Everyone is Becoming a Technologist

We're just gonna go straight into the wins of the week, and we're gonna make Julian go first.

right. Well, well, well, something that will nice segue into the episode. I was in Tasmania, uh, last week, so that was a huge win for me. My first time to the state first time to Hobart, obviously, and it was just such a beautiful place. I had the best time. The people were lovely. The city was beautiful.

It was this kind of colonial architecture, but you just... It, it felt like it was surrounded by nature and so untouched. And you know, just as importantly, I was invited to this conference, this, um, ethical investors conference or ethical advisors conference, and was just lovely meeting all the people.

Everyone was so welcoming and thoughtful, and I got to engage in some brilliant conversation around, uh, ethical technology use and just how they make their choices in their financial portfolios around whether these companies that they invest in or include in their portfolio are actually doing ethical things. And to me, as a technologist it was very nice to hear non-technical people talking and paying attention to technical topics like AI data centers and Palantir and all these different things that would influence whether people would want to invest in their products. So it was very cool, and I probably butchered that, what they do.

I probably com- They'd be like, "Did you even listen during this conference?" But it was wonderful, and we had the best time, and I got amazing feedback at the end of it, and I'm just... I'm very excited for what

That's

next.

the... Did you get to go sightseeing?

Yeah, so I for those of you who might know me from other things, I w- had... I've been on this other podcast with this a gentleman named John Dwyer. His podcast is called JD's Journal, and I've been on there twice now. And he now lives in Tasmania, and shout out to John actually for driving four hours down to come and see me, and he drove me up Mount Wellington. I have incredible photos from the top there, and it was just the most beautiful experience.

I had, um, a lot of sightseeing as well as work, so it was good

Why is everything so, like, big and far and crazy and down under? You guys, don't you have the biggest tran- tarantulas and biggest crocodiles or something?

I'd say it's

Not the biggest Julians.

yeah, no, a very small jewelers. No, we, we have, um, a fraction of the population you guys do, so there's

just more nature than people now.

Yeah, it's beautiful. I wish, I wish that was, uh, the case and not here in South Florida. It's ridiculous. All right, what's my win of the week today?

Tell me

Ah. You know, I'm doing these, um... This is kind of cool 'cause this ties in a little bit to what you're doing a little. Um, I've been doing these consulting hours through our school's consulting thing.

I don't know what the, what it's called, but it's called Pine Crest Next, and we help other schools with a lot of things. It's not just AI, but that's obviously my role. But we'll host events for VR and doing ma- or doing, like, math right, and reading, and entrepreneurship, and professional development.

So we host all these kind of events and collab sites. But we started doing consulting hours for AI because I think people have found that they really want more than just the collab sites or the trainings for, like, uh, you know, the 30-minute trainings we normally do. And so I consulted with two people today, and it was totally different.

Both, both of them were different. One's trying to do a workflow and solve a problem, and it's just, just talking to them almost like a project manager w- you know, does and trying to get them to, um, use the correct terminology in order to get their prompt to do what they want or to build an app script or whatnot.

And the other one was talking about, you know, just onboarding the rest of her team and getting everyone in. And I think for me, maybe it's a personal win. It's just this fun problem-solving. The thing I missed by not coding so much anymore is just listening to someone, they come up with this problem, and I'm just like, "Okay, let's talk through it, and let's break down the steps and go linear," like you do with coding and kind of...

Well, you know, not linear, but somewhat linear, and try to solve this problem. And I think it's been a, it's been a brain boost for me, and I, I,

felt that, I felt that back in 2017 when I was learning Python, that whole your brain hurts phase.

Yep

after my brain's been hurting for 30 minutes, and then I'm, like, just a little bit energized.

So I think that's a, an interesting win, and I, I think that also ties in kind of to what we're talking about today about becoming a technologist 'cause I feel like finally maybe I'm becoming more of a technologist than I thought

Well, there's, there's an essence of technology leadership to it, right? And I think, know, as I've been going on this, this journey and doing the... Oh, God, I used the word journey. I apologize. Um, now as I've been doing this stuff and moving and more into the CTO space and taking on these roles you know, people sometimes be like, "CTO, like, come on, isn't that a bit much?"

I said, "No," because it, it's exactly what the role is. I'm not just building software. I'm not just solving one small problem. As you were talking about, I'm understanding the problem. I'm understanding the business outcome. I'm understanding the people. I'm getting people on board and winning them over so that when change does inevitably come, don't hate me and, you know, set fire to my house.

So, um, it's really, there's so much more to it, and I feel like that's the path you're going down. That's what

Yeah.

I,

I think so. I think so, and I, I, I was just, I put this link in our, our show notes. I'm sure Julian has not had a time to read it, and I actually read it in July s- but it was from OpenAI about how AI is expanding what people do at work. And they went through this whole fact about these occupations are now, like, crossing jobs, and maybe it's, maybe that crossover is because of AI.

And I kinda... I'll put that link in our show notes, hopefully, if I remember, because I sometimes forget. But

Yeah

I, I find it interesting that technologists probably stay the same. They're just now being able to widen their gap of things that they can do. No? I feel So, uh, case in point, my tech director was having issues with the internet, and we, we outsource some of the stuff, security, whatever, because it's a school, and you don't really have the, the people to deal with that.

And something was happening, and he was getting all the error logs, and he was using the two different two different AIs, the AI from the company and another AI to cross-reference what the AIs were talking about in the logs 'cause he's like, "No, I see it in the pattern. This is what's happening. I see it in this log.

You know, it's this." And then one log was like, "Oh, yes, you're right." And the other, the other AI was like, "Oh, yeah, that's really what's happening too." And so he's able to solve the problem

Hmm.

and become... Even though he is the, you know, the IT director, but he's not an internet problem solver

Yeah.

but now he's able to solve these problems, so it's quite interesting

Yeah, I, I, and you know what? You see that across all roles and, and sectors per that article, right? People having access, and o- one of the eye-openers to me as I've been doing this is so many companies are using, uh, M365, Microsoft 365, right, for their business stack to handle email, to do all that kind of thing. But that includes Copilot. If you pay for that license, it includes Copilot that then has access to your email, to all of your company infrastructure, such as SharePoint and everything. I say that because what it means is people have their documents, who have their work living in this Microsoft environment suddenly have access to a tool can see all of it, parse it, and then build some pretty cool stuff around it. You know, so their work is changing, they're getting more comfortable with technology, they're starting to take... And this is exciting and scary at the same time fine line between the two. They're starting to take almost risks and try new things, experiment, such as create a dashboard for something give me these updates, go through my email and tell me what's most important today.

Um, you know, potentially even look at this customer database that we have stored safely in our environment, but have a look at this and tell me who haven't I connected with in six months? Who sh- who is a high potential client that I should be checking in with? These kinds of things. And so people are, you know, starting to play more and more with tech related to their specific role that generally has nothing to do with technology, and I think that's cool and scary

I was gonna say it's kinda scary. I'm o- I'm on that scary phase because another example today, I spent, ugh, I spent all of, almost all of yesterday and all of today down a, a research rabbit hole. I've been trying to do two different things, m- all, all on the cybersecurity side. But for example, I'm trying to find out what really, really happens to your images when you import them into an AI, and obviously all the AIs have different scenarios.

So for example, Canva AI, which is an educational tool it doesn't really say what it does to the images. It says, "We don't train on them," et cetera, et cetera. I- it's supposed to be safe, and we, we delete your data. And you're like, "Okay, whatever." And then you look at Gemini, Google Educational workspace, it retains no memory.

You ask it what it's, well, you know, what it does, and is there any metadata, are there any embeddings what do you do with the images? And it... And can you, can you find out who the person is? Are you taking any of the traits from these, these what are they called? Pix- I don't know. They're not pixels, but numerical...

Do you know what they're called? I had it earlier. But, like, when a picture gets split into num- numer- numbers in the AI, it's not, it's like, it's not the pixels, but

Hmm

almost, I guess, encodings, but that's not the right word. But anyways, and it says, "Oh, we can't find anything." But nobody really knows, and they always just say, and these are these technologists, right, "Oh, it doesn't train on your data."

Well, I was like, do you even know what it does i- in these, these models? Do you even understand what happens? Because it's, it's not text. It's not really a large language model. It's doing something else, and no one seems to understand. And so it's funny because I become... I feel like I'm becoming more of like this technologist, but it's because I have these, like, tin- tiny little understanding specks of knowledge that I've had somewhere else, and that's scary, right?

If a person's becoming a technologist and they don't really have any, any background. No?

Well, yes and no. I-- but I think the fact that you're able to ask those questions, you know, for you specifically, Kelly, puts you in a great spot, you know. Um, other people... And a perfect example that, that's very similar to what you're saying is in this financial sector, right? There are a dime a dozen, these financial advisory AI products that are coming out, and they're really exciting.

They do a lot of automation. They do this kind of thing. They say, "Yeah, we meet Australian leg-legislation," blah, blah, blah, right? But the legislation currently, and, and, you know, maybe don't quote me on this just yet because I-- it's been a few weeks, and things change every week now, right? But the legislation is that no customer data, no-none of this stuff should be stored outside of Australia, right? What it doesn't say or it didn't say at the time when I was looking into this product was, well, where is it processed? And as long as the product says, "Well, we don't save your file outside of Australia," then that's okay, But if, for example, if this product relied on Claude, on Anthropic, Anthropic doesn't have any data centers in Australia yet.

They're building "

Yeah, they do. Oh, they don't have it yet? I thought they already had

They've started constructing it. They've-- So they've been hiring and building, but

them

think... You know, again, I'd have to look into this for recent updates. But when I looked into it, they didn't have physical servers here, so you can't say, "I want my stuff processing onshore in Australia,"

Угу

it goes to Singapore or the US. Now, by nature of going into a server over there, getting put in RAM, and then-- or video RAM, and being processed and all that stuff, is anything saved? I don't know. Does anything get paged to the hard drive if memory's full? I, I mean, I don't know. And so where does the buck stop with this? That's the question I'm constantly asking, and I don't have the answers because no one seems to have the answer. Governments certainly don't have the answer. It's all a bit of a gray area, and it kind of is just, well, where is it stored? Where's the end product stored? So but where did it go to get there? And so I think on one hand, being a technologist is exciting because we get to ask those questions. But with everyone becoming a bit more tech-savvy with this stuff, we're kinda chasing our tails trying to plug these gaps and go, "Hang on, did you consider this?

What about this? What about this? Are you liable for this? Are you at risk here?" And what do we do? So-

It's crazy. I was think- you know, I'm gonna throw these questions out 'cause I think they're pretty cool, and I think this all stemmed from this idea that I feel like I'm becoming more of a generalist. I do not know one thing really well anymore, I feel. I feel. I have either... I have this little bit of FOMO and imposter syndrome going on at the whole, the whole same time, and every time before I have these, these consultant calls, I'm just like, "Oh, my gosh.

What if I can't answer the question? Oh, no worries. I'll just open up the 20 AIs I have going on at the same time. He'll never know. Hold, please." Um,

me read

my screen quickly. Yeah

but I wonder how much technical... This is a question that was... I, I generated, was like, how much technical understanding does a, a technologist actually need now, right?

I think a fair bit.

A fair bit, right? You have to know something. I don't think I would be where I was, I am today if I didn't have this podcast because I've learned so much from just talking to people in the industry, and I pull up... At least, at least I have some vocabulary terms. I might not really understand them completely, but I just don't feel that people can be technologists relying on AI.

If AI is only as smart as the user, supposedly, and I'm not putting in enough technical information into my conversation, are they really getting that far?

Mm-hmm.

So

And, you know, as, like you're talking about, this podcast has given you a good foundation and really built you up. So that's the experience that you have that you then feed into AI, LLMs, and whatever to get more robust answers and to ask the right questions. But had you not had that experience, of actually doing the work, of actually coding Python, of interviewing people, learning things to be able to interview people well, um, you know, connecting with other people, having these conversations, going to PyCon, if you hadn't done all those things that actually built your skills, then you wouldn't be that great working with AI.

You'd be chasing your tail. So similarly, you know, as a technology leader, having to know about the technologies that exist and then be able to draw on my experience from 20-something years going, "Oh, yes, I remember what Acti- Active Directory is. Yes, I remember network stuff. I remember, um, PABX systems for phones.

I remember this. I remember storage backup solutions." And, and so on. Because I have that experience, I'm not just relying on an LLM to go, "Hey, how would I do backups for a company of this size doing this?" You know? instead,

It's like talking into your wrist. Help me. I have no idea. Explain to me active directory like a kindergartner.

mayday. I need help. What do I... I gotta use the bathroom quick, boss, give me a second. Um, you know, you're, you're not doing that. You're actually calling on the experience, and instead, your question might be, "Okay, look, I, I've been out of the game for, for two years. What's, what are the current backup solutions available to a, an environment like this?" And, you know, or something like that. So I think you have to have an understanding to, as a technology leader, of the environment, of the industry, what's coming out. it's great to have years of experience playing with the tech so that you can make these calls, and I think you should constantly be playing as well.

I think it doesn't matter if you're a, a, you know, CTO of a multi-billion dollar company or, you know, a, a small, medium-sized business, you should be playing with it after hours. You should be learning, you should be tinkering, should be building things to just get a real understanding. Because then when the company needs it, you can say, "Actually, that's not a great idea, and here's why."

That's what I tell everyone all the time. No, that's not a great idea. Don't just go to Claude and turn on all the connections.

I think no is the most powerful word as a

Why?

leader right now. And I get, "Oh, no," or maybe, "Not yet."

I've been, ever since Jeff's Jeff's episode about cybersecurity, I just keep saying, "I have the zero trust philosophy."

Ühüm.

I'm, I'm all about zero trust. But I think that goes into being a technologist. I don't know. It comes into that. What about, Oh, well that's, uh, there's another question. I like this idea.

Where does knowing enough become dangerous?

That's, yeah, that's a tough one. I think, where does knowing enough become dangerous? I think it's when... I don't think you can ever know enough enough, right? But you can think you know enough. I think it becomes dangerous when you're talking about implementing something with production access, production data, um, and just overall things that if you get it wrong, it's going to impact the whole business.

Could lose customer trust, could lose customer data. So I do think it, it in the space where you are given privilege. So if you happen to be... So another, another slight example is if you're the director of a company, you know, and some of these small, medium-sized businesses, if you're the, um, director, you will generally have a decent amount of access to tools. if you would know, if you know enough to be like, "Oh yeah, I know how to create a, a dashboard. Oh yeah, yeah, I know how to to create this tool. I'll just do that," then that's when it becomes dangerous because you know enough to create something or an AI to create something. But if you don't know enough about the security aspect of it and what you should and shouldn't do, then it, I think there's a problem

I kept thinking if Sean was here, I'd we pick on him because I can hear, I can hear it right now. Knowing enough when he first had his first Adafruit hardware, and then once he had the one little Circuit Python, then he started building all this crap in his,

In

in his, his... Well, in his house,

in the, in the classroom.

Every time I turned around, he was building something with some sort of board, and I, I was going through the classroom today trying to find my PyPicos, which I wish I knew where my PyPicos were because they've walked out of the classroom. There's like 30 of them disappeared. But,

oh

I'm pulling out and I'm like, "Oh, I have an ESP board.

Oh, I have a FeatherWing. Oh, I have this skull, and I have all these boards." And I said, "You know what? I bet you I can do these stupid boards now that I have AI." And I grabbed a, I grabbed an ESP board, and I was like, "What can I make? I wish Sean was here. I'd make something." But yeah, I think that it-- Knowing enough does get you in a dangerous spot, I think.

It's kind of, um, with the kids when they know that they can make something in, in our, in our AI that's for them. They started building all these games, and I was like, "Oh, no. A game." I just, I just probably ruined it for every teacher out there, so now they're making all these battle games in it. It's pretty funny.

All right. I like these questions. I'm gonna ask you some more. Does AI make expertise less important, or does it make n- knowing when you need expertise more important? That's interesting. Does it make it less important, or does it make knowing when you need... I, I can... Mine's the latter. I, I

Yeah, you go

time I dig into the, into AI, I'm like, "Oh, my God, I don't know what the heck we're talking about."

And then I go down another rabbit hole of something else, and I have to ask it to find me an article, and then I have 45 articles I'm reading through and I'm just like, "Oh, my gosh. I need to know what this is."

Yeah.

And

with you. Absolutely the latter. I don't think it makes expertise less important at all. I think it makes it more important. But the, the key is that because everyone's using it to tinker and play, having someone to talk to about it is one of the most important things. And I'm not talking about someone who's the same, in the same echo chamber as you or someone who's just your, you know, your neighbor next door, whatever, but someone with the expertise, you know?

So you and I can bounce these ideas off each other throughout the week, and you might say, "Actually, that's a terrible idea." And I might say, "Well, why? Tell me." You know. Um, and so a lot of people, it's that typical shiny ob- object syndrome-type thing, right? And I was just chatting with two of my tech mates, which we'll talk about in another episode, actually, an ethical episode we'll, we'll do. But I was talking with these two guys this week, and I was floating this idea that I had because people around me were doing it, and I said, "I need ch- I need you guys to be the the, the counterweight to this discussion, because I know how you, you operate." And so their expertise came in, I was like, "Bang, that's exactly what I needed to hear." Right? So you can easily get swept up in the shiny stuff, so it's, it's really important to know when that that's the point when you need some expertise to come and balance it out

I wonder though, because I was thinking, you know, in Python I'm self-taught. Were you self-taught?

Yep

Yeah. So we both were self-taught Pythonistas. So are we gonna be self-taught... Are we technically self-taught technologists? And will other people become self-taught technologists if they want to?

Well, my, my, my counter question,

Kelly, is where is the school of technologists that exists? There's no such thing. Uh, honestly, I feel like to be a technologist, it's through experience.

It's through experience, right? It's just through doing. You can't just go into... Yeah.

man. Yeah, it's, it doesn't work, it doesn't work like that. Uh, you can go to uni and, and do all this stuff. Actually, very exciting, you, you might be interested. One of my mates is, um, who listens to this podcast, so shout out to him. He'll know when he hears his name what, that I'm talking about it. But he's just made a decision to become a teacher next year.

So I'll

Run. No, run. No, I'm just kidding.

no, but he's going to be doing maths, physics, that kind of stuff. He's, he's got a big brain, and they have a new course or syllabus out here, I think. Something around enterprise computing. And yeah, I don't know what's in it, but apparently it comes out next year or something, and he, that's where he's been the bulk of his career for the past 20 years. And, um, even so, back to the point, with ev- let's say it is enterprise computing, they teach you about servers and infrastructure and networking and things like that. That's great, but you're not gonna become a technologist leader or anything just from having studied that. You, you've gotta get out in the field and play. You've gotta make those mistakes, cause outages that cost millions of dollars, you know?

That's like the people that are trying to hire the 15 years of AI. You need to have 15 years of experience. I'm like, oh, please. The only people who have 15 years of experience are the people that were actually doing the machine learning and

Угу

let's go call, uh, Fei-Fei Li. She was started in 2007 or something.

I was laughing. But yeah, I think, I think that's interesting. It's like, uh, my title at school now is Applied AI. I mean, blah, blah, blah, it has other stuff in there. But Applied AI is, is quite interesting, and that's like a new... A lot of people are now using Applied AI, and what that means is we can apply I can use AI to solve my problems.

No, I'm just kidding

No, I was gonna say, well, what's the elevator pitch for that? If someone comes up to you and says, "Oh, Kelly, you know, ap- applied AI, what does that mean?"

Means that I can use AI to solve problems and to yeah, I don't know. That's it. I can't tell you what I really... I, I really don't have an elevator pitch because they, we were talking about names, and I was just like, "I have no idea what a title would be." I, I out- crowdsourced people on LinkedIn, and I still didn't see anyone really come up with anything great.

They have director of AI. What's director of AI? At a school, what are you doing, director? Are you writing your agents? What are you doing? Write a prompts, you know? It's, it's not like, it's not like for me, a- as maybe I shouldn't say this, but for me, it's not like I'm Fei-Fei Li who was developing image recognition and understanding what's going on with the machine learning.

So sometimes I feel the expertise that I have, which is a lot, and I can spew off a lot of things that I feel confident are true, but I think it's, it's different than maybe if you were a technologist back 15 years ago or 20 years ago

Yeah. No, absolutely. So I think this is a, a good segue to the next question

Okay

what does this all mean for our own careers?

We're, we're just gonna be more generalist, I think.

Yeah.

I, I was I was having one of those moments today where I love, I love, honestly, I do, I, I joke around with my job, I- but I love what I do. But today I was having the, "I hate AI" moment, you know? Just like, "It's gonna make us all stupid." And I was getting on the, on the negative bandwagon o- in my head, because I feel sometimes when the kids, w- for example, in exploratory, they were building in an, like a, a game.

And I told the kid, I was like, "This, this is not good. You know, the game, I don't understand what's going on." He's like, "Can I just have more tokens?" I was like, "No, you cannot have more tokens until you tell me why, what are you gonna use these tokens for, or these credits?" I know. Well, I wanted him to explain because I went to the game, and I was like, "This is dumb.

It's like a flappy bird." And bless his heart. It was a good game, but it wasn't... He, what he wanted to do, he didn't really understand, so he couldn't communicate the prompt clearly, and so he was just gonna end up spending another 10 project tech tokens or credits on a non-formulated question. And I was like...

He says, "Oh, I need more energy, but I don't need more energy because, you know, not everybody can have energy, and every time I click the button to jump up, I lo- use, use up the energy." And I said, "Yeah, but if you just sit there for five seconds, you get all the energy back, so I don't understand the point of your game."

And so, you know, I was going into that. I was rambling. But that's,

no.

my thought process of this, "I was hating AI today," because there was no thought process, so.

Okay, two, two things.

Okay

in cheek. mentioned that the, the game might have been a bit dumb. Was that dumb with a B? Because I hear that not many people know dumb spelt with a B

Yes, it's done with it.

oh, just checking.

unless you're having a Dum Dum, and then there's no V. Those are those little lollipops. Have you had those?

No, I've never heard of that.

Okay, we're g- next time when you come to PyCon, I'm gonna bring you some Dum Dums because the root beer Dum Dums are my

What a

favorite.

What a name. I love it. No but, but s-side note, you mentioned, like, this token thing. This is something that's foreign to me. I'm, I'm not a school teacher, right? So the students have limited token usage and then-

So this is, uh, uh, maybe I'll do a plug for Dora Palffy, who owns the, the company Imagi. She was on our show, I don't know, whatever episode, a while ago. I met her at PyCon, and she has this company, Imagi, where they make these little, tiny key chains that are NeoPixels. Well, now she also has the Lovable app embedded within Imagi that's completely personally i- identifiable information is gone, so the kids can use it, and it's not guarded.

As the Lovable, you can't use it if you're 18 and up because you know how you have Supabase, and you have these databases, and you can't guarantee that the, the kids' information is, is safe. So she put it behind her walls of her company in Imagi. And as teachers, we buy, or I buy, the project credits because it runs on tokens, which is nice because then you can teach kids about what it means to use these tokens.

Because normally, when we use Gemini, they don't have a token quota. When we use Canva... This is all my technologist skills coming out, by the way. When we use Canva the whole school is on a AI usage token, and so we get close every month to running out. And when we run out, Canva just kind of like slowly generate.

It takes

Mhm.

to generate. Where these tokens in Imagi, I can ration them out. And I have a lot, and I literally could give them tons, but I give them five in the beginning for them to generate a game. And I try to explain to them about communications. And you can draw, you know, generate websites, and you can talk about front-end and back-end, but this goes into the idea of becoming a technologist.

Kids don't really understand how to use the prompts correctly because they don't have the vocabulary, and they don't have the knowledge to actually create something really worthy. They create amazing things, don't get me wrong. And they love them. They look great. But yeah, so that's the tokens and credits

No. Oh, that's, that's really cool. I like that. It's kind of like force... It's adding a bit of scarcity, scarcity,

Mhm.

in there, and that way they have to be thoughtful.

Yeah.

I, I really like that.

And I love the power. "Oh, no you can't until you come up with the good prompts."

Well, okay, so to the, that question, what does this mean for our, our own careers? I actually think this means we have to step up, right?

And I know this sounds a bit dramatic, but I think as people with the experience, the lived experience, the experience in the field with work and whatever, I think this means there's an expectation or a hope that people like us will step up and actually support colleagues and, and friends and family and everyone else, right? So, uh, a perfect example is someone yesterday said to me, "Oh, can you help me connect ChatGPT to my work email?" And I was like, "Are you what, mate? What did you just say to me? I don't think so." And, and, you know, I said, "You... But you can't... You don't, you don't just abuse them and..." "Oh, what are you talking about?" The question was "Well, how come?" They said, "Oh, you know, I want it to help me with this, this." I said, "Okay, that's fine, but you know we have Copilot, right, as part of the company license." "What's that?" Right? So it's... And that's fine. People don't know what they don't know, right? But this is where I think it takes people like us to be able to stand up and, and people listening to this as well.

It, if we don't wanna see the demise of, of, of our companies and our friends and families and people, we have to step up and we have to stand up. I think that's the biggest change to the career is, and that's why I'm loving, absolutely loving what I do now because it's not about power like you, Kel. It's

power of the token.

power that I am the token keeper. No, it's... But I, I think it's more about you know, the confidence that we've done this and we know what we're talking about to guide these people. And more often than not, they're, people are receptive. So I think everyone is becoming a technologist, not everyone has that lived experience as a technologist to do it safely, do it effectively, efficiently, and so on.

That's good. I was gonna ask the next question, but this is... I'm, I'm hoping to get one of these, a, a guest speaker. I reached out for him, and he was a little busy at this time, but he's really interested. I I did a self- a selfish plug because I had listened to a podcast. Anyways, and he had mentioned Hightower, what's his first name? Sorry, I just went blank. Kelsey Hightower

on this, on this podcast, and I said, "By the way, Kelsey Hightower was on this episode with us." And I think he was like, "Oh, really? Kelsey's on this show? I'm coming." But,

join too.

but this last question had me thinking about this. So thinking about you, me, and Sean, all of us kind of met each other through Python, but we've all sort of branched out a little bit, and I think a lot of people are branching more.

There's some people that are still holding tight to Python. But instead of going deeper,

Mm.

we chose to go outwards. Happy with that? Yeah. Yeah, I think s-

happy with

yeah, I think

Yeah

for, yeah, I think for me, I think I think Python was helpful, and it's still helpful, and I think it's still fun to, to try to solve a problem with Python

Yep. look, I'll, I'll give you a real, real example from yesterday. I'm working on some data migration for one of my, my clients, right? And once upon a time, writing that Python script would've taken weeks, right? Not because of a lack of expertise, but because the constant to and fro. "Okay, that worked, now that broke," or, "This data's not in the right format."

Oh, or, "The source schema requires formatting of this kind." Oh, but that wasn't in the documentation, so now I have to figure that out, trial and error, see what works, right? So there's all sorts of different things there, but that was tedious. Tedious as all hell, and I... When I was doing some of this yesterday, I was using AI to do it, and m- what was more important to me was not getting stuck in the nitty-gritty of I don't know, a Python model or a, a, what do you call it, a class or anything like that. It was just about, look, I need this done because the business needs to move forward. This is not a production script that's going to run. This is to help me, know, mold this data set into this data set, and I will go through it. We'll fact check everything in our own way. But so for me, I didn't-- I don't want to dive deeper into Python for that, because now there's also a business expectation that these things should be quicker because of AI. So there's a kind of a trend change in the industry. There's an expectation change from leadership that you don't spend six weeks on a little script like that. Um, and even for me, it's interest, right?

Instead of sitting there bashing my head against the wall, I'm like, "I don't have to." But having the knowledge of how it works, so when the script didn't work, I went through it and I was able to read it and be like, "Okay, yeah, this, this, this," right? But writing it from hand, from scratch, as much of a, as a badge of honor that can be, and something to be proud that you can do it, um, for me, at this point, I'm like, "Nah, man. I, like, you take care of that. I'll fact check, I'll review, I'll test." And a- again, I'm gonna preface this with this is not a production script, right?

Okay. But at the end of the day, you know, I'm enjoying the conversation and the output and the guidance and leadership that comes with it.

Okay, cool. I was just made me think of Simon. I didn't get to read the whole article, but Simon Willison's post that came out yesterday, today, he made... I just quickly skimmed it because, you know, I have like 5,000 blogs and Substacks coming into my emails. But his was, he made a quick script with Python to identify like the sounds in his room or something.

What?

That's what I'm telling you, you gotta read it. I don't know. He, he found out there was some sort of library or something that could... Because large, the large language model couldn't hear the sound, but he gave the AI ears, and I think that because he has Python knowledge and understanding that he- it could happen, he can fix this, he can figure it out, he didn't have to spend his time doing it.

I'll have to find the actual article.

That's cool

yeah. Where was it? It was, it came out yesterday or something. But yeah, so he made a, he made a l- sound location app.

to his LLM.

to get the sound to... He's great. He's great.

I like nerds. We, we come up, we, we over-engineer

Technologists, not nerds, you

Oh, sorry, sorry, technologists, yes. Same thing. It's a the- the- the- thesaurus thing.

thing. All right, anything else to add? Why don't you explain real quick what, what a CTO does, you know, now

can

in

I

in your world?

In my opinion, this is what they should be doing. I think they should be leading the technology-- technological direction of the company. So I think it differs if the company itself makes a technology product, CTO there is gonna be at-- in-involved with that, right? But I think if you are a company that is doing any kind of other service that's not a direct technology product that you sell, then a CTO's job is to influence the internal direction of the technology.

That doesn't just mean the infrastructure that runs the company, but also the way people use technology the way that the company approaches technology, the way they engage their customers, and anything as well that will help the company be more successful using technology. So there's a bunch of pieces there, but thing that makes you a CTO as well is you have to be a-- you have to have people skills. I'm sick-- I'm gonna get on a, a soapbox for a moment, but I'm sick to death over the years seeing CTOs have-- who have no idea how to communicate, no idea how to communicate changes, how to talk to people who just don't care about the people. You know, I think a CTO these days should have more people skills than raw technology skills. you can't... what's the point in doing it if the people aren't on board, you know? So I'm making some changes in some of the companies I'm at now that are not popular, and I sat there with the stakeholders of the different teams on Monday this week and chatted with them on them and said, "This is probably gonna suck for you.

Let's talk it through." And then they-- the main question was, "Well, why?" "Because we have this limitation in the hardware stack or in the technology stack, and this kinda makes sense for everyone. Do you reckon you'd be able to give it a crack and try it with me? We can test it first." And they're like, "Yeah, of course.

Thanks for including me." know, and so that's what I think. I think CTO needs to have this beautiful, perfect mix of people skills and the technology understanding. There you go.

Yeah, there you go. I like that. I'm not gonna add to that. I think we're at 42 minutes, and we said we're gonna do a 30-minute show. Look at us.

Oh man, it's lunchtime. See you.

It's bedtime.

This is perfect

anything to add?

No, nothing to add this week. I, I do wanna talk about ethics in technology one of these weeks, so I have some- something to talk about with that. But that's, that's about it. I'm, I'm good. I'm good, Kel. What about you? Anything

tomorrow I'm meeting with some members of the Python Education Workspace again. That's one of the things that Trey does for PyCon and Python. PyCon, P- for the... Yeah, something like that. And so I'm meeting with them, and we're gonna talk about the Ed Summit again because there's a lot of things that go on before the Ed Summit, even though the website hasn't been...

I have, have not released the PyCon 2027 website. I've looked, so it's not out yet. But I do know the dates are there.

Oh, nice. Yeah, yeah.

so you should better get your booking, get your booking.

All right, man. All right. I will. Um,

Long Beach, it's close. It's before it moves to the East Coast. You better get on it.

Yeah. Oh yeah, three flights away.

Actually, one, one last thing I'll just say because you, you mentioned it. I'm still happy to hear that there are Python meetups happening.

Yeah

a shout-out to, uh, Jeff Jacobson, Chris Patty, and Fay Shaw, who were all at the Boston Python meetup tonight or yesterday or something. And just so cool that people are still getting together and talking about the stuff that hasn't been squashed by AI and the way the world is going and stuff. I love hearing that that's

happening.

someone has to be a specialist in Python so that, you know, they can look inside the, the black boxes and figure out what's going on because some- most of us won't really know what's going on because, well, I don't know, we gotta have those really, really smart people like our, like your bestie.

Mm,

Yes, definitely.

Yeah.

all right. So, well, for teaching Python, this is Kelly

And this is Julian

Signing off.

Nice. I love it