Systems That Stick: Tam Nguyen on Practical AI for Teams

0:05

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

0:12

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

0:18

AIdriven world and prepare for what's next.

0:29

I I gotta say your hair looks amazing. Thank you. I washed it. The people who

0:35

the people who saw me on the Daily AI show early this morning can attest. It needed a good washing.

0:42

It needed to be washed. I have to tell you um though I don't Do you know the the nice folks at the Daily AI show?

0:49

It's I certainly know of it. Yeah. Yeah. So, like Carl Yay, Brian Oair, Andy Holiday, uh Beth Lions, um

0:57

Jun Mi Hatcher. Fantastic. Here's the thing about them. They all

1:03

sit so still. Like when you know there's four four of

1:09

us and everybody is sitting still the whole time looking at the camera for 60

1:14

minutes. I I feel I feel very judged. I Yeah,

1:20

Kyle, you and I, we do. I mean, now that I've seen myself Yeah. like

1:28

I felt like I was like I was the whole time I was like, "Don't move. Don't move. Don't move. Don't

1:33

itch. Don't scratch. Don't Don't look over there." But um yeah, so I

1:39

those are not our people. media trainers. Media trainers would say that we should be more like them. But but I

1:45

would argue that we're just being us, you know, we're just being us. But when it was me and then three people sitting still like

1:52

statues, like Zen Buddha masters. Wow, that's rough. Well, yeah. So, but this morning I I sat

1:59

still. It was actually very very difficult to sit still that long.

2:04

That's awesome. That's awesome. Well, let's let's jump in. I'm excited. How are you? Uh, where's your your head

2:12

these days with with how ready are you for AI? Do you want Why don't you go first

2:18

today? Okay, I'll go first. I've got I've got a good one. Okay. Well, I don't know if it's a good one. I've got an answer.

2:26

Um, so this week I started um I created this five-day back to basics AI crash

2:33

course that I'm giving. So, so it's it's basically on the same channel that I do my lives every night, but it's a bit

2:38

more structured, right? It's a bit more sit still and, you know, do do the one thing rather than just full-on chaos.

2:45

Um, and what there there's a couple of things that

2:50

have struck me. I haven't done just a true back to basics kind of education in

2:56

probably more than a year because a lot of people on the AI uh the AI learning lab are irregulars. They're

3:02

the people that show up all the time and and you know, we don't necessarily need to go back to the beginning. And so

3:08

there's something really refreshing about just assuming that anyone that's watching may not have tried even chat

3:14

GPT may not have tried anything and really thinking about well what do you talk about how do you talk where do

3:21

you start right you know how do you talk about prompting how do you talk about the tools how do you talk about all the features within the tools how do you

3:27

answer their very legitimate questions about what do all these numbers mean for the different chat GPT models like what

3:32

is this nightmare right um so that's that's one piece of it the other piece of It is

3:39

it is essentially impossible to do it back to the basics. There's no there's

3:45

nothing basic right now in AI chat GPT itself. I did a I did I have a slide of

3:51

all the different features that I tried to go over in an hour. It was more than 25 of them. 25 distinct features like

4:00

memory and system prompts and deep research and Sora and codecs and just

4:06

just custom GPTs versus projects which are almost identical but they're study mode study mode

4:13

yeah voice mode just all of it right so so even the simple straightforward tools

4:20

are no longer simple and straightforward and so it just reminded me of a how far

4:26

we've come, but also how how intimidating um this this what we do is if you

4:34

haven't been playing along and I feel like that's going to get that that

4:41

shock or that jarring sense of oh my god there's so much to learn is just going to get worse and worse and worse for

4:48

people who are not up to speed right now. So, so the idea of AI ready is just start playing. And it also very much

4:54

reinforces for me um the idea of play first is super important

5:00

because I can imagine like I was overwhelmed in just trying to put together here's what we're going to talk

5:06

about. Yeah. I can't imagine being new and just like having all this stuff kind of, you know,

5:12

blasted out of the screen at me. Like what am I supposed to do? Just play. Just play. just have a good time with it

5:19

because eventually some stuff will start to seep in. You'll start to learn. But I I don't know. So So for me, I just it's

5:26

it's been a really a really interesting week. Tonight we're doing creativity. So we're doing

5:31

generators and video generators and song generators and things like that. So So it's it's going to be a fun night, but

5:37

still all of those tools have gotten so much more complicated and so much more sophisticated, even the simple ones,

5:43

right? Yes. Well, okay. So, can I just tell you how to do your job a little bit?

5:49

That would be awesome because you know I don't know how to

5:54

you and I both know for me with the image what maybe it was just like the bridge too far but when it

6:01

came time to learn the names of all of the image you know all the all the like

6:07

creativity related AI platforms that's when that's when my brain said just know

6:14

all the names. Yeah. So, I don't know if there's a way to like choose the ones that sound more

6:20

different like remember like Korea and like decoherea

6:26

de decompose. Yeah. Did de like it just descript like

6:31

they all are the same. So that would be really my only thing would be help people somehow or another h not have to

6:39

remember 30 brand new nonsensical words, letters, numbers, sounds.

6:45

Yep. Um I I told you I was watching the show and I loved how um I loved watching you go

6:53

back to basics. I loved seeing you Wait, did we get Did I miss Did the box metaphor come out yet?

6:59

Hey, that was night one. You missed it. Okay, I missed it. That's sad, but actually I'm okay with it.

7:05

Well, it was funny. I was not gonna do it and then the the crowd demanded it.

7:11

Yeah, I'm sure. They were like they they couldn't live without it. You've got to see the box. I got the box

7:16

out. Everyone was happy. It was good. Yeah, that was the box is good. But what I noticed in the comments that I really loved seeing

7:22

was so many people like I was there at toward the in the beginning and then toward the end so many people saying

7:28

like, "Thank you so much, sir, for doing this. This is my first time. Never seen anything like this. You're so energetic.

7:34

Appreciate you. Thank you. Like a new crowd. I don't know maybe was how

7:39

you guys promoted it or something, but uh really appreciative people. Well, it was it was really good. One of

7:46

the things that that Brandon noticed last night was um he did a he ran a poll

7:51

at the very beginning of the night and I think 80% of the people said they use chat GPT every day.

7:57

Okay. Some of the people that use chat GPT every day, I was showing things. They were like, "Wait, it can talk?

8:04

Whose voice is that? I didn't know it could talk." Like there there were things in the list of 25 features. Even

8:10

if you use chat GPT a lot, if you're not like explicitly exploring into the

8:17

corners of it, there are very likely things it does you have no idea that are even possible that are remarkable tools,

8:24

right? And so it's like that's a thing that's a thing that that kind of blows

8:29

my mind a bit is that people that are fairly deep into this AI stuff had

8:35

revelations in the in the back to basics just sort of here's some stuff to look

8:41

at even just you with you saying there are 25 features I started just like quickly

8:47

thinking of of them and I realized in a moment like huh I don't even use 03 like

8:53

I can't be bothered I'm that spoiled. I'm like, I don't know. Four is fine. You know, like

8:58

it could be. But it is. That's true. And it is. It's so good. Like four is so good.

9:05

Like like I you know, part part of this there's an ego thing. And for me at least, I'm curious what you think of it.

9:11

There's an ego thing of like you at some point you just have to go, h I don't have enough going on up here to even

9:17

need a reasoning model. Like there may be a use case or two where I pull it out, but for the most part, I'm not

9:23

asking reasoning kind of questions. Well, also like when they say when they

9:28

say like, well, if if it really needs, you know, if you if it's something that's really hard to figure out and I'm

9:34

like I feel like anytime I'm thinking it's like effortful, like how much more

9:39

effortful? I guess I don't do things that are that hard or something because I don't I can't relate to the all of a sudden like no, this is a really hard

9:45

one. Mhm. Um I think we might have I think we talked about this this weekend, but I

9:52

find it so fascinating that so many of us are like we hit these milestones in

10:00

our like AI educating community building journey and I told you independently

10:08

even though you and I talk all the time I didn't know that you were doing the back to basics. You didn't know that I was launching a foundations, you know,

10:15

for women. Like it's foundations for beginners, AI foundations for beginners. I mean, I we

10:20

could have launched that any time in the last two years. Yeah. It was the same week as you launched yours because we

10:26

just it's time. We have to go back to the beginning or how whatever the

10:32

version of the beginning. I don't know if we need to go back to the beginning being all the time, but build some

10:37

foundations with people so they can take on what's next. Yeah. and and you know I mean the the

10:42

fact of the matter is even though it feels like we're just getting started we are two and a half years into right the

10:49

post chat GPT world um it seems decently

10:54

likely that tomorrow um chat GPT5 shows up

11:00

which is exciting but again it's going to be a whole new thing to understand it's going to be a

11:06

whole new probably way to talk to it yeah Um, I think if the promises

11:15

of Mr. Altman are correct, we're not going to have to deal with models anymore. That would be lovely.

11:20

That it will just do the sort of dynamic, you know, model choosing on its own, which that that it should be. Like

11:26

I if I never see another: 04430

11:32

again, I'm good. Like we're we're done. No more. Yeah.

11:38

Just answer my question. You guys figure out the numbers. Keep

11:44

track of that stuff. I have very simple request. I'll ask,

11:50

you answer brilliantly, and we're done. And then we're done. So

11:56

is So what do we want to talk about for a couple more minutes? Do we want to talk about um study mo study mode?

12:02

Because that's kind of exciting. Do we want to talk about chat GPT5 coming out?

12:07

Do we want to talk about uh AI powered pricing and the freakout around that?

12:12

What do we want to talk about? I like the AI powered pricing. I'll the the for the one thing to pay attention

12:18

to this week, I've got I'm going to I'm going to bring up the slide from last night because I I want people to be able

12:23

to just screenshot it and we can we can just talk about that really quickly. Um the the AI the AI powered pricing. So

12:31

this was I think Delta Airlines, correct? the CEO of Delta Airlines said, "Hey, good news.

12:38

We're embracing AI and everyone's talking about AI. We love it. It helps our planes fly and we're going to now

12:46

use it to figure out who you are and we're going to charge you the most

12:51

possible you're willing to pay based on everything we know about you." Isn't awesome?

12:56

Yeah. Yeah. And he's like, "Oh, it's been so well received by Yeah. by the board." Yeah,

13:03

the shareholders. I know. And I love I love the audacity

13:08

to just be like this new pricing model is called what's the most you will pay.

13:14

Unbelievable. And you know, it's funny. It

13:21

there there's something about AI that is so seductive that I that I

13:27

think it's I think it's tempting for people to go, "Oo, this is AI." or oh, I

13:32

want to use this AI tool that or this is an AI strategy, right? Like the the more

13:38

we learn about this stuff, the more we go back to the lessons we've always learned, which is the technology is not

13:43

as important as wait, why are we doing this? What are we trying to accomplish? Right? And and I just I think there are

13:49

going to be just sort of a rolling series of gaffs, you know, for the for the coming years while people just adopt

13:57

technology to adopt technology, talk about it just because they think it's cool. Um and, you know, make use

14:05

technology for decisions like that which are, you know, a choice.

14:10

That's a that is a choice. It is a choice. But what's what's tricky about it is that it's the kind of thing

14:19

that makes it really easy for everybody to say AI is nothing but evil. Yeah,

14:26

because it really would be awful if every single thing we needed to transact

14:32

about was based on all these bajillion data points that we've so willingly

14:38

given over, you know, and you could say like, well, you know, it's kind of sort of been this

14:44

way. Yeah, you could say with the airlines, it's kind of sort of been this way. You don't know who paid what. You for sure didn't pay the same price as

14:50

the person, you know, back there or up there. Um, but when you start thinking

14:55

about things like gasoline and food and medicine and that all that and what

15:04

regulation may or may not be put in place to protect us, I think I

15:11

think it I think this is an example of when the AI looks pretty freaking bad.

15:17

Yeah. Yeah. It's a bad look. It's it's a it's a bad look. It's a bad look. It's a bad look. AI.

15:22

Yeah, it's, you know, and and listen, I mean, and and you know, I think people are justified in being frustrated with

15:29

with, you know, the the tech the corners of the tech world that, you know, are are just celebrating the tech for its

15:36

own sake and not really thinking about how it's going to impact all of us. Um, okay. So, let's let's let's talk

15:42

about a thing to pay attention to this week. So, um, I think what I'd like to do is I

15:47

just I just had it up for a second, but I want to pop this screen up. And actually, I think what I'm going to

15:53

do for a second is make it full screen. And I'm going to encourage anyone who's

16:00

watching right now, take a screenshot of this. Okay,

16:06

this is, believe it or not, last night on the back to basics um uh the the AI

16:12

crash course when we had our chat GPT 101. This was the slide. In one hour, I

16:18

I at least showed all of these things. I think I didn't show deep research, but in one hour, I showed all of these

16:24

things. Um, here's here's what I want to talk to about this. Um,

16:33

even if you use chat GPT a lot, there are likely things that either you've forgotten it does or or um didn't even

16:42

know that it did. Um, for the next week or so, I would I would encourage you to

16:47

just just kind of reconnect with playing within chat GPT. Um, you know, have you

16:56

looked at your system prompt in a while? Have you looked at your memory, right? Um, do you even know that those things

17:02

are there and that you can turn them off? Do you know what a temporary chat is? Um, have you explored non-ext

17:09

prompting? like you can upload an image that you make in midjourney and say um

17:15

give me the backstory here and it will write you an amazing story right and you don't have to do prompting right you can

17:22

just talk to chat GPT with advanced voice you can put it on your Bluetooth in your car so I I I think a thing just

17:30

because chat GPT5 might be coming out and it's going to get sophisticated a

17:35

lot of these tools may change and get integrated but just kind of refamiliarize yourself and start playing

17:41

again and start to, you know, push into the corners of what chatbt can do. I feel like that might be time well spent.

17:49

But you tell me, you look at that list. Does that just melt your brain and make you want to run away? Go go go back to a

17:55

podcast where you can't move. It makes me want to cancel the rest of the week and just hang out in chat GPT

18:02

and just like do all the things. And and I think I

18:08

think this is my defining moment of

18:14

yes, you need to know how to work with an additional frontier model. You do

18:20

just in case. But if you think you're like trying to find a better one-stop

18:25

shop, like no, just it makes images good. It makes words good. It makes

18:31

It makes words good. They got GPT5 coming out. They're the ones that like no one's really caught up

18:37

with them which is remark. I mean, they've caught up with them, but nobody's really clearly surpassed them, right?

18:42

And and yeah, until further notice, like having a real handle on what

18:48

you don't have to be good at everything, but having a handle on what chat GPT,

18:53

the ecosystem of chat GPT, let's call it that. It's not a single tool, right? It's an ecosystem. having a handle on

19:00

what the ecosystem makes possible I I can only think will be valuable for you

19:07

100% just and and to your point I think that if people could go in there and set

19:16

a timer for an hour and do not what they usually do like

19:22

you're not allowed to do the thing you usually do press the buttons that you don't know what that is like press that button

19:28

button. Dink around with it. Ask ChatGBT. How do you How do I do Sora?

19:34

How what is Codeexps? I don't even know. Yeah. So, um it's got to do with code and an X,

19:44

right? Like, you know, um and do some do some fun stuff. Yeah. I think

19:51

like I took a picture of it because I'm like I need to refresh my memory. Yeah. I couldn't believe it when I wrote

19:56

that down. I'm like I thought it was going to be just a column like a single column. I had to go two columns. I had

20:01

to categorize stuff. Yeah. Because I'm like wait some of these things are related to each other. Like custom GPTs and projects in chat GPT are

20:10

very very related. They're kind of variations of each other. Like projects is kind of like notebook LM. Custom GPTs

20:17

are kind of these things you can share with the world but you can sort of do that. It's you know but again it's just

20:23

like go play go play. It's funny because I told Michelle like Michelle Monty Silva who who is the executive director

20:30

of the She Leads AI Academy and who's the our foundation's course is her baby. Um, I said the other day I'm like, "But

20:37

we got to I got to do my thing where I go around the whole OpenAI ecosystem." I show everybody all the things. And I

20:43

just now realized no, that's not possible in a you know, foundations 90minute one-time workshop, but I can do

20:51

some cherrypicking, but it's very robust. You you you can't. In fact, I I would

20:58

argue that showing that much stuff in an hour was a

21:03

tad irresponsible. Tad irresponsible, right? Because if I were sitting there

21:08

and someone's like, "Oh, chat GPT 101. I'm going to learn how to prompt it." And I come in there and then I'm just

21:14

assaulted with the 30 things it can do. I'm like, "You know what? I'm good. Call

21:20

me. Call me when when someone figures this thing out." Right. Yep. Um, at the same time knowing that

21:28

it is a very deep well for 20 bucks a month, even the free version of Chat GPT

21:34

is so incredibly valuable if you take the time to kind of learn what it can do. And I think to your point, just go

21:41

do stuff you haven't done in it and play. Like don't put any expectations on it. Don't tell yourself you've got to

21:47

learn it. I think I think one of the worst things you can do is just say I have to learn this and then it just makes it like you know eat your

21:54

vegetables you know that's how I am. Yeah. Yeah. Just go play just go play Gen X very Gen X. It's going to love it

22:01

and it's you beta. You betcha. So um

22:07

so one thing Okay. So the people who haven't been you know obsessing about

22:14

this stuff for the past two and a half years. What is it going to be like for them when

22:21

the advent of CHGBT5 wakes them up and they're like, "Okay, this big thing, I knew something. I knew

22:28

it had a five in it. It happened today. Now what? What's it going to be like for

22:34

them? Do you think if they don't have any background and they're diving into five?

22:44

My instinct is it's not going to matter. Okay. Um

22:51

Sam, in in the in the Theo von interview, which I know drove you a little bit crazy, but but

22:58

there was there was a point in that interview where where Theo said, "Is there like what scares you? Is there

23:03

anything that scares you?" And Sam said, 'Oh, someone sent me this email and I didn't even understand what

23:09

the question was and I put it into, you know, five and it just answered it, right? And it kind of shook me. It was

23:15

like, oh, I don't even need to be in this conversation, right? And then and then he said, and then I just moved on.

23:22

And one of the things that I've noticed in the lives and in the AI salon and even, you know, you and I talking about

23:28

stuff, it's actually kind of shocking to me how quickly we normalize

23:35

this remarkable technology. So my instinct is five is going to come along. It's just going to be better at stuff,

23:41

right? Yeah. And then people that are new to it are just their first experience with chat GPT is going to be like, "Oh, this is

23:48

this just works." Like it will probably be a similar revelation as we had with Chat GPT 3.5.

23:55

Yeah. It'll be at a whole different level, but for them, they'll go, "Oh, this is amazing. Cool." They'll have their Kevin Mallister moment and then I feel like

24:01

they'll just normalize that as, oh, you can do that now, right? And then I think we go back to our lives

24:07

in some weird way. So, I don't know. I don't know. Like I I used to think I I

24:13

used to feel like there was going to be this moment where a thing launched that woke everyone up and all of a sudden

24:20

it's just we're all using it. I feel less and less like that's going to happen. I feel like it's just going

24:25

to be a normal adoption curve, right? We've still got the crossing the chasm. We've still got this big chasm of people

24:31

that are just on the other side like I don't want to try it. I don't care. Whatever. Or just not even aware at all.

24:37

Yeah. And then you got a bunch of people that are trying it that are like, oh my god, this is amazing. You got to do this. And I think it's just going to be a

24:43

relatively normal adoption phase. But for people who are not playing with it

24:48

now and come in and use it, I think that initial Kevin Mallister moment is going to be increasingly mindblowing.

24:54

Increasingly mind-blowing. And it's possible that they that it will lend us

25:00

more credibility as well versus people who went on and thought it was like some magic easy button and they would write

25:06

the like the best email of all time. Didn't like it and then they're like, "Oh god, that Kyle Shannon and Ann

25:12

Murphy. They're always try what? I don't know what's wrong with them." Husters. The Husters. What are they How

25:18

are they using this thing? You know, they obviously don't know good writing. Yeah. Exactly. Oh, yeah. Yeah. Yeah.

25:26

Yeah. Yeah. Yeah. I got that when I got when I got yelled at on LinkedIn for all the art things. I' I've never been called a a no

25:33

talent loser more in my life. And it, you know, it it just it just reinforced

25:38

all of my insecurities from childhood.

25:44

Listen. Well, listen. Hey, I have an idea. Why don't you tell the good people about

25:51

this swell little group you put together? Swell group. So well we have a little

25:57

bit of of an interesting update. We always have updates. I shouldn't say we say it like it's unusual but she leads

26:05

AI is a kind of it is an ecosystem to use the the word that is becoming the

26:10

word of the day. We it includes an AI academy which is external facing education for all for the peoples for

26:18

the for the women peoples. Um it we have a consulting agency that's coming online soon. We're putting final touches on our

26:25

website which will allow us of course as you do to put your shingle out. Yep. Then we have our Sheilleadi society

26:32

which is our membership community. And one of the fun things about the Sheiles AI society is that we have social

26:38

Saturdays. One of my friends pointed out to me it's it's like aa like it's there

26:44

be here on these screens from: 10:00

26:52

damn Saturday. So, if you are a woman in AI or you want to be like, if you're

26:57

totally obsessed or if you're just curious, like we'll be there. We'll be there. The door's open. Come on in.

27:02

We'll get you some hot coffee. Um, and you can sign up on our website,

27:07

which is sheleadsai.ai, and you can join us. And what's new is that we are doing

27:14

themes every weekend. So, this weekend is content generation and visibility. And we're not doing like the, you know,

27:22

not like the, "Hey everybody, here's how to write a prompt," but like, you know, a little, not the stuff that you see on

27:28

LinkedIn, the stuff that we But you're doing, you're leading with a use case. It sounds like we love use cases.

27:34

Yeah, use cases are the way to go, which is not what I've done this week, by the way. Oh, yeah.

27:39

I went back to my old my old habits, but that's all right. We're very dogmatic in about about you

27:45

use casebased learning. Um and then the weekend after that we're doing um AI and

27:51

hiring. Uh last weekend we did agents. So now we're kind of making sure that

27:57

people know what they're what you know they can pick and choose. Come anytime. And

28:03

um we're excited about it. And what I've what I've been super super excited about I'm almost done with my little rant

28:08

here, Kyle, but one of the things that you and I have done like with AI

28:13

Festivus, we have events at times when

28:19

some people might think, well, that's not going to work because everybody's home with their family. Everybody's

28:25

doing the holidays. Everybody's hanging out with their kids on Saturday mornings. There's no way. Well, guess

28:30

what? There's a lot of people in the AI community who don't have that same exact

28:35

Norman Rockwellesque life and who have who are looking for places to be and

28:42

things to do on weekends and in the holidays. And I like that we are there for those people in our community.

28:48

I love it. I love it. And you know, I'd al I would also argue that, you know, the the people that we're attracting are

28:54

curious enough about AI that they're prioritizing it. One of the one of the

29:00

things that drives me the most crazy in the world is when someone says, "Well, I don't have time for that." Well, no, no, no. You you chose not to make time for

29:07

that. Those are two different things. Very different things. And so, what I'm noticing is that the

29:13

people that show up to our events on a regular basis are prioritizing learning and doing the self-improvement and doing

29:19

the upskilling and things like that. So to that end, um the AI salon uh is a

29:24

e week after chat GPT back in: 2022

29:33

first meeting. Um uh you know, similar to to she leads AI. It's a group of of

29:38

really remarkable, generous, curious people that are exploring AI. Um we talk

29:44

about the cycle of AI readiness, which is play first, mindfully create, and generously lead. And generously lead

29:50

looks like learning out loud. Like people learn and they generously share what they're learning. Um and we also

29:56

have a subscriptionbased area of the uh the community called the AI salon

30:02

mastermind for people that want to kind of step up the game and get into more focused conversations. Like when you do

30:07

want to dig deep, it's nice to do it with people who are on a similar kind of level of commitment and and just really

30:13

really in it to be professional about it. And speaking of which, so please uh check out the the uh the salon and if

30:19

you go to community.thesalon.ai, that'll take you directly to the mighty networks community itself. Uh and it'll

30:26

just be one step less for you. Uh and the this uh this whole podcast here

30:33

exists because of the AI readiness training program which we put together along with Vicky Baptist out of AI

30:39

Festivus. So, if you are looking for a real deep dive in how to shift your mindset for AI

30:46

um and be ready for what's coming independent of the tools um that's what this training program is all about. So,

30:52

go check it out. It's five deep deep modules uh that use the uh content from

30:59

AI Festivus as sort of anchoring points and and then we've distilled out of all those talks some some common themes in

31:06

five different categories. So, really remarkable training. So, I'm super excited about it. Get a certificate, yo.

31:13

Yo, get Yeah. Put it on your LinkedIn. Put Put it on your LinkedIn, yo. Put it on your CV.

31:19

So, listen. We've got um an exciting guest. Why don't you tell the good people about Tam and let's bring her up

31:25

here. So, one of the things that I like about Tam is that she's also a Pacific

31:32

Northwest girly. So, she's up in Seattle, which is just um hop skipping a jump from me in Corvalis,

31:39

Oregon. Um I got to meet Tam a couple of months ago through just the extended women in

31:45

egg women in AI community. And like so many of us um out here, she's, you know,

31:53

not doing this stuff alone. None of this stuff can happen in isolation. And I'm

31:59

very interested in Tam telling us about where her next like where she's headed

32:04

right now because I'm loving the direction that I'm seeing. And what I think what I think is going to be the

32:10

case is that Tam is going to be one of the people who

32:16

is like if you've been using Chad GPT for like five hours, the next thing you

32:22

should do is start following Tam and start doing what Tam Tam tells you to.

32:27

Um, she has a gift for breaking things down in a way that is so clear and like

32:33

just not noisy. Just is this the problem you're having? Then do this. And I

32:38

really appreciate that because that's really good, right? That's not how my brain works. So Tam, awesome. So Tam, welcome, welcome,

32:45

welcome, welcome. Thank you, Kyle. Thank you, Ann. I appreciate the introduction. And yes, uh

32:50

I I I've been told that my communication style can be very direct in both good

32:56

and bad ways. So take it or leave it, right? Direct is good. So before we jump in and

33:02

start asking you a million questions, why don't you tell the good people who you are? Tell us tell us a little bit about your journey.

33:08

Yeah, my journey has definitely been nonlinear. Most probably like most people here, probably like you guys as

33:15

well. Um you know, I started out in accounting. I was in that for eight years. did property counting, real estate, both commercial, residential for

33:21

a while. And I just got bored of the financial year end clothes and the cordy

33:26

clothes and doing taxes and you know just uh to me it was it was just not

33:32

something that that aligned with what the work that I wanted to do. Um and I realized over time I just uh have always

33:39

had this this systemsoriented solutionsoriented mindset. So, wherever

33:44

it was that I went, I was always kind of finding a better way to do things. And my mantra has always been like, I'm

33:50

gonna leave this place better than I found it. Um, and so I after accounting, I

33:56

transitioned over to Fentech where I, uh, have worked in various goto market roles. I've founded teams. Uh, the most

34:03

recent corporate position was at RAMP. So, I was helping, um, accounting firms just adopt technology across the board.

34:11

And, yeah, it was great. And now uh I have become obsessed as you say with AI

34:17

and I think um I really agree with your point on play first because that is exactly how I've learned

34:25

how to do everything that I know now is just by testing and I've spent dozens

34:30

even hundreds of hours playing with AI. Yeah. Yeah. Yeah. That's great. And then and

34:37

what's your what's your current uh what's what's your current thing that you're doing that you're most excited

34:42

about? Yeah, and uh that's a really great question. Thank you for asking that. Um you know, for the last couple of years,

34:48

I have been really diving deep into not only AI, but also marrying that with the

34:54

automation aspect. So using tools like Zapier, like make.com, like natn, and

35:00

then combining that with AI. Um, I've been a little bit more of a generalist

35:06

up until recently and only recently have I decided to narrow down into content

35:12

generation specifically and going into LinkedIn first and expanding into, you

35:17

know, blogs and newsletters. Um, I think this is something that a lot of people using the automations

35:22

using automations with AI. Yep. Yeah. Yeah. Yeah. Um, just totally geeky. I'm just curious where what is

35:29

your tool of choice right now? Of of the three you mentioned, which is the one you like geeking out on the most?

35:34

Oh, I so my philosophy is around human plus AI. Uhhuh. AI doesn't work without humans. AI

35:41

doesn't work without the context that humans provide. AI doesn't work without the strategic reasoning and critical

35:47

thinking that AI that uh humans provide. And so there always have has to be a human element to the process. Always.

35:55

Yeah. Um and having said that I when I choose tools

36:00

my I like n a lot but it is really technical for a lot of people. It's

36:06

actually developed for developers. Yeah. Um so when I'm working with businesses I

36:12

assess well you know where are they on their technical journey? That's great. That's great. Who is going to adopt the system after

36:19

I'm gone? Yeah. So those are the kind of things that I think about. So between make and n I

36:24

like NAN a lot. As far as AI is concerned, I use both GPT and uh and

36:31

Claude quite a bit. So you may have heard uh Claude is really really great at writing

36:36

and I can attest to that. Oh, that's cool. That's cool. That was that's actually the perfect answer. I asked you a a crappy question which is

36:44

what tool do you like that and you that's the the right answer is depending on who I work with I have different

36:50

favorites. That's that's that's an awesome answer. All right, I will shut up now. And questions, thoughts.

36:56

All right, shut up. Um, now when you when you're doing your

37:01

content thing, you're obviously going to, you know, discover all sorts of different areas that you want to help

37:09

people doing, help people like expand and explode. Are you going to do you think you're going to end up being like

37:15

a influencer personal branding kind of coach or like a CEO personal branding

37:22

kind of coach or just like an everybody kind of educator? What are you thinking is your ICP?

37:28

Yeah, I've been thinking about this quite a bit and if you have played around with uh content posting tools,

37:34

content creation tools, SAS products, the issue that we encounter with those

37:40

tools is that number one, they're very easy to use. Two, once you buy into the system, then you're paying this monthly

37:46

subscription and you're stuck with them forever. Your data lives with them. You don't own that data. They do. Um, and

37:53

the other thing too is it's very very hard to build businessspecific

38:00

con context into the SAS into the platform. And so what you get is this

38:06

really generic output. Let's say you use a tool that creates LinkedIn posts. That

38:11

post is going to follow the same template that you know thousands of other people are using. And it doesn't

38:17

know anything about your business. It doesn't know anything about your ICP. It doesn't know anything about the services that you offer. It doesn't know what

38:23

your brand voice is. And so what I have really been focusing a lot is what kind

38:30

of context do we need to feed AI? How does that need to be structured? What

38:36

does that actually need to entail um to make AI give us the outputs that we actually want? And so my my ICP to

38:44

answer your question, long roundabout way of answering this is going to be solo founders, small teams who can't

38:50

afford to hire an an entire marketing team uh and even potentially digital digital agencies. So where I come in is

38:58

um I'm not a marketing agency. I'm not a content guru. Uh I'm also not a SAS SAS platform. I'm going to build bespoke

39:04

solutions that work for you that actually take your business context into consideration. Awesome. So, you're focused on that part

39:12

of Are you gonna Are you a context engineer? H I I kind of have a physical reaction

39:19

when I hear the word prompt engineering. I do too. How do you all feel about that? I hate it.

39:24

No, it was really just a trap to get us all to agree that we're talking a bunch of traps today.

39:30

Yeah. So, it I feel like it is so much more

39:35

important to have the right context. Yeah, me too. over knowing how to create

39:41

the right the perfect prompt the structure. I I totally agree. I the the prompt engineering thing for me

39:47

feels quite exclusionary, right? It's like, you know, it it kind of perpetuates, oh, you have to be an

39:53

engineer to be good at this. No, you don't. You can be, right? You know, um

39:59

but you but you don't have to be. So, yeah, I'm right there with you. That drives me a little bit baddy. Um,

40:04

I I have a a recent little life hack for context. Can I Can I share it with you,

40:10

Tam? Yes, absolutely. And and everybody, all the all the nice people, everyone who's listening.

40:15

Yeah, just me and you, Tam. Everybody else, just turn turn down your volume. Okay. So, I don't have like a Teams

40:24

account or whatever. I have normal, you know, chat GPT Pro or whatever plus. I can't I I don't think we should have to

40:29

be able to tell the difference between Pro and Plus. it's the one I pay for, whatever it is. Um, but I have tons of

40:36

contractors and tons of collaborators who are all working on the same general

40:42

topic. She leads AI and so, but what happens is that we're all super

40:48

well-meaning and everybody wants to promo stuff. And I'm certainly not going to begrudge anybody who wants to do

40:54

marketing for our organization. But because people aren't all feeding the

40:59

same context in or accurate context in, we get some messaging kind of drift to

41:05

say the least in including some of the stuff that we absolutely don't say.

41:10

Right. There's like a a a suite of of like ways of describing women in AI that

41:17

we we just don't use those words. And so what I finally did is I sat down. I was

41:23

like, I don't I don't believe a brand guide is the end all beall that it used to be anymore. I think that the brand

41:30

guide still still it's it's it's old school. Like you're thinking somebody's

41:37

cross-referencing your your sacred brand guide while they're working in AI and that's not happening. Not not at least

41:44

with people I work with. So we create these packets. We call them pack packets of context. So if you're working on the

41:51

this that or the other event, here's the sheile leads AI packet plus this event packet and then they can upload it into

41:59

their custom GBTs or their projects because I don't want to just give people

42:04

my stuff anymore because everyone knows how to make a better mousetrap for them, for their brain, for the way they work.

42:11

And this is a huge shift for somebody like me because I don't like Well, you should see Kyle's Kyle's This is the

42:18

kind of Gen Xer he is. Tam Kyle saves his files without proper naming

42:23

conventions on his desktop. I do on title one version 2.5.

42:30

Yeah. Final final final. Yeah. 3.4. Yeah. 3.4_draft_draft.

42:38

No, but you make a really great point there and you know it's it's I mean it's kind of like because I used to work

42:44

pretty heavy in oper heavily in operations as well. It's like creating SOPs or standard operating procedures

42:50

that is they're incredibly detailed. They've got the looms. They've got the step by steps that no one ever looks at

42:56

ever again. Right. It's this false. It's a padlock on a paper bag.

43:02

Well, right. You know, I'm I'm curious as to your thought, Tim. I feel like we're potentially entering an era of

43:09

co-creation. One of the things we did at the AI salon is we created um a midjourney mood board, a style uh for

43:16

the AI salon and we give that out to people and we're like, "Hey, if you want to ever make images that are like on

43:22

brand for the AI salon, just use use our thing." So, it's it's like it's kind of like a brand guideline, but it's more

43:27

like a brand like here's a brand goodie, right? And and you can go create within

43:32

that. I mean, do you think there's something to do you think we're entering a more dynamic era of brand expression,

43:39

brand engagement, customer engagement? Yeah, I mean, again, context is everything. And I think that uh one

43:45

thing too is when you feed these documents to AI, it's not a static document. It's a breathing

43:52

thing that needs to up be updated constantly as your business direction changes. And I think that, you know,

43:58

myself included, sometimes I can forget to do that. It's painful to update up update these things, but your AI is only

44:03

going to be that much more relevant if it has the most relevant information about your business. If it's Yeah. If it's got it. I've got

44:08

before we jump onto our our our three questions, our our our lightning round for for our speakers. Um you mentioned

44:17

you mentioned that you're narrowing your focus to to content generation and LinkedIn posting and things like that.

44:23

I'm curious what led to that. Is it is it that you find your customers are more

44:29

ready for that? Like like what is the thing that that triggered you to narrow the focus? Was it a was it an arbitrary

44:36

choice? Are you like I've got to narrow the focus or was it more based on what you were seeing in the marketplace? Well, first of all, it's a problem that

44:43

I had to figure out for myself. So building yeah,

44:48

so we should probably talk. But I also see, you know, you see those those LinkedIn content gurus with hundreds of

44:55

thousands of followers and they give you all these posts that, you know, they they seem like they're people are

45:00

engaging with them. They got likes, but we have no idea if they're getting customers, right? We don't know. And

45:05

they're not really providing tactical strategies or implementation strategies.

45:10

You can make money with CGBT, right? Or here's a hook or here's, you know, um, how do you actually apply

45:16

that? Yeah. And you know, I think of content as a a business acquisition channel. And

45:22

so, it's something that deserves the respect. It deserves the time, the investment.

45:27

Every post that you put out there, who are you reaching? What is the what is it you're trying to teach? What is the goal

45:33

of the post? Are you driving people to your newsletter? I think it's a really good idea to get people off LinkedIn and

45:40

following you on a newsletter because you own that audience, right? If LinkedIn changes the algorithm

45:45

Yeah. or you know something blows up, all your all your followers are gone. Yeah. Amazing.

45:50

So, it's uh it's really about starting with content and moving into revenue generating systems. Everything from

45:56

generating lead magnets that get email addresses, getting people to follow your content, getting people to buy your products.

46:02

That's really what I'm focusing on. And um less about content guru strategies and more about taking those strategies

46:08

and applying them to AI and automation so that we can do it at scale. Yeah, I love it. It's just so it's so clear and

46:15

it's not I'm selling AI automations, right? It's like you want some revenue for your business, I'm gonna hook you

46:20

up. Um I'm going to remind you what Tam said, which is that social media is a

46:27

business acquisition tool. Yeah. Which Tam did not say entirely clearly,

46:34

but we do need to tell people that we would like to work with them. Member

46:39

member Kyle Kyle forgets to tell people that a lot. Tam,

46:45

this is a this is an intervention moment right now. Exactly. Yeah. Here's my product. Here's my service. What are you trying to sell?

46:51

Put it together. Never does. I'm bad at that. All right.

46:56

Perfect timing. Let's move. Let's change the subject.

47:02

Okay. Um, so so we asked these three questions of every guest. So, I'll I'll

47:07

hit you with the first one. And the first question is, what was the tipping point where you knew you had to go all

47:13

in on AI and and what happened next? I mean, it's hard to ignore. It's all

47:18

over the place. It's so easy when you see all this news and you get overwhelmed and you think that you're

47:24

behind and it more and more the longer you wait that is the case. But it's

47:29

actually surprising how many people still don't know the basics of GPT. And so I think that finding a network of

47:35

people who are also in the same stage in journey is important and learning together but um you know hiding our our

47:42

head in the sand is is not an option anymore. And so I have always been uh that of a learner's mindset and I'm just

47:48

obsessed with technology anyway. So it made it really easy for me. I'm just I'm a nerd. Okay.

47:54

was there was there a specific moment did you have a specific aha moment where you're like oh my god I didn't know I could do that or like what what was

48:00

there a moment or was it just sort of a long progression for you where you're just like I'm going to geek out from day one

48:05

I hate creating proposals ah and because you can either spend a ton

48:12

of time hyper you know tailoring this thing and making that customer feel great and then

48:18

they just ghost you. How does that feel? or that looks so bad. You create something general that you

48:24

know could speak to anybody that anybody could have created. And so again, it just started with focusing on a problem

48:30

that you already have and trying to solve it. Yeah. And and so it did it really good and you're like, "Ah, this is this is

48:36

better than doing it the other way." Yeah. Yeah. Absolutely. Yeah. We're we're not going we're not going back to doing it that way again.

48:43

No. I can't Can you imagine them taking away AI at this point? I like I can't I don't

48:48

know what I would do over my dead body. Exactly. Yeah. So, oh boy, that so yeah, I

48:58

woke up this morning and was reading an article by this faculty guy who's saying

49:03

that, you know, AI is going to be the end of higher ed. And and by the way, as

49:10

that as though that is like literally the worst thing that you could ever imagine, spoken like a faculty member of a small

49:16

liberal arts college. Um it's not going away.

49:22

So okay, sorry. Um Tam, based on your area of expertise, which I know is you

49:29

have many areas of expertise, but so maybe choose one. What's an what's an AI trend that you're paying attention to

49:35

right now and why? You uh you prepped me for this one, but

49:40

I'm still not quite ready for it. I think uh for me it's so hard to get it's

49:46

so easy to get like overwhelmed like I mentioned before and so focusing on one specific area that really solves a

49:52

problem a painoint that you're experiencing so deeply right now is what I'm focusing on myself which is

49:58

why I've created this whole business around content the content kitchen um and so for me it's you know what kind

50:04

of context do I need how do I train AI how do I feed it the right templates frameworks

50:10

uh and then how do I automate all that you How do I put it into a database and give it access give AI access to these

50:16

tools outside of GPT outside of quad and how does that all interact? Can I trust

50:22

it? AI likes to hallucinate and it also hallucinates very confidently. So I'm

50:27

very interested to see how GBT5 Yeah. you know shows up. Yeah. Yeah. Exactly. My my nickname for

50:34

chat GBT is mansplaining as a service. I love it. It will confidently answer you even if

50:40

it's wrong. Yeah. So, I hope that answers that question. I don't I don't know if that answer that

50:45

question. It does. Yeah. Yeah. Yeah. Yeah. Absolutely. How do you How do you How do you I mean, you you said you you're like geeking

50:52

out. You like geeking out on this stuff, but what are your strategies for how you pay attention to what's happening? Like

50:58

like where where do you where do you pull the signal from the noise out of because there's so much noise right now.

51:04

There's a lot of noise. Um, I do use Perplexity for research and it does pull

51:10

up the relevant YouTube videos which I honestly don't have time to watch sometimes and so just summarize them.

51:16

Right. So, Google notebook LM has been wonderful. You know, there's there's I I've really liked uh Substack recently

51:23

and my newsletter is also hosted on Substack, but I like it because people don't tend

51:28

to withhold all the good stuff. Yeah, they'll give you everything. And so you can actually um even build an automation

51:36

around, you know, substack.com domain names that you receive, these emails that you receive and just create one big

51:44

Google document that you feed into notebook LM and then you can turn that into a

51:49

listening experience. Now video, now they have video and it's really good. It's really cute.

51:56

Yeah, it's really cute. Little little infographics. Instant infographics. Yep. Yeah, it's really good. No, that's

52:02

a good one. Like, yeah, sub substack's an interesting one because it it really is each of the the platforms have

52:09

different personalities, right? And that that one has the personality of here it is. Go figure, you know, take it. Go.

52:16

Yeah, that's awesome. Brilliant. Brilliant. Um, okay, last question. So, this this is one that is

52:23

um is a favorite of ours because it's because it's named after the show. Um,

52:29

what is AI readiness to you? And then what would you say to someone that's just getting started with all this AI

52:35

stuff? Uh, I mean, I've been harping on this already, but you know, start with a

52:41

problem that you already have and try to apply AI to that. Uh, context is so much

52:47

more important than prompting. You can have a mediocre prompt as long as you feed it the right information. And um I

52:53

think of it like training a uh like an intern.

52:59

And you may have heard this before, this analogy before, but what does an intern need to know in order to be successful,

53:04

right? Do they need to know about your business services, how you write, you know, who they're writing to? So, uh I

53:11

think that's something that's helped me allow out a lot. um something that a lot

53:16

of uh GPT users that I'm surprised really surprised about um who aren't taking advantage of the features. So you

53:22

mentioned all these different features that are now available for plus users. Yeah. If there's one feature that I would

53:29

recommend any paid user to try, it is uh the GPT projects.

53:35

Yes, it's my favorite favorite thing ever because you can more than custom GPTs.

53:41

Yeah. a ton of a ton of documents, a ton of context, feed it everything that it needs to know about your business and

53:46

then every chat that you create inside of that folder has all of that context. There's no more

53:52

reprompting, which can be so annoying, which can Yes. Yeah. It's so good.

53:58

Yeah, it's so good. If if nothing else for me, it's like I I can find the other chats like I I knew

54:05

there were some other chats in there. Um Yeah. And it remembers it remembers things that we talk about because other

54:11

because previously you would start a chat with chat GBT and if you've chatted with it long enough it'll just break.

54:16

It'll stop talking to you. Yeah. Exactly. Exactly.

54:22

Yeah. Is that how how is how are you dealing with the context

54:28

window challenge? Like are are there tools like is Claude better at it than OpenAI with you know when and when

54:36

you're putting together these automations? I assume if you're doing it in little chunks, you can probably get around it. But like where where do you

54:41

run into in in your professional applications? Where where's it still janky, right? Where's it still not quite

54:48

ready for prime time? Uh all of it. I would say there's really I was hoping there would be an answer

54:55

like this part's really good. No. Yeah. Well, in in the context of con uh content creation, uh the fact that you

55:02

can upload these massively long context contextual documents into a folder in

55:08

GPT or in cloud is is massive. I mean, these these files sometimes I have files that are 20 pages long or more,

55:15

but you can't feed that into a make.com or Zapier automation, right? It would be so costly to do that.

55:22

Yeah. So, I think it's just picking and choosing, you know, what do you really need and what parts can you and should

55:28

you automate? Yeah. And that that gets that gets Oh, sorry, Ann. Go ahead. No, no. Well, I was going to say I know

55:34

this sounds so basic, but just in case anybody didn't know this this weekend, I was putting a bunch of stuff into Oh,

55:41

oh, I was putting six months of transcripts from social Saturdays. So,

55:46

they're super long. It's a two-hour conversation. Sometimes people stay after so it's three hours or four hours

55:52

and I wanted to get them all into one project but I ran into the limit of you can only have 20 files.

55:58

I just but can you I just said can you put all of the text from all of these files into one file one file and it did

56:06

it for me and then I just put that back into the project and that was that. Yeah. Felt like cheating. Loved

56:12

it. Another feature. Yeah. Tell me. that just released recently in last week or or so I want to

56:18

say is the connectors. Yeah. Oh yeah. Have you utilized that?

56:23

I'm afraid I'm afraid I'm going to attach it to like my credit card file or

56:29

something. I don't I don't know how to use them. I hooked it up to maybe you can help me. I hooked it up to notion

56:35

to my instance of notion and then I said, "Hey, do this thing and then write it to my notion thing." And it said I

56:41

can't write to notion. So I just gave up. Right. GBT can't do any edits yet.

56:48

Can, but you still want to be careful because it could wipe out your entire notion. See, I don't I don't That's why I wanted

56:55

to do it is because I don't have any data there. So, I just wanted to play and then the the first thing I tried it

57:01

couldn't do. So, I was just like, I'm gonna take my ball and go home. I just Which we try to get other people not to

57:07

do, but I do. Exactly. Exactly. Tam, tell everybody how can

57:14

they work with you? Where can they find you? Tell them how awesome your newsletter is and your website and the

57:20

lead magnets. Um, we want people to be able to find you. Our people need you. Yeah, definitely. Um, reach out to me at

57:28

hellwithtam.com. You can visit my website techwithtam.com. Um, I started a

57:33

publication on Substack. It's called the contentkitchen.co co and I talk I'm going to be sharing a

57:40

lot more strategies around applying uh content creation strategies to AI and

57:45

automation and how you can scale your own business acquisition channel and learn how to sell yourself a little bit

57:52

better. So um looking forward to sharing more there. Um but yeah, love it. Love it. That's awesome. That's awesome. And how

57:58

many how many businesses have you started? Uh I don't know how many how many domain

58:03

names do I have?

58:10

I registered more domain names in the past year and a half. Yeah, it's a problem. Yeah, they're all

58:16

gone. All of the anything shorter than five letters. Um, if you want theai,

58:21

they're gone. Yeah, you have to do crazy You have to do crazy things like, "Are you ready for

58:26

AI.com?" with some combination of letters, numbers, and asterisks.

58:33

Tam, this was fun. I'm so excited that you're out here. I'm excited about what you're doing. People

58:39

need to need to work with you. Lots of them. I love this. Thank you so much for having me on. It truly is a privilege to

58:45

be here. Uh yeah, this is the most fun that I've had in a little while. So Oh, that's good. That's awesome. Yeah.

58:50

No, it's you know, hang around with us. Yeah. I I love your insights. I love your approach, too. You're just you're

58:56

just so clearheaded about this and and not seduced by the technology, which is Yeah. It's the perfect place to be. And

59:02

I I think what you're up to is just incredible. So, um I I think I need to hire you. So,

59:08

you do. Well, you know where to find me. Yeah, exactly. All right. Thanks again.

59:15

See you later. Bye. Bye.