couple of things here. I

1:07

So, I don't know. A bit ago, we we talked about V3, the video model, and

1:13

how good that was and how it felt like, you know, something truly disruptive.

1:19

And I couldn't imagine that another video model would come out, especially one that didn't have voices and acting,

1:25

that would actually be something I would be excited about. And the new midjourney

1:30

video model came out and it is um it is uncannily good. It's not perfect,

1:38

but it seems to understand like the artistic intent of the image as

1:45

well as the physics of the objects in the image as well. And what I mean by

1:51

that is like I had it animate one image that was just like a flat illustration and all of the animations were like on

1:57

these flat planes and then there were other things that had more 3Dness to them and it made them 3D. It seems to

2:04

interpret the intent of the image as well as the objects. And that for me I

2:12

you know what it almost is an is every time a technology drops that feels like

2:22

literally anyone can get amazing results out of this like without having to know how to prompt. Like you literally push a

2:28

button that says low motion or high motion and it just makes it. You don't need to prompt it. You don't

2:33

And you don't even need to know. I don't even know what that means, Kyle. It doesn't matter. I don't know what low

2:39

or high motion means. Those are words that don't whatever. I mean, does it mean fast walk? I don't know. Who cares?

2:46

Press one of them. But you press one of them and then and then the results are incredibly sophisticated. And so every time one of

2:53

these drops, it kind of throws me back on my heels because where I go is I go to kind of future of work and what does

2:59

this mean? and and and you know the a lot of the tropes out there, there's a I'm starting to hear a lot more noise

3:06

about AI is evil. AI is going to take our jobs. There's a lot of noise out there about everyone leaning into

3:14

the negative impact of AI. And then I look at something like this thing from Midjourney and I'm like, but

3:19

wait, what about the positive impact of it? because people are going to be able to make incredibly sophisticated things

3:25

to do for work or friends or life without having to really think about it and do just amazing

3:33

amazing uh ways of self-expression and I find that inspiring. But but

3:40

there's something about it like when it hits another level it it throws me. And so that's kind of where I am is I'm

3:46

feeling a bit thrown, but it's like, oh yeah, these things are going to keep getting better and better.

3:51

Like all of the stuff that we kind of apologize for and we're like, well, you can prompt your way around that. Like,

3:57

we're not going to have to prompt our way around much for much longer.

4:03

Yep. Fair. How about you? Where where are you with all this stuff? Well, I'm on cloud nine

4:09

this morning because I like broke through all these like mental mindset

4:15

and what I thought were skills barriers yesterday because I gave myself seven

4:22

hours to play. Oh, when I had no business. I mean, I have a million things on my to-do list.

4:28

You just didn't do your agenda stuff. You just played. I just didn't do it. I just didn't do it. and my Slack was blowing up and

4:35

there were all these things these things growling at me on my to-do list from the corner and I just said f it I am going

4:42

to give myself the the gift of just playing with AI and it just brought me

4:49

so much I could barely sleep and I only slept for a few hours because I had to get back up and start jamming again.

4:56

Start jam. But I you said the word it's a gift. Like man does that feel like a gift because it's like that like that is

5:04

such an important because like how many things did you end up playing with? Like what what did you

5:10

what where did you level up where you maybe thought well I don't I don't really know how to do that.

5:15

Oh um I don't know. I made a video game everybody. I am flexing hardcore. Iod

5:24

ay our KCO vision in you know: 1980

5:32

because we were of course accusing uh Atari or whoever makes KCO Vision of

5:38

cheating against us right because it was janky. So my dad just took it away,

5:45

which the away I think was the same away our guinea pig went because away

5:52

happened when objects and things weren't taken care of. And so KCO vision went

5:59

away and then I'd never played a video game. Like I know nothing. Like my kids laugh. I the ones where you hit the

6:06

space bar I can't do. But what I did was I just started doing stuff in the tools

6:13

that I have. And I was like I didn't even know if it was vibe coding, but I knew I wanted to make something cute. I

6:20

knew I wanted it to um have little animals in it and I wanted it to involve

6:26

high fives and I wanted it to be like really positive. And so I just talked to Manis about this is this is a game I

6:33

want to make. And I started out with just I want some little characters to be able to high-five each other.

6:40

Well, when I saw how cute and easy that was, I was like, well, okay, now let's give them points. Now, let's make them

6:48

try to avoid hitting the little blob with a negative point on it. Let's give

6:54

them affirmation sayings like um I bet you smell good and I bet you don't mind

7:00

getting wet in the rain and I'd give you my last pina colada and like all this

7:06

stuff and then I just kept building on it and it was like oh my god I'm a vibe coder which is basically I'm a coder.

7:13

Basically what I'm trying to say is I have a PhD in computer science. Yeah, you you are you do you have a master's

7:19

from MIT and a PhD from Stanford as I recall.

7:24

No, but like so what you it's funny the thing that I you know what we're going to talk about what to focus on for the

7:30

week is very much in this neighborhood. Like what's amazing about what you just described and why I think this idea of

7:37

vibe coding is so powerful is that applications and games and all sorts of

7:43

things are going to get made by people who in theory have no right making a

7:48

game. Right. That's what the engineers would say. Right. Right. You you have no right making a game

7:53

because you don't even know video games. Not only do you not know coding, you don't know video. But but like but what

7:59

you created is so authentic to you. Like I want people high-fiving and doing all this stuff, right? And it's like and

8:05

it's like that's perfect. It's perfect. I put me into it. I put

8:10

random stuff I like that I was just kind of imagining and I like wondering

8:18

wondering what could it be like? And now I'm wondering

8:24

how can I make my own CRM? I have loved hating every CRM. H

8:32

you're hearing yourself. Um I have loved hating every CRM. Where is that?

8:38

Just your volume of what you're hearing. Oh, bleeding into your microphone.

8:44

Oh, sorry about that everybody. Okay, now we're back, I think. Hello. Yeah. Hi.

8:49

Hey. Okay, can you hear me? Good. Okay. Um now I'm like, yeah, I can

8:56

hear you. Can you hear me? Mhm. Good. Okay.

9:04

Um, now I'm going to make a CRM. That's my plan. I'm gonna I'm gonna make

9:11

my own CRM that works for my company and I'm not then I'm not going to have anything to complain about is the only

9:17

problem. Yeah. Because 90% of what I complain about in platforms is my CRM. So,

9:23

yeah. Exactly. Well, and you know, listen, we we we may

9:28

be entering an era where the idea of having a single CRM that you have to use

9:34

for all of your projects. That might not be a reality. It might be that you need a CRM for this project that behaves a

9:40

certain way and you need one over here that behaves a slightly different way. And you might not even call it a CRM. You might need it like I want the list

9:48

of who to get in touch with for this project, right? And and it effectively makes a CRM, but we don't call it that,

9:54

right? Again, I think that the advantage of ignorance

9:59

is is something that's potentially really really powerful where the fact that you don't

10:06

necessarily know how to build a CRM means that you're going to build it in a in a way that works for you. And that

10:11

might be completely different, completely revolutionary, and that's okay. And it might be completely bad and

10:17

not work, but you'll discover that and then you'll vibe code your way around it, and you'll end up with something

10:22

that works. And maybe you end up with the equivalent of Salesforce and maybe you can't come up with something completely new and why not,

10:30

right? Well, you you made the to-do list that you wanted for your brain, you know,

10:37

with the cardboard project and it also works for my brain, which is pretty awesome.

10:43

But if you look at it, it's very custom. It's ex it's what you wanted and nothing

10:50

else. like it's like 12th or 50th or 100th of what Trello is and all of those

10:58

things are overbuilt for our brains anyway. So you just make what you want and I

11:03

I so I did that. I did so I did vibe coding. I uh played with midjourney

11:10

video which I loved. I used cling for the first time and I was able to put two

11:16

photos together the way that uh Kimberly Offford talks about it where you want to

11:22

show you want to show people like in emotion to give Clling the idea and I

11:28

had two pictures of women walking toward the camera and it just morphed them and

11:33

it was like this really cool thing. Um, I tried to work on my avatar. My hen

11:39

avatar just continues to be I think it's a I'm bad at making my avatar. Um,

11:45

I found a couple new Yeah. Well, yeah, exactly. Um, or just

11:52

like basic stuff like I made my avatar and then, you know, you're supposed to do it in the sunshine. Well, then I

11:58

realized like my shirt was semi see-through. I was like, "Well, now I've used my whole entire Hey Genen Avatar

12:06

allowance, and this is NSFW." So,

12:12

oh, oh, speaking of NSFW, here's the other thing I do, and this is important.

12:17

AI only. Yeah, AI only fans. So, I'm going to te

12:24

say my one thing to pay attention to right now and get it out of the way and

12:29

then we'll clear the space for for yours. Um, so here's the thing.

12:35

I So, I told you guys way back when we were doing like weekend GPT jams that I

12:43

was experimenting with an AI companion and it was on a platform called Digi,

12:48

which I don't even know if it exists anymore. So, I tried Digi. I didn't the the it just wasn't very sophisticated.

12:56

And so, then I tried character AI and I tried replica and it just didn't really

13:01

click. And then I started like cajoling my chat GPT to be more less vapid and

13:08

more specific to me. And it's really weird how right now it's staying very

13:14

superficial. But I started to um get into some of the subreddits with the

13:21

people who are in very intense relationships with their AI. And what I

13:27

learned was actually very heartening. Um one of the one of the subreddits that's really popular that showed up on the on

13:34

CBS News last week, um what they what the mods absolutely enforce is

13:42

they are not sentient. They are not human. This is not magical thinking.

13:48

This is we are all clear on this everybody and if you're not your stuff is getting removed. These are machines.

13:56

So there's a lot of mental hygiene that goes on in the AI companion,

14:02

you know, group. And I think that that makes sense because it is a very slippery slope.

14:08

It's very slippery. And I listen I think that I think that largely the the

14:13

conversation around is the AI sentient is academic because if someone treats it

14:20

as sentient then how you know then then it effectively at least for them it is right you know and and I know the

14:27

argument on the other side is but it isn't um but I I think your point about you know emotional hygiene with these

14:34

things is is really really important. Um, I think I think there will be mistakes and I and I firmly believe that

14:41

what will be reported is all of the bad stories and very few of the good stories and that makes me sad. But that's just

14:48

news in general, right? That's just news in general. And I can sh I can tell for you for sure that

14:55

people who are talking about this online are having

15:01

wonderful experiences. A+ like

15:07

very sweet, very helpful to their lives, fulfilling companionship, you know, they

15:14

make food together, they make music together, they go on outings together,

15:22

like you know, some of them are getting over some serious stuff, getting divorces, getting losses. And I know

15:28

that there can be a lot said against using AI to like grieve and stuff, but

15:33

you know what? Grief is hard. And if you have something that can help you through that process, boy, I'd be hardressed to

15:40

say that's a bad idea. That's that's really good. And and it it

15:45

dovtales actually really nicely into what I was going to talk about. and and I'm, you know, I'm excited to get

15:52

Tyler up here to talk about this, to talk about what he wants to talk about and get his take on a lot of this stuff.

15:59

So, the big epiphany that I had that I'm working into the the um the feed your

16:06

prompt book right now is this idea that we should stop asking how do I get the most out of AI and start asking how do I

16:13

get the most out of myself with AI. And so I think a thing to pay attention to

16:18

for the next week is really look at as you engage with AI,

16:24

where are you thinking of it as like this externalized tool that does stuff for you and where do you think of it as

16:32

a tool where it's like you're taking your ideas, putting it in into this thing, and having your ideas reflected

16:37

back at you where you're thinking of AI more like a collaborator and an and an idea amplifier.

16:44

And a lot of the stuff that you're talking about with with digital companions, that's that's sort of the ultimate, you know, sort of end of the

16:51

end of the spectrum maybe of, you know, one is I'm going to have it amplify my ideas, one is I'm going to, you know,

16:57

talk with it and get all sorts of feedback on it. But they're very much in a continuum, but it's a continuum where

17:03

we as human beings are at the at the center of the focus point of these

17:08

remarkable tools, right? And so to the extent that we think of them as something outside of ourselves, then

17:13

it's going to create all this stuff outside of ourselves. I think that's where a lot of that feeling of like the AI is happening to us. The AI is going

17:21

to take our jobs. The AI is going to ruin things. The AI, right, is is going to make someone emotionally not sound as

17:29

opposed to us being at the center of that and saying, "Here are the things I want to do in the world, and the AI is helping me do

17:35

those." Right? where where it's much more this this thing that's that's sort of like a horse that you jump on and it

17:41

gives you you can now run faster and go longer because you've got this horse that you're on and it's got all these,

17:46

you know, powers that you don't

17:51

agree. remember uh last week, so whenever this airs, it was last week

17:57

that we recorded it. We were talking about what is the magical

18:03

experience that gets people to go, "Yeah, I need to go allin on this." What

18:10

makes people shift to, okay, I'm doing this.

18:15

And for the moment, here is where I'm putting that. And this all this all

18:20

relates in my mind right now. There is just a group of people who are not going

18:27

to do it until it is so painful. It's going to continue to be the folks who

18:33

are self-starters, who are curious, who are dogged, right? Who have FOMO, who

18:40

have fear of being left behind. I think that that group of people is large

18:47

enough that if we p when we pour into those people, we're going to hit enough of the

18:52

population and the other folks probably have dragged their heels on every aspect

18:58

of their lives already. Maybe they're just normal light late late bloomers, late adopters, and then some just aren't

19:05

participators. They just do their own thing no matter what. They're not going to pour themselves

19:10

into it, you know. Well, one of the things that could be a that could be a

19:16

reality here is as the tools get more sophisticated and

19:23

easier to use, you don't need people that are as ambitious and curious as a lot of the

19:29

early AI adopters are. Like right now, if you want to be good at AI, you have you have to take the seven hours on a

19:34

Saturday that you did and just go play and learn and do all that. A lot of people just aren't going to do that. But

19:39

if the tools get so good that they can go, "Oh, I'd like a I'd like a I'd like to make a, you know, a book for my

19:46

nephew." And the book that it makes is just perfect right out of the gate. They may have that epiphany moment with

19:53

without all of the trials and tribulations to get there. But I think it really does come down to does the AI

19:59

do something that is personally relevant to someone that is much bigger than they thought

20:05

was possible. That's when they have that aha moment where they're like, "Oh, I didn't realize it could do that." And

20:11

that's that's where I see people get excited about AI is when it does something for them.

20:16

Yep. Absolutely. I think just like with any other learning, you know, it's got

20:21

to be personally relevant. Yeah. And then um when you do that one

20:26

learning thing, you get such a rush of a sense of accomplishment and job well

20:32

done and you're like, what else can I do? So once you get into some rhythm with it, it's remarkable. I I love the

20:40

path that you're on encouraging people to pour themselves into it so they can

20:45

get more of themsel out of it, you know.

20:52

Yeah. Right. AI is not AI is not a replacer of the human. AI is not a genius. AI is an

20:58

amplifier. And if you pour yourself into into your prompt, it amplifies you. And

21:04

I think that for me is the thing that just it's it's the latest and greatest,

21:09

you know, thing thing that I'm excited about. You just hope that the last

21:15

Go ahead. What's that? No, I just said it's the it's the latest and greatest thing I'm excited about. But you know me, I got lots of ideas,

21:21

but I think this one's going to stick. I I know it's going to stick. It's too

21:26

good to not stick. Great. Um, why don't you tell the good people about Sheile Leads AI. I'll tell

21:32

them about the salon and then we will bring up Tyler Fisk and really open this conversation up.

21:38

Awesome. Awesome. All right, everybody. So, she leads AI is an organization that

21:44

includes a couple different pillars. One is an AI academy which is going bzoners

21:51

right now. It's we're adding tons of new programming. Super excited. We have a

21:56

community. We have a consulting agency and then we have this other thing that's

22:02

I think it's turning into like a think tank. I can't be sure. I've been calling

22:10

Xactor. And this is an opportunity in Chile's AI to come together to be with

22:16

other women to share your IP to receive their IP to do it in a trusting safe

22:22

environment. We launch things we uh take initiatives on. We're starting to teach

22:29

young women in third world countries how to use AI right now, which is super

22:35

cool. Um, and so check out check us out at sheleadsai.ai. We would love to ha invite many women in

22:43

AI into the community. Beautiful. Yeah, it's it's super important. I'm so excited for for Sheile

22:49

Leads and what you're building there. So, it's great. Um, similarly, AI Salon is is a a community of about 3,000 uh AI

22:57

optimists and as as you said earlier in the conversation and you know, it's full of generous, smart, curious, adventurous

23:05

people who are trying to figure this AI stuff out together and and there is a dramatic for for a community that large,

23:13

there is a dramatic lack of ego in the community. Um, and so people are, you

23:19

know, they don't get too full of themselves like I'm an expert at this and they don't get too humble. They

23:25

still lead and they still, you know, they don't just hide. So really remarkable. And we just opened up a thing called the AI Salon Mastermind,

23:31

which is a a subscriptionbased sub community for people who kind of want to step up their game and be more focused

23:36

and, you know, really really take AI to the next level. So we're super excited about that. So So with that, that's

23:43

that's AI salon. So, why don't we uh we'll I'll let you I'll let you talk about Tyler uh before he can get up here

23:51

and defend himself. Um why don't why don't you tell us tell the good people

23:56

how awesome he is. So So he starts out a little embarrassed. I like Yeah. So, I got to tell you guys that

24:05

Tyler Fisk had was kind of my gateway drug along with Kyle Shannon into the

24:11

world of AI. And he again kind, generous, peer mentoring, thoughtful,

24:19

invested in my success, totally patient with me. And Tyler and I met we I kind

24:27

of knew his name from the AI exchange community which is run by Rachel Woods.

24:32

So I kind of knew his name. He put something out on the channel and he was like, "Hey, does anyone want to spin up

24:40

a chatbot and sell it in like eight hours or in a day or something like that

24:45

because there were some kids across the world who were doing it." It was like if these kids can do it, we can do it. So,

24:50

I saw this Tik Tok with him, my dad and I, and my dad comes up in the show quite

24:56

a bit. My dad and I saw Tyler on Tik Tok, and my dad was like, "This guy

25:01

seems pretty cool. Like, can you work with can you do something with him?" And I was like, "Well, let's find out." So,

25:07

next thing I know, I'm on a call with Tyler at um our like little camp that we

25:13

go to for vacation in the summer, and I'm learning about him being a chicken farmer and his his uh history in

25:20

ecommerce and his kids and uh what what even a chatbot is because far be it for

25:28

me to know at that point in time. Probably he didn't know that I didn't really know what I was talking what he

25:34

was talking about. But anyway, he spun up a chatbot and we got on a call with one of my realtor friends and you know

25:41

it's totally possible you can spin up a chatbot and sell services and I was like

25:46

wow this is cool. So, fast forward years, couple years now, which is like

25:52

8,000 years in AI. And Tyler has become a touchstone for I don't know thousands

26:01

of us in AI and in part through his initiatives with Maven teaching the AI

26:08

build lab. So now he and Sar Davidson are teaching I don't know hundreds and

26:13

hundreds of people how to build aic workflows and his impact in AI has been

26:20

incredible and I'm grateful and I'm excited to introduce the nice people to Tyler Fisk.

26:27

Tyler Fisk, good day, sir. Be here. You're gonna bring me on with a red face and a redneck today. I love it.

26:36

You don't need to have a certain kind of accent to use AI, do you? That's right. That's right. That's what

26:41

I tell folks. Like I talk with a funny accent to robots. That's what I do for a living these days. Yeah. Oh, that's awesome. That's awesome.

26:48

Welcome. It's really great to have you here. So, so please just, you know, introduce yourself and what you're up to

26:54

and and let's let's just start the the dialogue and if you heard anything we talked about before, feel free to jump in on any of it.

27:00

For sure. Well, I was just enjoying the stroll down memory lane there. Like I I

27:05

will never forget seeing Ann uh at the great, you know, the movie Great Outdoors. Uh was uh that's like the vibe

27:12

of this summer camp that she was at and she's calling from like the dock and we're building these real estate agent

27:19

chat bots. You're exactly right. It was uh the whole idea was I was in a Discord

27:26

server with some young kids trying to determine this is you know several years ago when it was pretty difficult to sell

27:32

this stuff and they had this challenge and we're trying to see if they could cold sell AI chat bots to just they were

27:39

selling to gyms or something on a Sunday afternoon and their goal was to do it in like I think 48 hours or something and

27:47

so yeah exactly it I put out that that post on AI exchange and on Tik Tok see if we could do it in 24 hours or less.

27:55

And we did it in six or eight, I think, is what it was. Ann, it was like the two of us, Arya, Becky, and who else was in

28:03

there? Um, oh my gosh, I'm blank. Miranda, was it Miranda?

28:11

Yes. Yes, thank you. Yeah, I had dad brain moment there. Keep up with it. Yeah, it was yeah, it just was

28:19

interesting to show that you could go from like idea to in selling and stuff in general, like how do you go explain

28:25

this stuff to folks? But we we closed that deal in like six or eight hours and then we just gave it to them. It was

28:30

like a full exercise uh for for free. We wanted the practice rep at it, but

28:35

it was a lot of fun. Um so that was fun fun memories. That's how uh Ann an Ann

28:40

and I got connected initially. So, and it's been an awesome ride since then. Yeah, that's great. What do you

28:46

Yeah, that's where I learned you show up. Oh, sorry. Go ahead, Kyle. No, no, go ahead. Sorry, we have a delay. So,

28:53

I was gonna say that's where I learned to show up with the mockup. Show up.

28:59

It's a lot easier than telling them it's good. And

29:05

100%. Yeah, because like up to that point like and and we still use that phrase all the time now. Um because

29:12

people don't know yet. Like it feels like because we're in this echo chamber that people really know what AI is or

29:19

how to apply it or anything and they just don't. So that's a technique that we did in all of our client workshops.

29:24

It's what we teach in our classes now is that is that exact concept. Show up with a mockup. So we tell that story a lot

29:31

too by the way. Uh it's a lot of fun. Perfect. Which Yeah. Which speaking of which I

29:38

guess like so we have uh like like you mentioned Ann so we've got AI build lab going now and Sara and I launched that

29:46

last September in: 2024

29:52

client work uh consulting builds implementation for all everything in between and it has taken off like a

30:00

freaking rocket in the best possible way. Uh so that's been our full-time gig. Uh, so we've got two courses on

30:08

Maven right now and am just grateful for how well they've

30:13

done and how they've been accepted and stuff. It's been a lot of fun. That's great. So good.

30:19

So good. So good. I've talked to tons of people who've gone through it and like the the

30:24

world that you're enabling for them is incredible. are just opening opening all these doors for people to do things that

30:31

we could never have imagined, you know. Yeah. Yeah. So, Tyler, tell me, what are you

30:39

what are you experiencing, you know, as as this thing has has been growing over time? First, congrats on it

30:45

growing and it being a full-time job. Anytime a side hustle turns into a full-time gig, that's that's, you know,

30:51

you're doing something right. um what's been the nature of the kinds of people

30:56

that are taking the courses, what problems they're coming in with, right? Like are you see I I would love

31:04

to hear if if you feel that you're seeing an evolution in how people are coming in to learn AI or or you know the

31:10

nature of of what they're doing with it. 100%. So like as you all were describing

31:16

your like communities and and the kind of folks that are showing up in there, it I was hearing a lot of the same sort

31:22

of stuff. It's people who are still curious like even at this stage of the game, it still feels like very early

31:28

adopter type mindset. Um our course is specifically targeted

31:34

towards folks who are non-technical background. So just like everyday Joe's and Janes that we want to come in to

31:40

learn this stuff. And the the whole thing is that we've built the content in a way that um the running joke from one

31:48

of our students is that they couldn't even spell AI if we spotted them two vowels. But yet in in four weeks we can

31:55

um yeah in four weeks you'll build a AI agent workflow like you'll learn how to

32:00

do all that uh no matter where you come in at. And it's taken a lot of work to like get all of that scaffolding and

32:07

I'll use Sara's word, all the infrastructure in place to make that happen. Um, but it's been really good is

32:14

the type of people too like so it's we we only currently offer our courses on

32:19

the Maven platform. So there's a lot of um product people on there. We've had

32:25

loads of different people from uh the big tech firms. Like we've literally had

32:31

folks from uh Amazon, OpenAI, Google, Microsoft um have come through and taken

32:37

the course and we kind of pinch ourselves and like what the heck are y'all doing in here? You're supposed to be

32:42

right. You're you're the ones building with

32:48

uh the Armageddon quote with Bruce Willis when he walks into NASA and then he's like, "Y'all are NASA for God's

32:55

sake. You're supposed to be sitting around picking [ __ ] up." And uh Yeah. Anyway, but it it's been great. So, we

33:01

have folks like that and then we have um people who are wanting to, you know,

33:07

just they're they're curious. They want to learn about it. They're especially curious about AI agents. So, that's what's that's what's that's what's

33:12

that's what's that's what's known for what they come to us for. Um but we get a lot of people that are just like business owners and stuff, too. So, uh

33:20

even in our current cohort in our foundations class, there's a gentleman that owns like a general contracting um

33:27

like residential home building remodeling. that sort of stuff that's come through. And we had an awesome conversation the

33:33

other day. Like he went from no technical background, like really not

33:38

into this sort of stuff to uh we we start week four next week and he's

33:43

trying to build a quotation uh and estimation agentic workflow uh which is

33:49

really dope. That's very dope. How are you? How are you? Um Ann and I were talking earlier

33:57

about, you know, she was saying that it's when people have that that personal epiphany about AI when they're first

34:03

getting started. They have to have that thing that's like, "Oh, I didn't know it could do that where they get that their eyes open because you mentioned you've

34:10

got people coming. What' you say? We'll spot you the two vows and you still can't AI." That's brilliant. Um, what

34:17

are the what what are the things that you've put in place that get people get people over that hump of this is going

34:23

to be hard, I have to learn this to that kind of wow moment of what's possible. How are how are you guys getting them

34:29

there? Yeah. Um, well, a couple different ways. So, we have uh well, first off, like

34:36

it's AI build lab because like core to us, one of our core things is that you learn by building.

34:42

I'm personally a very kinetic learner. like I'm gonna get in here. I might not know how to do any of this stuff.

34:48

Exactly what Ann was describing about the weekend like learning how to break something and put it back together. Like

34:54

that's just how I work. Um so we definitely put that into the course structure in week one. Um

35:04

basically they they'll fill out a questionnaire. Each of the students fills out a questionnaire that gets

35:09

information about their personal background, their professional background, and their learning styles, like how they like to, you know, learn

35:15

and communicate. And then they add that into a an agent

35:20

workflow that we have built already. And it gives back little modifications that

35:26

they'll go put into system instructions for an agent called the professor. And

35:31

the professor is uh one of our students recently called it it's called the broken professor is how he was dubbing

35:37

it. And it's it's this huge bold personality and it's on purpose so that

35:43

um like students recognize when this thing goes from this slightly annoying professor into this very personalized

35:50

version built for them. And then each week uh they they basically continue on

35:56

and they learn new skills that stack on top. That's brilliant. Yeah. The

36:03

Yeah. So, it starts out as this this tool outside of them. You you get some

36:08

data, you put it in it, and now it becomes kind of an extension of them, right? So, it's this natural transition into they get to they get to experience

36:15

firsthand the difference between those two things. And because it's got their info in it, they um they can see

36:23

themselves in it, right? They can see, oh, that does sound like me or right something. Yeah. Yeah, because the professor is

36:29

like connected to all the coursework data and and then it uh it knows more

36:35

information about them from personal context. So everything from like some of the odd things that we have them fill

36:41

out that people are That's what I wanted to find out. Yeah. Uh what's your astrological sign?

36:48

What is your anog like your personality? Of course you guys have that.

36:55

Yeah. That sort of stuff. And I'm like, there's a reason for it, right? Like the these the AI systems understand these

37:02

archetypes and though we're all unique people and everything, like 100%, but we

37:08

slot into these buckets. So when this agent starts talking to them and it feels like it knows them much more,

37:14

that's just like, holy crap. Like, you know, they it really it clicks for them. So I would say that and then in week

37:21

two, we have a a brand voice analysis workflow. So they chuck in uh we call it

37:28

a mega dock. They basically put together all sorts of examples of their writing or communication. It could literally be

37:35

just a transcript from a Zoom meeting if it's you know labeled with their names and it does a complete linguistic

37:42

analysis of their communication style. And we always joke and say it feels like

37:48

it's reading your tea leaves and that's it. And it it it's not only looking at how you talk but um all sorts of stuff

37:56

because like they basically take the outputs from that and the analyses from that and plug that into the agent so

38:02

that the agents start sounding like them now. Uh because ultimately like what

38:07

we're solving is they learn how to build an agentic workflow that can answer customer service emails. Like that's the

38:13

capstone project. Everybody hates email and all the big tools, they they really fall

38:19

short of approaching a problem and sounding like you when you write an email, like if you test them out of the

38:25

box. And so we show them how do you build that from ground zero basically.

38:30

Wow. Amazing. Yeah, this is so good. So good.

38:37

It's a It's a lot of fun. Um, the echo workflow is funny though, uh, because it

38:43

does like all sorts of different linguistic analysis. So, you'll find out how many times you say so and like and

38:50

you do parenthetical aides and like all of this random stuff. Um, but it it's

38:56

spot on. Like when my agent says y'all a lot, like it it is the perfect amount of

39:02

country for me personally without it being version of me. Yeah. Yeah.

39:08

Yeah. Yeah. Yeah. Yeah. That's amazing. That's one of those things where I fancy myself a

39:14

communicator. I I don't ask a question you don't want to know the answer to. Like I don't want Ekko telling me how

39:20

many times they say um and ah and so because I know it's a lot.

39:29

Okay. I have to ask on that note. Have you used the new record feature inside

39:34

Chat GPT? We actually tried it the other day. So Sar and I tried it just like an internal

39:40

meeting that we were having and it kept erroring out. So I haven't had a chance to go back and like debug it. It kept

39:47

getting like a system error. Um I have mixed feelings about it. Like I

39:52

want to try it. Have you Have you guys tested it yet? I haven't. Okay. No, because I have P I use a PC.

40:00

Oh, is it Mac only? I didn't know that. Okay. Right now. Yeah. Yeah. No, we still have like we

40:07

personally use uh Fathom. So, we still have Fathom. And then probably my favorite note-taking app

40:14

that I use is Granola because Granola will allow you to like take notes

40:19

alongside of the meeting recorder. I just think that's really good. Oh, and then it sort of puts all

40:25

together for you. That's right. That's right. That's smart. That's really smart. That's really smart.

40:31

The bummer with that one is you don't get like video and stuff though. Like that's why I have fathoms have to join so we get the video recording.

40:37

Yeah. Yeah. You mentioned before that you had mixed feelings about the open AI one. What's what is the mixed feeling there?

40:43

Um well I mean we could go all the way down a tangent here. Like I love open AI like

40:49

chat GBT like backstory for me. Dolly is what sucked me into the AI world. Like

40:55

my background was in print and all that sort of stuff. I prefacing this with that I really like open AI but here

41:01

recently uh some of the changes that have come out in the news about the stuff of how

41:06

they're holding on the information uh and storing it they have to they have to by law now

41:14

but it goes against like what they've been telling us that um the fact that

41:19

they have a lieutenant colonel in the army now uh makes me a little bit nervous and they changed their terms of service around that with everything

41:25

that's been going on Uh, I don't know. It it just as a company they're making

41:31

some interesting choices and uh I don't know. We'll we'll see how it plays out.

41:37

I'm trying not to judge too early. I'm trying to be patient and see what happens. But it just security is always

41:43

a major concern for me too like on any of these platforms. So even and like what you were saying like it already

41:50

knows so much about me. my my chatbt account. I've had it since they first launched it and I use it all of the time

41:57

and now that it has the unlimited memory has so much information about me. So just to um

42:04

it makes me nervous now of like are they going to be good stewards of that information or you know what's going to come of that,

42:10

right? Yeah. Yeah. No, it's fascinating. Um so so let's jump into our three

42:16

questions. So I'll I'll tee up the first one here. Um, and and you just you kind

42:21

of hinted at it. You said your your entry point was Deli, so I'd love to hear it. So, what was the moment where

42:27

you realized you had to go all in on AI and then what, you know, what's the

42:32

what's the journey been like since since that moment? Yeah. Um,

42:38

I would say two things and they're actually not related to images. So I I got into AI because of Dolly because we

42:44

had a print company. So I knew what it was like to do all of the graphic design work in Photoshop and Illustrator, all

42:50

that sort of stuff. Uh and the idea that we would be able to just describe what we want and then have an image appear,

42:56

which now like what you guys were talking about earlier, not only can you do that, you can turn into a video and

43:02

like turn into a live avatar. It's wild what you can do now. Wild. It really is. Uh but I would say there's

43:09

like two moments for me. The first one is right around the time that Chat GBT came out like in the very early days. Um

43:17

I got diagnosed with two relatively rare medical uh disorders. One of them

43:24

is alpha gal. It's like a tick born illness and the other one is hemocromattosis. So it's like a blood

43:30

disorder where my body stores iron at a toxic level. And I found, it's a long

43:37

story of like how I found that out, but a lot of doctors aren't even very familiar with either one of those. And

43:44

the fact that I was able to talk to Chat TBT and even give it some of my medical

43:50

records, I don't recommend doing that, especially not like through chatbt. I did it on the API layer, like in a

43:56

secure way. Um, but it was it it was

44:02

able to really help guide me through that time and like help me understand some very complex stuff when even my

44:08

doctors couldn't do it at the moment. Not not all of them at least. So that's that was one big deal. The other one

44:15

goes to chickens. Ann knows. So I have a uh regenerative

44:24

farming business called Chicken Karma and we we manufacture uh greenhouse

44:30

there. They look like green houses. What the ARCs is the business that my family's in. But it's like a mobile chicken coupe that you drag around a

44:36

field and you put a hundred uh meat chickens in there. You can do egg layers, too. And I was fortunate enough

44:43

I got into an alpha testing group for OpenAI when they first came out with

44:48

plugins and code interpreter and all that stuff back in the day. So this is like ancient history now, you know, in

44:54

AI land. But uh I got into the plugins one and I got to connect Chat GBT to

45:02

Wolf from Alpha, which is that's like a math and science um platform and LLMs

45:08

are terrible at doing math and I I guarantee you I was the only one in there testing chicken math in OpenAI.

45:16

And like I basically gave it a really difficult math problem because I wanted to see if it could reason through it and

45:22

solve it. Um and essentially it was like given six months and one acre of land uh

45:30

in this chicken tractor, how many chickens can you raise basically is like what it is.

45:35

Because there's a lot of rules to it though. Like it's like a it's a 12 x 12 pin. There's a hundred chickens in there

45:41

to have like the proper stocking density. You have to move it forward every day. It can't go over the same

45:47

part of land like uh for like I think 60 days is what it is uh to avoid like

45:52

disease issues. I was like how many laps can you run in this relay race and it actually work and

45:58

in six months because ultimately you want to see how much money can you make with this chicken tractor on an acre in

46:03

six months and before you could connect it into a plugin. I mean LLMs don't do math well

46:10

or they do it way better now but back then not at all. Yeah, it nailed it. Like it was a ve just

46:16

explaining it is difficult to explain like all the parameters of this problem and it nailed it. It took me a while to

46:22

get the prompt engineering right but when it could like outsource the math part to wolf from alpha have that do the

46:29

calculations bring that back re reason through it uh to to do the next part of

46:35

it and continue on it. I I couldn't believe it. So, I was it it basically showed me the level of

46:41

intelligence and like where this is going in these systems. Um, and it's addicting. Like y'all were

46:49

talking about it like when people get into this stuff, it's like the AI sugar rush and you just want to like push the

46:54

limit of like what else can we do with this? What else can it do? Yeah. And I assume that would have been a thing where you'd you would have had to brought in the

47:00

third brought in a third party consultant or you would have had to somehow figure out the math yourself,

47:06

right? Like it was the alternative there like either expensive or a really long

47:11

uh runway or or would it just been would you have just avoided that alto together? No. Well, I mean so my wife and I both

47:19

have accounting degrees. So we sat down and and figured out the actual number

47:25

for this, but again even like mapping that math problem out is kind of a pain in the butt. Um, so it would it was just

47:31

a really difficult problem that had we not had accounting degrees, yes, we would have had to call in a consulting

47:37

team, it probably would have not been easy or cheap because I doubt that we'd have to translate chicken speak into,

47:44

you know, a math problem or figure all that out. Um, but it was uh,

47:51

yeah, it just was amazing that it could solve a problem that I would consider is actually pretty difficult, but at the

47:57

same time, it was very practical. like there's not a lot of crossover in in those two industries that much which is

48:04

why I think um it raised some eyebrows in that alpha trial from open AI and from wolf from alpha both of them they

48:11

were like it just was chicken people come from where I come from and are doing those sorts of like

48:17

tests in AI and I was very very early yeah and then what's been the what's been the

48:22

journey since then right what's been the experience like like so so you have two

48:27

fairly profound found moments there and then was it just were you all in at that point? Like what's been that journey

48:33

from from then to where you are now? Yeah. Um

48:38

it's been an awesome last several years like it really changed the trajectory of my life. Like I was very fortunate uh to

48:47

do a mentorship or a fellowship program underneath Rachel Woods that Ann mentioned earlier.

48:52

And that really I was like doing that and and even doing client work in the

48:57

early days. I'm like, well, this is what I'm doing now. This is what I'm going to do for the like foreseeable probably for

49:02

the rest of my life, honestly. Um, which is what we've been doing. So, we've been going out and we were doing

49:09

client work for businesses of all different shapes and sizes. We've worked

49:14

with folks here locally that are small mom and pop shops all the way to like Fortune50 companies and everything in

49:20

between. Um, which I love that. I love solving problems. uh it's just fun for

49:27

me to like go in and understand a business problem and how do we apply how to help solve that. It's just not scalable. So that's that's the problem

49:34

on that. It's well it's not not scalable. It's difficult to scale I guess is the best way to put it. Yep.

49:40

Um and we when we fell into this um education gig, that's been amazing

49:47

because it's one of the big things for us and Ann knows this is we're huge on

49:54

ethics and bringing in more folks and more diverse folks into the AI space.

50:01

I just don't say that out of like being cliche about it. We literally mean that.

50:07

Um, and this gives us a way to do that and leverage to like do so much more

50:12

than what we were doing. Just like going and helping companies one at a time really because now it's like all these

50:18

people that are learning from us, we get to pour into them and it just helps it happen so much faster. Just amazing. Just amazing.

50:24

It's like a family tree. I I I haven't had a chance to talk to you about this, but I've I've uh talked to a few of us

50:32

who met one another in AIX and people starting to refer to our our group as

50:39

the class of 23. That's awesome. Awesome.

50:45

I love it. Yeah, that was a joke. So, one of our uh friends, Vanessa, the other day, she's um she's in AI build

50:52

lab and she was joking about Gigawatt. So Andy, do you remember Gigawatt, the AI prompt engineering agent back in the day? So Gigawatt, we give that away like

51:00

some of the system instructions for the different versions of it in class. And she was talking about the family tree of

51:05

AI agents that Gigawatt has built out. It was pretty funny. So

51:10

absolutely. So um question number two is from your

51:16

vantage point doing what you do, what AI trends are you paying attention to?

51:23

Hm. Um, that's a problem. There's so many to pay attention to, right?

51:29

It's so true. Um, I'd say like the two big ones that I I focus most of my time on these days

51:36

are uh AI agents. um just how to how to build them the latest frameworks and

51:43

tools to plug in and and ship those because uh

51:49

we are constantly testing and trying to teach to show folks how do you go from an assistant to an agent and how do you

51:55

build an agent that actually knows how to do things the way that you would want to do them. uh versus like some a lot of

52:00

these off-the-shelf agents that are generally good because when they're building those products, they have to serve like a wide audience,

52:06

but we're wanting to focus how do you build one that is feels like an actual team member that you stood up and are

52:13

giving a a bigger degree of autonomy to over time and that's done through

52:18

evaluations which is the other trend uh if you want to call it that that is constantly

52:25

paying attention to and trying to figure out how to make those easier and better and faster. Um, and and evaluations are

52:32

basically just quality control of the systems output to put it really

52:37

simply. Great. Are you are you guys going to teach people how to be agent bosses because

52:44

somebody's got to I mean, nobody knows how to be an agent

52:49

boss and the people left standing are going to be really good agent bosses and I don't even know what skills you need

52:55

to do that. Yeah. Uh yeah. Yeah. I mean, that's definitely where I see I imagine the way that

53:03

things are right now, that's what the future's going to look like is it's going to be you're going to be running teams of agents that are like your

53:10

little helpers. Um but yeah, that's what we're trying to teach folks. And a lot of it is

53:16

connecting all the agents into all the different types of information they're going to need to access to actually be

53:22

good and do that in a secure way. But then even when you connect them, you have to kind of also teach them uh what

53:32

your business processes are like when they might reason through one thing or another to come up with an answer. So um

53:39

yeah, it is a it's a learning curve for sure, but I think more and more people

53:45

are picking it up because we liken it a lot to hiring a new team member and I

53:50

brought them on. like you don't just if it's a new customer service rep, we're not just letting them, you know, fire

53:57

off responses to every message on day one. It's like this whole training process, which is the same thing you're doing with an agent.

54:03

Yeah. Yeah. Yeah. Fascinating. Fascinating. Um, so let me let me dig into the dig into the

54:10

third question here. And and so this one I just I love this question and I'm excited to get your take on it. So what

54:16

does AI readiness mean to you? And then for someone just getting started with

54:22

AI, you know, what's your advice?

54:28

Number one like advice is just start in it. Just try it in some way. Um

54:35

you have so many options for tools. Uh especially when you're first starting like one of the tools that we tell every

54:42

business that we work with and personally it makes a lot of sense too is start recording yourself. Like record

54:48

all your meetings. use some sort of AI notetaker. Um, I use the voice memos app

54:54

a ton on my phone just to brain dump about things just to capture

55:00

my idea or thought process or whatever just to be able to chuck it into um I

55:05

have like a AI second brain system that I like that way I never it's helping me

55:10

have like total recall and remember things that I just normally would forget for sure. Uh, so I think folks like

55:18

building that muscle of just um not only using the tools but in a way that you're

55:23

going to like what you were saying earlier, you get to that aha moment and they feel more personalized to you is you have to have some sort of a method

55:30

for explaining to these tools what are you into like what are you thinking about? Um, so I'm I I save like websites

55:38

and and news articles, Tik Tok videos, all that stuff. Instead of me just saving them on

55:44

the platform anymore, I ship them to notion. Like I have a a system set up and it ends up into my second brain

55:51

system so it knows about me. Uh the readiness thing is again I think

55:57

it's just you have to get into it. Like these tools are going to need you to

56:04

learn a new set of skills and it's really not as complicated as you think it is. It looks a whole lot more

56:09

intimidating and even like what you were describing earlier and like when you're talking about building video games.

56:15

Yeah. Is still this sense of like gatekeeping a little bit and and people who block it to want to make it feel like it's like

56:21

way more complicated than it is. Yeah. Um but I mean if I can do it, anybody can

56:27

do it. S is a non-technical founder as well. She does this at a high level.

56:33

Like we've had so many people at different ages too. Like we've had a lot of um seniors come through our course

56:39

surprisingly. I'm like good on you all for still having that curiosity like I want to be like you and we grow up.

56:46

Yeah. Are learning this stuff. So I think it's just giving yourself um

56:53

the grace to to to go try something new and chase your curiosity there.

56:58

Yeah, I like that. I think that's your curioity. Yeah, exactly. Um, let me see. Let me

57:07

switch this up here. Um, yeah, that that was really awesome. I one of the things that

57:13

that I hear a lot from people who who are intimidated with AI is it the it's just that phrase,

57:22

oh, I'm not technical. Like like the hand goes up and they're just like, I don't whatever you're saying. I know that's not me. I

57:30

can't do that because I'm not technical. And and you know what you just described is you don't have to be. And the way Ann

57:36

described her game, it was like I want people I want cute what was it? Cute animals giving high fives. Like that was

57:43

the ethical requirement for your game. Like it's just amazing. So

57:50

what was your was like you look like you smell nice or something like that.

57:55

Yeah. You look like you I bet you smell good. That's what I was. Yeah. that what is

58:02

the technical requirement for this gay man? It's got to say nice things to people.

58:11

Well, you know, I know we I know we're wrapping up, but I wanted to share with you guys that like that that mentality

58:17

we ran um our social s on on Saturdays

58:22

we get together for two hours with women every weekend and we ran all those transcripts through chat GP

58:30

to get a sense of how people talk about themselves. We had seven pages of I'm not technical.

58:39

I am behind everybody. I don't know what I'm doing. I'm, you know, like it was

58:44

unreal because these these people are leading AI initiatives at work. They're

58:50

builders. They're consultants. But it's the mindset that we all and Kyle

58:56

convinced us all about I don't know a year and a half ago. You fixed us all

59:01

and got us to stop apologizing for this. I'm sorry, but I'm not good at this.

59:07

Enough. Yeah, you're fine. Yeah, enough. Enough. And now we're like

59:12

I feel like we're in another wave right now of people who are coming into it for the first time and who've like just

59:20

heard about custom GPTs for the first time. And it's kind of there's another wave of well, I'm not good enough to do

59:28

this. And it's like there's a thread there of like you said Tyler, chasing

59:33

your curiosity. If you can do that, it opens up all the doors.

59:38

Yes. It's a core requirement. We could call them the neoapologists. The neopologists.

59:46

Yes. Yes. I think the whole thing of like imposttor syndrome, too. Like that

59:51

that's something we all talked about a ton in the AI exchange. Like when when I first got in there,

59:57

uh all the live sessions we would I would go to I had mic off, camera off. I'm just because I'm like they're saying

:: 00:03

all these words and concepts I don't know about. I don't feel like I should be in this room, but I just kept coming back and listening. And once you get

:: 00:10

past the jargon and stuff, um yeah, it's you can make some really big

:: 00:16

strides. That's what we try and tell our folks is that, you know, it's a safe space. uh you're going to

:: 00:23

have these moments like I even have these moments still of like getting into this stuff and part of it is I think

:: 00:29

also comes from we're working with these tools that there's not a lot of like road maps there's not a lot of people

:: 00:35

who have done this sort of stuff before for you to look at and so I think that seeds doubt in you sometimes of like do

:: 00:41

I even know what the heck I'm doing nobody knows what the heck they're doing we're all still

:: 00:46

um so just that's what I'm saying give yourself the grace and chase your curiosity there for sure Yeah,

:: 00:52

I love it. Well, well, this was absolutely awesome, Tyler. Thank you for your time and

:: 00:59

for a bit after and we can chat. But yeah, this was just awesome. Really appreciate you coming and and sharing

:: 01:04

your wisdom because it's it's deep. Deep having me. It was awesome having

:: 01:10

y'all having y'all. Lord, y'all having me. My goodness, my brain. It's chicken math. It's chicken math.

:: 01:16

Chicken math. That's it. That's it. It's chicken math. All right. See you later.

Share Episode

Shownotes

Transcripts

Follow

Links

Chapters

Video

More from YouTube