From Gen Z Skepticism to Local AI Models: Jyunmi Hatcher on What's Next

 Forget trying to keep up with ai. It's moving too fast. It's time to think differently about it. Welcome to the AI Readiness Project posted by Kyle Shaman and Anne Murphy. They're here to help you build the mindset to thrive in an AI driven world, and prepare for what's next.

Ann Murphy. Hey, re shaking. How you doing? Hey, hey, hey, hey, hey. I'm good. I'm good. I'm excited. I've got my new microphone. Can you tell I'm a professional now?

We can tell. Yes. I'm just now getting used to it. In fact, I had like a different version of your podcast voice in my head, but it's, it's very professed.

Is it? Okay, good. Yeah. Well,

yeah, professional.

Yeah, we're

very professional. Everybody

we'll dial it in. Yeah. You know how professional we are. Listen what we lack in professionality, we we make up for in get it doneness we make.

Yeah, exactly.

Um, it's funny because I bet a lot of people might not know that.

Like we are very capable of sitting up straight and wearing grownup clothes and having conversations that are, you know, like more on the, I think you just word used the word profession. Profess professional.

They're sort of profession. Professional. Professional. We're

professional.

Yeah. I'm professionalizing.

Yeah, we can, we can do that buddy. We, we choose not to, but we can. No, it's, it's important. You gotta be able to turn that on and off. But you know, listen, everything's changing. So, so is professionalism.

So is professionalism. And by the speaking of professionals. By the time this comes out, we may be at like episode 30.

So we are, we hit episode

25. We're, we're in the, we're, we're well into double digits, which, uh, congratulations. We're to double

digits.

You, you know, one of the lessons in life is, you know, one, start it and two, keep showing up. Start it.

Just

keep

showing up. Okay. Can we just, I know that, okay, we're gonna talk about ai.

I feel like I've got a little bit of a topic about how ready are we for ai, but I did wanna say about this thing about just showing up mm-hmm. How many conversations this is coming up in. Like really I'm reading about it, like it's in the New York Times. I was reading a blog about how showing up is an act of rebellion now.

Like, think about that we're supposed to be quiet and stay at home. That was better for, that's better for the powers that be if he, we would all shut our faces and stay home and stay separate. And, but

here we are. Here we are. And listen, showing up. We're, we're building community, we're showing up and we're telling people, come with us.

Showing up if you show up, dang. If you show up, thi good things, good things happen.

I like it. I like it. Um, how ready for AI are we this week? Um, you wanna go first? You have a, you have a burning how ready you are or you want me to jump? Well,

I have it, it's not, it may not be burning, but I think it's timely because, so I'm gonna go for it.

I like it

this, this time of year we've got lots of parents and guardians who are sending their kids off to college and either they are, you know, it's their first time experiencing a kid going to college or like in my case this is, this is the year that we be become empty nesters and how. How our kids are ready for AI is a bigger subject for most of us, even than how we are in our readiness for ai.

Mm-hmm. And so I thought I would share a couple of examples from Gen Z because both of my kids, both of my kids are Gen Z and they both have very different approaches to ai. Um. Oh, interesting.

And how old are they again? I should know. Uh, so

my daughter's 20 and my son is 17.

Okay.

So they initially, both of them are like, whatever you do, we are ride or die.

Like, we love you, you're amazing. You go do AI stuff, mom. But secretly in the background they were like, but AI is horrible. AI is the worst thing that has ever been invented. AI is gonna ruin our lives, et cetera. And we've kind of gone through different waves, so. My daughter, definitely, her name is Luca.

Luca, definitely has like a very pragmatic approach, which is, it kind of doesn't matter how I feel about it, it's happening. Right. That's, that's, that's healthy.

That's healthy, right.

Right? Yep. So she has learned all the things, and now she's a member. If she leads ai and she's going to be in charge of all the volunteers on site at the Create conference because Wow.

She's a camp counselor.

She gonna have her apply her, her well-earned skills with, with a bunch of adults that need corralling because they're all neuro spicy.

Yeah, exactly. Exactly, exactly. She'll be like, you over there, you go play hopscotch, you're useless. You over there.

Here's your juice box. Go sit in the corner. Here's

your juice box.

Sit down, shut up. So, um, but you know her, at her school, they say that they're a copilot school. Of course, the kids don't know what copilot is like. She's in on all these meetings and stuff, and she hears the, the, the leadership talking about ai and it's very different from her own experience. But anyway, so that's her.

My son, Theo, he's, he is going into his first year. He was, he's, he does not, he's very old school in terms of like, um, being real and authentic is very important and that's the only thing that matters. And being in the moment is important and what happened before it, none of that matters to him. It is like right now, that's what matters.

Yeah. And things that are fabricated are much less meaningful in his book. However, he's now going to run our, our, our, what is becoming a vast social media landscape for, she leads ai. And so I think this is a good example of you can hate it, but you may end up working for it.

Yeah, that's, that's a good point.

That's a good point. It's, I think. I guess they're not children at this point. I was just gonna say, aren't there child labor laws? You're just getting free labor as far as I can tell.

Hey, what? I don't know what you're talking about, Kyle.

He's my employee. So you, lemme get this straight. You got two kids that are anti AI working for the AI conference.

That's solid. I like it. No, it's good. No, so, so listen, I think that, um, I think that our educational institutions have done a massive disservice to our children by spending two, maybe three years saying, you know, AI's evil and you should never use it. And all AI is is cheating because it's not. Yeah. And it, you know, it's even funny.

I think if I think about Theo and, you know, he is, he's authentic. He's about in the moment. I think one of the perceptions of AI is that, you know, it can't be creative if it's just push a button and squirt out some crap. It can't be creative. But, you know, we, we just talked with, um, with Sid Harrow and, and you know, I asked Sid, or no, you asked Sid like, what if you were gonna start some new project with ai?

Like, what's your process? Like? What's your, where would you start? And the first thing he said was, the first thing I would do is I would sit down not at a computer. It's just like a piece of paper and just gather my thoughts.

Yeah.

And the people that I know that are using AI in interesting ways, they, they, they have what I, what I call chain of craft, but they have a very thoughtful and considered interaction with ai, whether they're doing it on their own or they're doing it with other people.

Um, it can be a very thoughtful collaborative, you know, creative expression kind of thing. And, uh, and, and that to me is, is uh, is something that is not on the surface of ai because it's so easy to demonize it, it's so easy to go, oh, it's just the lazy button. 'cause it can be absolutely the lazy button and it can just make crap and mm-hmm.

You can use it in a much more thoughtful and creative way.

Yep,

yep. You know? Yep. Has, has Theo played with AI at all?

Yes, he has. And and the way I, the way I got it him into it was I was, um, making images on Midjourney of him and his girlfriend. And they loved that. That was really, really fun. And he has used it for.

For school and for some story writing. And you know, he's, he is, he, he knows how to code and stuff. So he's actually not that he, like he's, he thinks that coding by yourself is cool. Like you don't need to vibe code, just code it yourself kind of a thing. Yeah, yeah, yeah. Um, so yeah, I mean, he gets the util, he gets the, he gets the, that it's useful.

What he doesn't, well the kids are very, very sensitive to the environmental impact for one thing. And then they are very unsupportive of having like, AI companions, both of them. And I've gone through all the details with them. I mean, I've talked to them about good, bad, and the ugly and they are very, very against having an AI companion.

That's fascinating. That's fascinating. Again, because I, I have an instinct that a lot of people's first experience with AI is going to be some sort of, you know, imagine a Siri or an Alexa that it, I is actually good and you can just talk to it about anything. Right? Yeah. So, so at some point someone's gonna do some version of that.

It could be Apple, who knows, but, but, uh, at this point doesn't look likely. But, but, um, but, but, but I think, you know, conversational agents, I mean, that might be a lot of people's first interactions with ai. Yeah. So that's actually a fascinating, fascinating thing that they're, that they're so avoiding it.

And I, you know, I wonder, I wonder if, if getting good with voice interfaces with AI is going to become a marketable skill moving forward, it might be. Right? 'cause because how we type things and how we talk are very different interactions. Yeah. And if voice becomes a primary interface for, for these systems, especially as they get more agentic,

yeah.

There's gonna

be people that are gonna be really good at talking things into existence, right? Mm-hmm. And, and, yeah. Uh, I wonder if that's a skill. I wonder if that's something to practice. I mean, and I wonder if that's another way in for, you know, for them is not to think of it as a companion, but to think of it as this is, this is basically just another interface to, to use these tools that might help themselves.

Yes.

I think that's, I think that's totally an approach that I could take with them. Uh, this isn't about whatever, this is just about learning how to work in the future. Yeah. Um, the. Thing about the voice interaction. There was something I was gonna say about that. Oh, okay. So that the voice interaction, they, they won't realize that they have, they're developing a relationship with an AI companion.

That's the other piece about it. Like, yeah, it's, you can use the whatever terminology you want. You could just call it a relationship. You're in the beginning of a relationship and they wouldn't buy that either. They don't think that we're in a relationship with AI when we're using it, whereas I absolutely think I'm in a relationship with AI when I'm using it so different.

It's different point of view.

Uh, yeah, I don't know. I don't know. I mean, I, I think in the end it's, it's, it's, they're gonna be largely academic distinctions. But I think some people, some people are gonna get emotionally involved. Some people will get intellectually involved. Some people will go, never, ever,

never, ever.

That's all fine. And if it, if it, if voice does become a prim primary interface, I could see that being, you know, a requirement on a, on a job application at some point in the future.

Can I tell you that, and this may seem off topic, but it's not people I promise. I was in a, um, I've tried this like three or four times when I've, a lot of the projects that I end up working on are fundraising for facilities, for spaces, and I'm working on this one for an art museum.

rdware probably by the end of: 2026

Of course, I don't say it in an, in a snotty way like that, but this design that you're building is like, you're gonna have to build, rebuild this whole interior space. And I'm, I know how much money we're raising, so I know what the budget is and I know that we won't be able to afford it, right? So I'm like, yo, let's just, whatever.

I mean, I have not been at all successful in any conversation trying to convince people who are building things that we are gonna be talking to our hardware. They really don't believe me.

Yeah, I know. So we're

continuing to build open floor plans

in today's day and age. You should show 'em the, the opening scene to the movie her where, you know, they, they all work in an office where their primary interface is talking to the thing and they're all distant from each other.

They're all in these kind of sound boothy kind of things. Oh, that's what I'm

do, I'm gonna send that to, I'm gonna send that, I'll send it to the person. She can decide if she wants to send it to her architect or not.

Yes. Say send it to her. Say you know her, and then where it says science fiction movie, cross out science fiction and just say documentary.

Exactly right.

Here's what's coming. Um, here's what's coming. So, so here's the thing, kind, kind of related to what we're talking about with your kids to, to pay attention to this week. And I, you know, I recently had this, this personal realization, I'm trying to figure out my role in AI and like, what, what's my real passion, right?

Mm-hmm. I'm passionate. Like, what I know is that it's not going away. And what I'm passionate about is getting as many people as possible to try it, you know, so that they can experience, you know, what it is and, and, and why it's good. Um, but what's, what's the thing I want them to do? And so the recently I've had, I I, I had this epiphany that the through line throughout my career has always been the thing that I seem to be most passionate about is advocating new technologies for the, the primary use of personal self-expression and.

The people that I see getting the most joy out of AI and having, you know, big epiphanies coming out of retirement, getting super inspired are when they realize, oh, I had this story and I wanted to tell it. And one of the exercises I gave recently was, go find some old piece of work that, you know, you never finished as a kid.

And, and you know, like explore it with AI and see if you can turn it into something. So I've kind of a follow on exercise to that one. And that's this, I think one of the valuable things about AI that you, you, if, if you're involved in a group like She leads AI or AI salon where we're constantly encouraging people to play outside of the lines, you'll probably do this anyway, but a lot of people, you know think, okay, here's the work I do and I'm gonna make a thing that's, that's in that work.

A thing that I would encourage people to do is take some piece of work that you've created that's in line with like what you do today, right? So if you're a writer, it's some piece of writing. If you're an illustrator, it's some illustrations. And take some nice completed work that you have and reimagine it in some other form that you're not good at.

Oh, so if you are horrible at music, you can't keep a beat. Take the annual report that you just, you know, did and see if you can reimagine it as a song in Suno. Or take the children's book that you wrote, you know, because you're an illustrator and you're a writer, and turn that into a novel Right. Or turn that into a movie.

Yes. Um,

that, that force or, or, or, you know, one of the things where people often get really hung up is, I can't code, I'll never be a coder. I don't have a head for it. I'm not good at math. Right. And, and they avoid it like the plague. They've, they've got themselves defined as, I can't do that. Learn one of the vibe coding tools, something like lovable and head over to lovable and.

Just turn your annual report into an interactive dashboard using code. Right. And it'll blow your mind. And so, so I think if you haven't done it yet, really purposefully take something that you've built and try to reimagine it in a thing that you know you're not good at. Mm-hmm. And that may unlock something for you.

So that's, that's my thing to pay attention to. And I know you've done a fair amount of that over the, the months and years. Well, you,

you gave me, you gave me the idea of co like an hour ago. And between our shows, I went and I, Dr. I dragged out. I have all these, um, 10 minute. SMI snippets of writing. 'cause I was in a writing group for a few years in San Francisco and our assignment was, we had a partner and every day mm-hmm.

We had to write 10 minutes to each other. Yep. And then because it was back then, we printed them out and we brought them the class. So I have these printouts and yeah. Thank God you have the printouts.

Right?

Yeah. Thank God. And it's the only paper I have in my office and I've always wondered why. Haven't I digitized this or why do I still have them?

They're from literally 22 years ago. Well, now I'm gonna take these 10 minute writes, which I skim through 'em and some of 'em are really good. Yeah. Um, of course then I thought they sucked. But that's, you know, being Gen X. Yeah. So now I'm gonna make them into like, well I'm gonna go back and I'm gonna watch the video of, um, of the LOL with Cindy Kon and Suzanne Wilker Jergens teaching us how to make.

Uh, products with ai

Yeah.

Merge. So I'm thinking about something like a triptych, like take one of my, one of my essays that has like three parts and make, make some images and do it in my office. Like, you know, I wanted to take these

Yeah. Like, completely reimagine it. Like, like Yeah. Like, then you couldn't even have imagined that that would be possible.

Yeah. That, that's exactly it. Yeah. And like, it's so cool that they're printed out now, and historically you'd had to get OCR software and you'd have to buy a scanner. Now you can literally just take a snapshot picture with your phone and chatt, BT can, can read it All right? Yeah. And you can say, help, help me refine this idea, right?

Yeah. Turn it into a piece. Turn it into posters. Right?

Yeah.

Why

not? I mean, how cool.

How cool. Yeah. And, and you know what, you know, back to the self-expression thing, the, one of the things that AI. What that excites me the most about AI is the speed with which I can get an idea out of my head and in front of me.

And the minute I can, I can look at my idea objectively. It, it inspires other ideas. And so like I find working with AI is this amplification cycle. So if you take these old ideas that maybe we've even forgotten about and you sort of bring them front and center again and and kind of reimagine them in this other way, it is going to absolutely inspire new ideas because you now have all of your life experience as context, right?

Yeah. So you're gonna take that seed of an idea, whatever you were thinking back then, you're gonna look at your life experience and you're gonna go, oh, imagine what I could do with that idea now who I am today, right? Yeah.

Yeah. Right. Yeah. That's so cool. That's the magic

of this stuff, isn't

it? That is, and and you know, I mostly give myself.

Allowance to do like nerdy business things with ai. And I feel guilty when I do fun stuff, but I, but I also know that it's like medicine. Mm-hmm. To do the creative part as well is so good for my brain. Right. It's calming and it's soothing. And before somebody told me about the water u the energy usage involved with like a mid journey image, I was, every single night I was making images before I went to sleep.

And that was really, really fun. I realized why that for a pastime might not be. Like I should, I didn't wanna be making a ton, a ton, a ton of images every single day. I decided that wasn't a good use of energy. But, um, right now, as I'm thinking about this opportunity to take my essays and make them into something, I'm like, yeah, that's a really, really good use of AI for me.

Really healthy, you know? Yeah.

And the, the other thing I would, I would argue, Anne, is that, you know, not only is it a little way to maybe recharge your batteries if it's, you know, you're not just, you know, making the donuts all the time.

Mm-hmm.

There are gonna be things that you discover in those things that feel like play, that feel like hobby, that you might realize, oh, wait a minute.

I've got this project I'm working on on right now for work, I could actually bring that in into

that. Mm-hmm.

Right. And so, so it's, uh, it, I don't consider any play in AI frivolous right now because so much is unknown about what's possible. Mm-hmm. And so even, even the thing like we were just talking about, if you context the voice thing as an AI girlfriend, then people are not gonna use it.

But if you say, Hey, wait a minute, what if this is the interface of the future, well then maybe I should at least learn, learn what it's about, right?

Yeah. Yep. Yep. Absolutely.

Beauty. So

what does that look all right?

That leaves us that, that leaves. That's it. So go play, go, go reimagine something in a new form.

Go play. And especially imagine it in a new form, you know, you're not good at, I think there's some magic there.

Yeah. Yeah. Um, so

here's where we are. You, you little Missy. You need to tell the good people about She leads ai.

Well, um, so I, on our last show, I got to talk about our reading club, our book club, which is called She Leads Reads.

But today I'll talk about what we did or that right now I'll talk about what we did at Social Saturday yesterday, which I wasn't there, which is the perfect reason why I came up with what we, what we're going to do. So we set up our, um, our social Saturday to talk about, uh, chatt T five. Mm-hmm. But we kind of banned like no prevaricating, no hand ringing.

We're not going through your trauma of chatt BT four versus chatt TBT five. That's not the conversation we're having today. The conversation that we're having today is how are you using it? Like, what's different? What's working? Well, give us your u use case, your workflow, the thing you tried that didn't work, whatever, and then.

Later on today, I'll take the recording of that and I'm gonna turn it into a field guide that that goes out to everybody who was there. 'cause we're like, we don't have enough hours in the day for all of us to go individually and play with all the use cases. That's great. We need to learn more quickly. So mm-hmm.

That's the kind of thing we do. And what we really value is that rapid transfer of knowledge that can take place in a safe environment. So she leads ai. Our mission is to unite accomplished women to advance AI for global prosperity. And we do that through events and through our community and through our, um, AI academy.

e're there every weekend from: 10:00

Beautiful. Love it. Love it. And the AI salon, if you don't know, is a, a, a, uh, a community of AI optimists and creatives and makers and builders and entrepreneurs.

Um, and it's, it's a really remarkable and generous organization. And, you know, I, the thing that most inspires me about the AI salon is all of the different points of view of people that, that regularly, someone will come in, they'll not be good with ai, so they apologize for that. But then they'll say, oh, I figured out this little thing that I do for my little business over here.

And like everyone in the room is like, oh my God, I never thought about it like that. The different perspectives, the different lenses through which people use AI and CAI is is absolutely inspiring. So, so that's, that's what the AI salon's all about. We have meetings, we've got a, uh, a subscription area called the AI Salon Mastermind for people that wanna dig deeper and collaborate deeper and, and teach each other in a, in a different way.

And, um, you know, start businesses together, things like that. We've got a lot of, lot of experiences of, of people meeting and starting businesses together. That's how we met. So, um, so that's what the salon is all about. And, uh, yeah. And it's, it's a really remarkable organization. And then the final commercial, uh, before we bring up Jun me is, uh, the AI readiness Training program.

, Hanukkah and New Year's. In: 2024

Um, and it was an absolutely remarkable event. And one of the things that came up during that event, it was all just people talking about AI and how they think about it and how they're doing it and what they're building. And people wanted. A course, they, they said, what are the lessons here that we can, we can pull out of this?

So we worked, it took us, it took us longer than we thought it would, but we, we got it together. And working with Vicki Baptiste Anne and I took the lessons out of those, um, 24 sessions and sort of pulled out evergreen content and turned it into the AI readiness training program. So I'll let you take it from there.

So the AI Readiness Training program is a self-paced course, so it's one of those numbers where you get to do things in the order you want to. Whenever you feel like it, and I dare say that it is among the self-paced courses you are most likely to finish. I know it's really, really hard. I know people like the statistics are awful about how few people actually finish a course.

The nice thing about this is that if you don't want to take the section on AI and business or AI and creativity, you could skip a whole entire section and still graduate and get your certificate. However, we encourage you the, it's the what you resist. You must persist the least. The thing you, the section you most don't wanna do, do that one first.

Yeah. Like, you may think that it's not useful for you to learn about AI and creativity, and I would've thought I would've said the exact same thing a few years ago, but wow. Has it opened up my world. So we, so the lessons are, um, sections of the talks from AI Festivus with each of our 34 speakers and then Vicki Baptist pulled together like a, a, a quiz that you have to check your understanding at the end of each section.

It's like AI in business, uh, creativity, collaboration, personal growth, community. It's not about do you use cling or crea or cap cut or, yeah. It's not about using

the tools, right. It's, it's about the mindset, right? Yeah. The mindset of AI and uh, yes. And getting ready for that. So. Exactly. So go check that out.

And with that, so are you ready for

AI com?

Yep. Beautiful. Um, why don't you tell the good folks about June? Me?

Well, June me is one of the people who I had the good fortune to meet in the AI exchange, Rachel Woods, um, AI community. He is one of the. The coho co-host of the Daily AI show, and it, um, lives just up the street from me in Portland, in Oregon.

Oh, that's, I know that and

that's awesome.

Yeah, yeah, yeah, yeah, yeah. So many of the people, we who are friends in common who live, live in this area, um, and I'm excited to hear from June me if he's gonna share with us, like he's on the airwaves every single day. I wanna know some AI stuff that June me has not said yet, that he can, he can, um, say hear out loud for the first time ever.

No, no pressure. Welcome June. No pressure.

Aha. Uh, yeah. Thanks for bringing me, uh, on, uh, it's always great to see you Ann, and, and meet you Kyle. What?

Welcome.

What kinds of things are we thinking here? AI that we don't talk about every day. That you know

specifically. Alright, well, well, why don't,

why don't, why don't, first, why don't you first give us a little intro, give us a little background about who you are and then what you up to, and then gather your thoughts and wind us into something, some secret corner of AI that some team you haven't talked about yet.

Some nugget.

Sure, sure, sure, sure. Okay. So, hi everyone. I'm June. Me. Um, I like to describe myself as a multi-passionate generalist and when that comes to, that comes to tech, media, content, strategy, all of those kinds of things. I'm a systems thinker if you, uh mm-hmm. If you will. So, um, I tend to have to stop myself.

t introduced very end of, uh,: 2022

Became hugely mainstream. Everyone started talking about it and, uh, you know, it's, it's a new tech space. So that's kind of what I wanted to go into it. Now, I, uh, concentrate on a few projects, uh, one of them being the Daily AI Show. Uh, like Ann said, it's, um, it's a daily show. Monday through Friday. We, uh, we talk about various, uh, AI related, uh, subjects.

Some, uh, on Wednesdays we have the news, which I have the honor of, uh, being the lead host on. So it's really nice to, you know, uh, dive into that hype cycle and go, oh, okay, I think this is what we really need to talk about, and those kinds of things. So yeah, if you get a chance, uh, come check us out. Uh, I think you'll learn a few things.

Yeah. But, uh, yeah, so that's a, that's a little, um, background on me. Um, so. In terms of AI that I don't, or I haven't really talked about? I guess it's be because it's fairly niche, so if, uh, yes, let's hear it. Yeah. So if you didn't know, I have, uh, a film background, um, and, uh, I like creating film and video and been in that space for, for quite a while.

Uh, and one of the cool things that I've been looking into in terms of technology is SBCs or single board computers. So if you've heard of the Raspberry Pie

Yeah. You

know, those kinds of things. And, uh, recently there has been more SPCs that have had AI components or an AI board. So the, I believe it's called the, uh, Julio.

Uh, AI hat, uh, is a board that can top out at about 27 to 40, uh, t flops or trillion floating point operations, and that's kind of the, um, the defacto, um, benchmark that people, you know, calculate. So in, uh, in comparison, uh, an H 100 does something like 2,500 T flops, right? So that's what all the big hardware, you know, it costs $25,000 for a single card.

So, you know, that's, that's enterprise, cloud, uh, infrastructure. But if you wanna run a very simple, um, small AI model, you can do it on this, on this SBC, and there's a few others that are out there.

Mm-hmm.

Now. That might be fairly niche. The one, uh, application of it that I think is very, very niche is there's a project called cpi.

And essentially it's true, uh, the project means to build a cinema grade, uh, camera using SBCs. And now with the application of ai, what kind of in-camera effects effects

can

add add onto it. And so that's, that's kind of my hyper niche. Uh, wow. Not really talking about it right now,

but that's crazy. Insane.

Wow, that that's so cool. That's so cool. Yeah. The possibilities are crazy. What's your, can you, can you take it, abstract it up a layer from the super niche thing and just talk about what you think? The implications are, or, you know, the value is of locally run AI models and, and, and. Mm-hmm. Yes. Like and how soon do you think we move from the hobbyist where you have to understand what a raspberry pie is and a hat and programming them and you know, where you have to nerd out a bit, you know, how soon do you think it's coming where, where those things are just gonna become the defacto way to go, but like, can you just talk to me about what you think the implications of onboard AI is versus Right.

Sending all,

um, hold your horses because you just asked me a question that I want to, or I love to talk about. Okay, good. That's great. So, um, alright, so, uh, right now. What everybody seems to know about AI is the foundational AI models, right? Chatt, Gemini, Claude, all of those, right? Um, that's all cloud interfaces.

That means you have to send your requests out there, have 'em, uh, spend the compute. Spit something back. Even if you're using an API, you're still making calls to their servers. You know, and going through that whole process on an enterprise level, people are concerned with, uh, security and PII and IP and all those kinds of things.

Right. And so the solution, rightfully

so,

right? Rightfully so. Exactly. Rightfully so. By the way. 'cause because, uh, I know it's been said for the last, at least 10 years or so, but, uh, data is the new gold data is the new oil, right? It's the commodity that all of these big tech companies have been trading in, and it is a value asset to them.

So I think what people are waking up to, uh, and other businesses are waking up to, is that, uh, maybe we should own our own data. Mm-hmm. Uh, maybe we should have like inherent data rights and stuff like that, if it's such a commodity. So if that's been identified, then it, it should be ours now. Extrapolating that is the concept of running OnPrem or on premises computing.

Right? Um, a good example of a AI enabled service or system or tool that takes this into account is N eight N. So, uh, n n's really popular right now. It's the new automation tool, you know, on the block. Um, but it's open source, like make, it's

like Zapier, like make,

like Zapier, those kinds of things, right?

Yep. It gets a little more on the nitty gritty side. So you do have a, you know, a higher learning curve, but you can do so much more with it. They, it's open source. They do offer a cloud solution if you don't wanna deal with any of that, but you could have a self-hosted version. And they have an AI agent, modules, and now all these other interconnections.

Um, and you'll see, you'll see things like, um, uh, how Claude and, uh, OpenAI have been pushing their connections, whether that to be the Google services or, or anything like that. Or MCP. This is all that another, um, uh, protocol layer that essentially allows you to just do more and have your AI agents do more, right?

Mm-hmm. But this concept of self-hosted allows you to do more on your own hardware or your own instances and things like that, which does help along with the security aspect and, and whatnot. 'cause you can, you control the sandbox, right? Mm-hmm. Now, if we're going to extrapolate out from that, is. Local LLMs, nothing new.

The open source community's been doing local LLMs forever, right? Yep, yep. And so you can go to Hugging Face, which is essentially just a giant aggregator, uh, platform for all of these local LLMs. You can use them with cloud resources for compute. But as these LMS are getting better and better, meaning they're more capable, but they're getting smaller and smaller, right?

So that you can run them on your own computer. Own your own compute, your own computer, right? And so you might have seen Nvidia talking about their Jetson package, um, or, you know, they wanna release their $3,000 pc. That's it's all, uh, ai, uh, focus. Um, you're seeing new chips being made that are all specifically about processing LLMs and similar models, uh, like the grok with a q, they, they've got huge.

Performance enhancement because that's what they do. It's all about the ai. So now that you're seeing, uh, personal, uh, or individual hardware level of things like, you know, Mac M four chips or if you have a GPU with, you know, 16 gigs of RAM or laptops, uh, you know, just with as much memory in their a hundred twenty, twenty eight gigs and, and those kinds of things can start running these smaller models relatively quickly.

Mm-hmm.

And there's, uh, and the secret here that a lot of people don't know is a lot of these models just run way slower. On older hardware but still run.

They still run.

So you've got all of this X enterprise, zon, multi CPU, you know, hardware that's out there because it's something like eight years old and you can buy dirt cheap with, um, an A GPU that is like eight, 12 years old that can still give you something like 400, you know, t flops.

And that's still better than nothing. And that's still extremely fast when you're running these smaller and smaller models, you know. And then we're starting to see more, uh, more models that, uh, can run on, um, that can run on ridiculously small hardware like your phone

mo mobile devices. Yeah, I mean, to, to one of the, you know, one of the things Anne said in the, in the setup.

To this is, you know, she found out the energy usage for a mid journey generation. Well, if you can generate something like that locally on your device, then you're only burning up the energy that's on your, on your device or your little computer. Right. It's, it's not taking up so much, much stuff. What do you, what do you think the, the implications of it are?

Like where do, do you have a sense of use cases where, where we'll see it first, where, where it sort of breaks out of, um, you know, sort of duct tape and hobbyists, you know, building these things. Mm-hmm. Where, where do, do you have a sense of when we're gonna see it first? Where, what Use cases.

Okay. Um, yeah, I think.

What it'll come down to is the concept of our personalized ai. Um, right. So there's companies like Hume ai, which their whole mission is to create an interface layer. It's an interface layer. Ai, right? They'll build in some sort of orchestration. So we have an orchestration layer, which we say, Hey, I wanna accomplish this today.

And it goes, okay, let me make that happen. And then it reaches out to all of this army of smaller ais that are hyperfocused on the specific component. And the orchestrator brings all that in and then presents it to you or accomplishes a specific task or goal for it.

And the orchestrator would likely run locally and then it would use as many local services as possible.

But if it needs to go out to the outside world, it would let you know that basically then it

would, yeah, exactly. Exactly. And I, and I think that's kind the, at least hardware perspective. Of things is because you won't have that ubiquitous, um, acceptance or adoption of AI until it can live in your pocket or on

your desk.

Right. And that, and that's that first step you kind of need or one of the steps that you kind of need before you have, uh, for lack of a better term, what we consider a GI. There's gotta be a, there's, there's gotta be a universalness. Yeah. Before we understand as Oh yeah. That's, that's where we're at,

is we, it, it just hit me.

It's, it's very similar to the early days of the worldwide web where the internet wasn't always just on, people had dial up modems. You had to actually dial up to the, to the internet. Right, right. And it wasn't until everyone had ubiquitous internet that it was there. And this is a very similar kind of bottleneck that while, while, while all the compute is out on these big cloud servers.

Um, that's a, that's effectively like dialing up to the internet in the olden timey days. That's fascinating.

Yeah. So I, I really see the, uh, that next step to being able to, to do that is once we can get that orchestration layer onto our own hardware, once we can do that, I think that'll be a, a, a, a big jump forward.

Love it. Love it. Wow. Very cool. That was Kyle. That was good. And nerdy.

That was nerdy. Was nerdy. I made my grocery list in my head list. No, I, that was actually, that was super helpful and I was gonna suggest, Kyle, that in your box metaphor, you may need to add the on-prem version of an LLM.

That's good.

But what you might consider is not, not using, using something that looks a little bit more techy than a box.

And then you can go around and just tell people, oh, you want your own LLM, okay, BRB. You go out there, I'll install it. You bring your little plastic thing and you like plug it in and you tell them that that's what it is. Yeah. It's, it's your own personal large language model because you know, as soon as it becomes even remotely affordable.

Yeah. How

many people are just gonna say, you know what? I need my own. I have to have my own, my stuff is so special, I have to have my own period and then the lid is off. Right. So, yeah. Yeah. Love it. Love it. So make sure that when you're doing this, you incorporate something that looks real, so that then you can go out and be like, I'm here.

I've got your on-prem, LLM look

exactly

as featured on tv. Anyway. Love it. Um, June, me what, so one of our questions that we ask all of our guests is, what was the tipping point when you, you knew you needed to go all in on ai and what happened after that?

Okay. Um, well, I think I, I, I kind of mentioned a little bit was mm-hmm.

tt G PT was. Right. So end of: 2022

2020, right.

And then, um, I think. I think sometime in January when I was talking with other people, um, you know, various projects that we used to work on, things like that. And, um, so there was a few ideas that were thrown around.

I'm like, we could use AI to build all of this, or at least develop it and things like that. Yeah. And that's kind of when it clicked on, oh, we are talking about education, we're talking about administration, we are talking about healthcare in all of these, just separate things. And I think that's where the realization was.

It's like, oh, this can apply to this, this can apply to this, this can apply to this. Mm.

Oh, you, you got the, the generalness of it. The, the, yeah. General intelligence. Right. Because AI prior to that was very specialized. Right,

exactly. And so when, when I realized that you're, you know. It touches upon so many different fields.

That's cool. And so many different applications. That's, that's a, that's a no brainer, right? It's just like with the, with the internet, it's, it's like, oh, this is information on a global scale. Oh, okay. You gotta buy in on that. Right, right. AI is, it's intelligence at a global scale. Oh, okay. You gotta buy in on that.

Yeah. We should, we should probably, we should probably have a little of that in our back pocket here. Exactly. Yeah. Yeah. That's really good. That's really good. Um, what, um, did you have, was, was there a, so once you got that sort of general thing, you said, oh, we've got all these projects. We could apply AI to them.

Was there a moment early on where it exceeded your expectations? Did, did you have one of those moments where you're like, whoa, I had like that, that it was sort of beyond what you expected? Did you, did you have one of those kind of tipping point moments?

I did, I did. It's a, it's, it's a little vanilla.

It's not really a niche solution or anything like that, but, um, I kind of hate writing papers. Mm-hmm. Um, I never, I never liked as, as someone who likes to communicate as directly as possible mm-hmm. Possible, this is the information that I need. Here's the information I can provide. Right. The, the concept of oh, and this and that, and filler paragraph and all that kind of stuff that did not, I did not engage with that very well.

So the fact that I could say, here, my 12 bullet points present this for me, and then I review it, make my edits and I'm done. Right. That,

that

immense amount of time saved that, that was, uh, that was huge for me. And then going on,

it's a huge one,

learning all of the, all of the mundane things that I really.

You know, you spend, it's so much tedium, right? Yeah. Uh, even, even when it comes to like production, it's, it's like, okay, well now we have to analyze this and we have to review that and go through all of this. And then, um, so as you go through each step, if you can save 10 minutes here, 15 minutes here, a half an hour there, by the end of it, I'm saving 10 hours a week, 20 hours a week, because I don't have to do the tedium bit bits as much.

Mm-hmm. And that, that sold me as well. That was a, I was like, oh, okay. I saved a bunch of my personal time. I'm,

yeah. Oh,

good.

Yeah. I, one of the things that, that struck me is how much I knew I hated all of that tedium kinda stuff, but I didn't realize how much I hated it. And what I realized is I spent a lot of my creative energy.

On non-creative things right on, on lower level things. And it, it just, it just SAPs your energy. So it, it, you know, I found it, it allows me to kinda live at this higher creative plane 'cause it's doing all that crap that I really hated. So it makes my life, it makes my life a bit better. Right.

You're preaching to the choir here.

Yeah.

Okay. So here's a good one. And this is, we, we, we've skirted around this, but there might be something new here. So just based on your focus and expertise and, you know, given the show and everything, um, what's the trend right now that you're following that, that you think is really interesting and why?

Well, kind of the big one is the smaller models.

Mm-hmm.

Right. Um, having more specialized or hyper-focused models who work in a team, right? Mm-hmm. Your collection of expert, mixture of experts to solve specific goals. Um, and that of course, you know, can be run on less resources. Um, I think there is a, um, and this kind of goes to the hyper niche, but, um, the edge ai, uh, applications, um, AI living as close to your day-to-day life.

I think there is, uh, there's a growing trend with that, whether that be as grandiose as embodied AI in robotics

mm-hmm. Or

simply having an AI layer to your. You know, your doorbell, your ring cam. Mm-hmm. You know, those kinds of things. Um, having that ability to have AI at the very edge touchpoint to our lives.

Mm-hmm. I think, uh, I think that's just, that'll just be a growing and growing trend. It's a little harder to see because you can't really implement it on your own. Yeah. You have to, you have to be kind of deep in, into those environments.

Yeah. Do you think Apple's gonna get their crap together and get Siri to not suck?

I think Apple needs to make the biggest purchase they've ever made. I agree. Buy a perplexity.

Yeah. Mm-hmm.

Yep. Buy perplexity right now, and if perplexity is for real about their collection of investors to buy Chrome, do it 100%. That'll be, that'll, that'll give perplexity on every single Apple device. That means Comet is the new sa uh, safari.

Mm-hmm. That means it'll be on every Mac, Mac device, Mac Os. Um, it's a huge jump into AI capabilities for Apple.

Yeah.

Um, and. Th Apple has a huge war chest. They, yeah, they've so much, they learn that back from the nineties. So Yeah. They, they always keep a few billion in cash ready to go for, for those kinds of things.

And yeah, they, they need to go all in on a perplexity or something like that so that they can jump back in. Then that will give them the runway to actually build out their own solutions. Yeah,

yeah, yeah. And things like that.

Yeah. Yeah. That, that one's been a heartbreaker for me. Like, have, have they missed it so bad.

Right, right, right. And then with Google coming out with Pixel 10 just now, with, uh, with all those AI capabilities,

with the, with the, on, on, uh, on device translation, did you see that stuff?

Yeah, exactly. It's, it's sort of like, uh oh, all of those said all of those things you said Apple Intelligence would have a year ago.

Ha ha. We have it now. We have it now.

Yeah. Well, like, you know, I feel, I feel like, you know. Apple is most disappointing. I feel like Google is most improved 'cause they're, they're 22 and 23 AI strategies were bad. Absolutely. So bad. They saw that they messed up, were bad,

and, and did a hard U-turn and it's like, oh wait, we messed up.

Okay, cool.

Yeah. Past, past six months. I feel like they actually have a product strategy now, and that, that was a disaster for a while. So, absolutely crazy. The edge, edge, the edge stuff's really interesting because it's, I, I don't hear a lot about it, a lot about it in the mainstream media, but I think you're right.

It's like as, as the. The models get smaller and more efficient. At some point someone's gonna crack that with the right interface or the right device. And it's just gonna be like, it's gonna be one of those moments where it's just here.

That's what Sam and, and Johnny Iver are. That's what they're doing.

Right. That's what they're, they're trying to do is have that interface layer and then, oh, well I guess you'll just live on our platform 'cause we provide everything.

Yeah. What's your speculation on, on what that device is?

Ooh, that's a tough one. I know we have seen a couple years of just personal devices tanking.

Um, but Oh, do you have, do you have one of the rabbits? He's got one. Yeah,

I do. Right.

Um, fun game boy design. I always love the design. Oh yeah. I bought it because I figured

if it's bad, it's good design, you know?

Right. Um, if, if I were gonna guess anything, it would have to be. It would have to be something revolutionary.

I don't think someone walking around with a medallion kind of thing's gonna work.

Mm-hmm.

Um, it, so someone told me once that, and I've kind of kept, kept this in the back of my mind, was that people, the things that, that, uh, really are successful are to take something that exists and just turn it a little bit.

Mm-hmm. Just put a little different spin on it. So this could be a new watch or mm-hmm. This could be, you know, um, some sort of attachment to a phone, or this could be some new device that just replaces what we use these things for. Yeah. Like how many people do actually use their phone for making phone calls.

Right. Very rare.

Right. But most

people text, they're on their, at YouTube or TikTok or those kinds of things. It, social media device. And so if you had an AI device that was just all about that,

then yeah.

That would probably do pretty well. Remember how popular the iPod was. Yeah, yeah. And there's a reason why we call podcasts still.

Yep.

Yeah, yeah, exactly.

Oh, very good point. Exactly. Very good point. Okay. Jun me, we, this is the AI Readiness Project podcast, which is a companion to the AI readiness Training program. So we are obligated to ask you, what does AI readiness mean to you, and what would you tell someone who was just stepping off the sidelines into ai?

Okay. What does AI readiness mean to me? Alright. On an individual level.

Mm-hmm. AI

readiness means, um. Creating the mindset of constant learning. Ah, good. This is not a technology where you just go, ah, I read the book on it, I know everything. No,

yeah, no.

Every day it'll change and you need to constantly learning.

So I think that mindset is probably number one. Right. Beautiful. I think AI literacy is probably one of the most, uh, important aspects of just general education that everyone will need, uh, going forward. Um, and it will take a long time. To fill the gap. There's always, usually, there's always a, a tech to societal gap.

And the faster that we can fill that gap, um, and train ourselves to understand that we run at this level or at this speed now, uh, I think that'll be helpful for everyone. Uh, in terms of advice, I would give someone who's just getting into it, just realized what AI is or just stepping off the, uh, off the curb there, like you said.

Um, I would say you, you need to expect, like, I guess that goes back to the mindset. You need to expect to be constantly learning everything, but don't be unmotivated or dismotivated, I guess maybe a better term. Um. When you feel that there's too much, because there is too much, there is too much, there is too much there.

Too much. And there'll always be too much. There'll always be too much. So give yourself

some grace for that,

right? Give yes, definitely. Give yourself some grace. Take a breather. Tech detox. Um, as much a tech person that I am, an information person, that I am having a detox, having uh, you know how they say work life balance?

I guess there's info life balance as well. Yeah. We need some days off. 'cause we are human beings. The whole point of AI is to be able to act as a mirror to ourselves so that we can improve ourselves individually and as mm-hmm. A society, as a species. And to offset all of the things that. Accomplish all of those goals that we know aren't really valuable to us as individuals and as a people.

Uh,

and that, whether that be mundane tasks, whether that be, uh, you know, pushing forward a post post scarcity type world, you know, all those things. And that's what it's there for. It's, it's supposed to be a tool set that we can use to go, oh, okay. We don't need to do any of this anymore.

Yeah. We can concentrate

on important things.

Right. Yeah.

Yeah. That's

awesome. That's awesome.

Yeah. That's really good.

Wow. Wow. This been awesome. Thank you. Thank you for, for coming and sharing your insights with us and nerding out. I love nerding out. That was, that was a good one.

You're welcome, Kyle. You're welcome. I know.

So good, so good. Yeah. We're often up in esoteric, you know, esoteric land and I love, I love that like what you're paying attention to is, is, yeah.

Cool. Little hardware.

Great job with the Daily AI show. June, me, I know you guys are in your hundreds and,

yeah. Hey, E, everybody in chat loves, uh, when you were, uh, there, Ann, Kyle, uh, when you can carve out some time, let us know. We'll get you on there. We'll get you, uh, on our bench and, uh, yeah, happy

to come in.

I'll, I'll, I'll sit in the bullpen if you, if you need someone, if you have a, a thin day, one day and you, uh, need to golum it, I'm your guy.

Awesome. Kyle. There's the one, there's that one issue though. I, I was giving you guys June me a little bit of a hard time, but it's actually, it's actually a backhanded compliment that when I was on the show, I realized.

How much I move around because of how still you all sit. You sit so still and you look at the camera and your hands are not in the frame. And, and it's just like I'm in, I'm, I'm like in the zen mode and then here's me like,

yeah. And so I told Kyle, I was like, oh my God, I'm so glad we're to be back. 'cause now I can move around. But now that I'm making short clips of our shows, I realize how valuable it is that you guys sit still. Because I cannot get a ten second video without me and Kyle all over the place. Yeah. Dammit. So Kyle, if you go on the show, you have to sit still, just so you know.

I'll, I'll take a Xanax before I go there. We,

we always value you being you, so

thanks. Awesome. Well, this was just awesome. Uh, yeah. I look forward to it. And, uh, thanks for, thanks for coming here and sharing, sharing your wisdom. It was really you,

me. Anytime, anytime. Take it easy. Cheers.

Bye.