Wax On, Wax Off: What AI Can't Teach a 14-Year-Old

0:00 The Pragmatic AI Podcast is sponsored by Tighten. T-I-G-H-T-E-N. We will take your AI ideas, prototypes, or even vibe-coded apps, and we'll take them to production. Scalable and secure. Check us out at tighten.com. Hey, and welcome back to Pragmatic AI season two. And thank you so much, Bill, for being my first guest for the second season.

0:23 ~ for those of you not familiar with what we're doing, the second season is pretty similar, but we're just kind of learning a little bit about what we, you know, what worked and did not work in the first season. ~ and so, as the first season, we're talking about. what it looks like to use AI in the real world.

0:38 And this is about having a clear-eyed perspective ~ to try and find a middle ground space between the maximalists who just always want to talk about AI is going to transform everything and the Luddites or the minimalists who think everything about AI is bad. We're trying to find a practical conversation right in the middle for builders, business leaders, and also normal people.

0:57 My guest today is my friend and I I need to find a way to talk about my friends from a long time ago, 'cause not my old friend, because that makes it sound like you're old. But anyway, my friend from when I was very young, Bill Van Loo There you Engineering robotics Educator, a creative technologist, and an independent electronic musician.

1:13 Bill, would you say hi to the people real quick and let's get started from there? Yeah, thanks so much for having me on, Matt. I really appreciate it and I'm excited for the conversation. Yeah, as am I. ~ so for those who are not watching on YouTube, you just need to at least go at some point and take a look because Bill's setup has the absolutely perfect setting for you being a creative technologist who also possibly moonlights as like a mad scientist in your basement, which is kind of what's going on because you mentioned, A little bit. Yeah, a little bit.

1:41 you know, creative technologist, independent electronic musician. Those got some like kind of mad scientist vibes with all the stuff you're creating. So can you tell Yeah, a little bit. people, walk through each of those independent roles and just kind of like What do you do on a daily basis? Yeah, for sure.

1:54 So my day job is that I run the engineering program at Pioneer High School in Ann Arbor, Michigan. I've been full-time at Pioneer specifically since 2022, and I'm actually starting my 20th year of teaching right now, which is a little wild, yeah. It does Yeah. not seem like it's been 20 years. ~ part of that is because I'm also a career switcher.

2:14 coming out of college, graduating in in '97, ~ I came out and did web development for about eight years in downtown Detroit, working for a variety of companies. Companies doing front and back end development, but then also a bunch of interaction design and project management and software testing. ~ I really loved that work, but I felt like there was maybe something else out there for me.

2:36 And so my wife kind of prompted me and encouraged me to consider teaching, which is something that I'd been thinking about anyway. And so I kind of, you know, accelerated that timeline and moved into education, ~ where I've been now for, like I said, almost 20 years. Yeah. So that was your first one.

2:52 So you are a teacher and you're not just a teacher, Yep. but you're a teacher specifically in technology, engineering. So there's gonna be some some things we talk about there. And what's your age range? Is it just Yeah, a lot of connections, Christian. all of high school? Yeah, so yeah, so my classes, ~ depending on which one it is, I might have ninth through twelfth graders in the same class, especially for that introductory class, which is super cool. You know, a lot of times students who are ~ like pretty sure they're gonna be engineers wanna take that class right away in ninth grade so they can do as many in the sequence as possible.

3:19 But then I also get folks like a senior who emailed me yesterday and was like, Hey, could I get in still? Like which class should I take? And so I tried to give her some guidance about the different coursework and all that. So yeah, ninth through twelfth grade. ~ I also do ~ work as the lead mentor for my wife's two middle school robotics team so I get to keep a foot in the middle school world which is hilarious.

3:40 They're such funny weird people in middle school. Yes. I love it. Okay, so that was your first one, So yeah. Yep. Yep. which was an educator, which is the the day job, Yep. Yep. That's right. but you also have creative technologists. What does that mean to you? Yeah, so that's a that's a role that I realized, you know, it means some specific things to people.

3:57 So a lot of people look at that role as being folks who might do things like projection mapping and building worlds in Unreal or working with lighting and I love all those things. ~ that role has taken a little bit more of a back seat in the last couple years. For a few years I was really leaning heavily into that side of things, ~ which is very normal for me.

4:18 Like my work and my creative life kinda go in these big cycles and waves. And so like it's that's a little bit on a a lower wave, you know, lower part of the wave cycle, I guess, right now. ~ but you know, basically I've it's it's my kind of playground for realizing like I do all these different things that are using technology in fun and creative ways, whether it is something like projection mapping or lighting or integrating that stuff into music performances, which I guess is like the natural segue into the third part there.

4:46 Yeah. and I want to go to that third part but for a second also for those not watching. There is very clearly 3D printers and other electronics projects type stuff in the background. So even as you say projection mapping, and some of us might go, I don't know what that is, we do know ~ that like there are things that are very tangible and ~ for sure. Yeah. Yeah.

5:03 familiar for those of us that are all a part of it, you know, and like i i they are three distinct things we said there, right? But there is a lot of overlap between them, and you've already started giving us some clues about those. So I wanna dig into those in a bit, Yeah, for sure. but tell us about electronic musician first.

5:18 Yeah, I mean so that I got bitten by the electronic music bug thirty years ago. You know, growing up outside Detroit, not in Detroit proper, but outside Detroit, you know, folks may or may not know that Detroit is kind of the birthplace of Tecno, Detroit Tecno. Yeah. ~ you know, it's one of the sort of like you know, massive sort of cities in terms of its effect on electronic music culture.

5:42 And so I grew up a bit of a remove from that, but like realized, you know, kinda early on in college that that was ~ something that really spoke to me. So I started DJing and then almost immediately after that started producing music as well. And so that's again, that's been a background thing that's been a steady part of my life. You know, through kind of ups and downs, like really over like the last thirty years.

6:02 I have a new record coming up soon, which is exciting. you know, I I tend to work where I I make a lot of partly finished things and then the grind is like finishing those things. I don't know if that sounds familiar to you or to any of the people listening, huh. Yeah, it's yeah. but that definitely is my is my thing. ~ so I kind of collect pools of material that feels like they sort of sit together and so I've got this new set of tracks that'll be coming out within the next week or two actually.

6:27 So that's pretty exciting. Yeah, ~ very cool. I love that. it's been I think two two years since the last record came out. So yeah, it feels good. Yeah. we will make sure that is totally is linked in the show notes and all your previous stuff as well. So Yeah, right on. okay, so ~ we are here in the context of ~ an AI podcast.

6:42 So before we dive into the way AI overlaps with your particular jobs and interests, I do wanna kinda hear like what's your backstory with AI? Are you an early adopter? Have you, you know, been trying it more recently? Is it involved in all the parts of your life, only certain parts? Like, what's your background?

6:56 Yeah, so I would I would really characterize myself as kind of a a skeptic and a holdout, which is not typical for me with technology. Okay. Yeah. I I tend to be a pretty early adopter. ~ I would say like the first couple of years, I viewed especially generative AI, like LLM, you know, based sort of generative AI, as a thing to like kinda keep at a bit of a distance.

7:21 ~ Yeah. you know, primarily I think because of the the sort of ethical and environmental concerns around it, al along with things like hallucination and and that. You know, so I would use it very occasionally. ~ you know, the the way I describe it to people is like, if I need some words that are good enough for this scenario, like I'll let it generate some words and those are probably fine.

7:39 Sure. Yeah. Yeah. You know, we've all had that like moment in our off we're like just write something to like check the box and move on with your life, right? So that had kind of been my my relationship with it. the big turning point for me really came back in like February or March. ~ So I'm the one of the faculty mentors for the high school robotics team at Pioneer as well. our one of our parent coaches is a VP for a software company and you know manages a software team.

8:07 And he was talking to me and some of the other mentors about just how radically the AI coding tools specifically were changing that industry. And I think I had a Mm-hmm. sense for it, but he really made the case that this was a C change, that there was really something there that was not there six or 12 months before.

8:25 ~ and what I realized. using AI a little bit and having and leading conversations with students about it and you know thinking about my engagement with it or my students' engagement with it ~ through one lens so to speak. but that really made me think I I kind of owe it to myself and probably to my students to some degree to take some time to really engage with these tools, understand more deeply. And ~ I work best when I can kind of frame the things that I'm doing with some, you know, some kind of put some kind of framing around it.

8:57 And so I thought, okay, I'm gonna do a summer of AI learning for myself. I'm gonna just dig in and really try to understand the tools, ~ understand what's possible, really just get myself a a new perspective. ~ and and the the big turning point for me came actually before the end of the school year, before the like summer officially started for me, ~ with a little project that I built.

9:21 And that was a that was a game changer. Yeah. So the I we've talked recently, so I'm like trying to pretend like I don't know what you're talking Yeah. about. but I think that's really fun because one of the things we're gonna be talking about is the practical ways that ~ we discover what AI does and does not make sense for, right?

9:39 Like that, especially as a skeptic. I was just talking to one of my friends yesterday talking about the fact that ~ my goal was to not just jump onto the AI bandwagon, but to really like kind of give it a skeptical, critical but open minded shake. And so sometimes when I found something I was really good at, the skeptic in me was like, make sure you're not just doing it just cause.

9:58 And it's like, no, this is actually legitimately practically helpful. And I think those voices are really valuable. Right. Yeah. Yeah. So I I I want to talk in a minute about electronic music in your students, but tell us about that experience you had, that very practical benefit you got from it. Yeah, you know, it is it is funny because there is that intersection already.

10:14 My one of my friends Josh, who I've done music with before and have known for forever, was making these little tools to convert synthesizer presets from one like synth, I think it was like an old hardware synth to like a new modern thing or something. And he was tossing little little tools at me. ~ and I was like, that's really cool.

10:32 And ~ I think that got in the back of my head. And what I realized was Okay, if these tools are good enough to do that, I'm sure they could do this other thing. And the the context in this case was ~ was for something that I and my students use all the time, which is our laser cutter. So in the engineering lab, we have a laser cutter, lets you take a 2D design file and cut it out, or you can engrave with it as well.

10:54 99% of the time, what we need to do is my students or I will design something in CAD as a 3D model. We want to cut that thing out of a piece of plywood or cardboard or whatever it might be. To turn it into a real thing. And unfortunately, the way that like the combination of our software that we use for CAD and the software we used to run the laser cutter, ~ there's just this little gap where you have to take the the 2D design file, ~ open it up, convert the line weights, convert the colors, and then save it as a new PDF so then the laser cutter software recognizes it, right?

11:29 And and what I realized was how many times have I run this cycle? Like I have run this cycle personally hundreds of times probably to cut examples and cut parts for my own projects that we're doing in the la whatever. And then of course students as well, right? Anytime they want to run the laser, they have to do this. So I was like, I bet I could make Claude do that.

11:48 And and so I had already kind of not settled on Claude, I think, by that point, but it was like in speaking with my friend Keith, ~ you know, that was the tool that he had specifically mentioned. ~ I liked Of the stances that I had seen them take, I think that set them apart a little bit from some of the other AI companies.

12:04 so I thought, let's let's give Claude a go at this. You know, I was doing this sort of in the background where students were just working on a project, and within a pretty short time I had something that was really close. I hit my usage limit, of course, because I was on the free plan, which I think everybody listening, if you use those tools, has has had that experience.

12:23 ~ I immediately was like, okay, this is the moment, bought a pro account, waited for my usage to reset. And I think it reset at two fifty. By three o'clock I was laser cutting the first thing that it had converted. And it's it was that moment where I went like, okay, the ability to build tools like that, I mean I Like I said, I wrote software.

12:45 I worked on interaction design. I did all these kinds of things, both front and back end. Like could I ~ design a tool and I gave it some clear specifications about wanting it to be HTML5? I just want it to be like super simple, run anywhere, super duper lightweight. Like don't make an app. I don't need that.

13:04 I just need super something really com like really, really easy. So could I have like learned how to use the PDF library and JavaScript and done this, that, and the other thing? Like I could. have but practically speaking there was no way I was ever going to write that tool. And but you know, but being able to do that with ~ with Claude, it wasn't even Claude Code at that point, it was just Claude, ~ really was like, okay, this feels like a game changer.

13:27 ~ So that was kind of my like dive down the rabbit hole, I guess. Was there any part of you that originally leaned towards I will give the file to Claude every time? And if so, like what what made you jump to think the way of having Claude build a tool instead? Yeah, I think a lot of that comes back to that initial conversation with Keith.

13:46 So he had specifically mentioned like part of the power of using the AI stuff is not just throwing a problem at it, but getting it to build a tool for you that solves a problem. And that that appeals to my nature anyway. Like I build all kinds of tools and design, you know, things, whether it's like a jig in the shop or whatever it is.

14:05 Like that that's very comfortable for me. And I think having that initial success was great, but also the framing around making it really super lightweight. ~ a thing that can just run in a browser that you don't need to like give to one of the platforms in general because so much of what I do of course deals with student information. And that's just a thing where we have extremely strict boundaries around that for the very best reasons, of course. So thinking about what are tools that I could build that might let me, you know, work with student data but are completely local.

14:38 It's just an HTML file that runs in my browser locally on my computer. Yeah. And so, you know, I worked on another tool. I have students fill out a getting to know you form at the beginning of the year. It, you know, outputs a Google Sheet, which is fine, but it's not the nicest thing to look at. So I wrote a little I my friend Claude and I wrote this little tool Ha ha ha ha.

14:57 that lets you just chuck a CSV file at it and it outputs a much nicer looking sort of student profile sheet where I can kind of skim Nice. That's very cool. through it and it just looks like a nicely laid out PDF instead of like reading rows in a spreadsheet. And I you know, so my hope is to use that at the beginning of this year.

15:10 And hopefully it just makes that process of learning students' names and faces and their background and all that stuff just a little easier. ~ And so I'll I'll make that available to my teaching colleagues as well in case they want to use that. Yeah. ~ I really appreciate the thoughtfulness of Knowing how to approach LLMs is one of the biggest skills that people learn as they start using it. Knowing what it's good at and what's not good at, knowing what prompts to get the sort of thing you want.

15:40 And a lot of people kind of pitch these ideas as like, you just need to learn the the top 10 prompts. And if you pay me for my prompts. And for me, it's always just been you need to learn what things it's good at and what things it's not good at, because that guides what things you want to ask it for.

15:54 In this particular example, It's much better at building a tool that converts ~ this file than it would be at converting the file itself. And we can talk about like determinant versus determinism versus non-determinism. ~ but you kind of came to that thinking pretty quickly as someone, it's not that you've been using AI really heavily for the last four years.

16:13 Right. Yeah. what was it mainly just that one conversation ~ with your friend Keith that that led you to, you know, having a greater level of understanding? Or what like got you to the point do you think? where you were capable of interacting it with a way where you are better discerning like what is this good at and how do I use it for the things it's uniquely good at.

16:33 Yeah, I mean I think there's two parallel threads that kind of that kind of come together for that. So thank you for the compliment, first off. I appreciate it. Worse. I it was ~ but no, I think there are two things really. So one is, you know, in computer science education, there's this idea of computational thinking, and problem decomposition is a really big part of that.

16:53 And so that's a core Mm. skill that we try to teach students. Like, how do you take this large problem and break it down into smaller pieces and think through it logically? You know, so that that kind of mode of thinking, I think, is something that I've been working with for a long time. ~ but I definitely go back.

17:09 Back to my, you know, my web development days, ~ you know, the last thing I did there was start our company's software testing practice. And that's like a little grandiose. It was like me and another person. You know, it's like it was a very small effort, ~ but it formalized something that we had never had a formal process before, where we would go to launch a new version of this site or web app, and we actually had formal scripts and test data and all the kinds of things that you need. I get very good at writing up steps to reproduce bugs and like being able to see.

17:39 This is the expect, you know, this is the input, this is the expected output, this is what happened in the middle. You know, is there a gap? Is that what happened? So that kind of like step by step breaking a problem down thing is pretty natural to me at this point in my career. ~ and so you're right, there's definitely a natural sort of I can think how the machine wants ~ you know, wants its input. I can I can ~ hopefully specify pretty well what I want the output to be. and then I think the thing that I've really enjoyed that Interaction design part of having a sense of what a good high contrast tool looks like in terms of accessibility and usability, Yeah. Uh-huh.

18:17 like understanding where labels should, you know, be or not be, or why this is important in the in the interaction flow or or not. ~ that part I think has has helped as well. And it's it's like muscles I haven't flexed in a while. So it's pretty fun to be able to get back into that and go like, I don't have to do what to me is often kind of the tedious part, which is to write the code. I can like think of the spec and think of the design ideas and the stuff that's important, and then I can evaluate the output and like do this iterative cycle.

18:45 And that's been that's been super exciting. That's very cool. So a lot of what I'm hearing you say is you're you're building on skills you already had, ~ but applying them in a new context. ~ in Yeah, for sure. and it's very interesting because I I again the vast majority of people who I see come to AI have this moment that we call AI psychosis where you're just absolutely obsessed and you're not getting any sleep and you're working on it all the time and every problem looks like an AI shaped, you know, problem or whatever.

19:14 And then at some point you come back out and you've learned enough through that process that you now kind of have the discernment of what it's good for and not good for. So you I I'm not saying nobody has done this, but of of the people I've talked to ~ on and off the podcast, I think you most quickly move to a point of having discernment and having judgment and stuff like that.

19:32 So it's it's very interesting thinking about what things led you there. ~ that's a great transition for me to the fact that you are a teacher. And part of what I wanted to ask you about is you are a ~ a teacher of technology to people Uh-huh. who are about to go off to college to decide what they want to study, you know, and Yeah.

19:50 they're, you know, my kids are in middle school in elementary school right now. So I'm I'm not I'm like, look, hopefully by the time they're going out of school, we'll have a little bit of a better handle on how much, you know, these things are impacting their jobs. But the kids you're working with are about to go out there.

20:03 Yeah. What has the process been like of figuring out how to talk to? And you mentioned a little bit about like you want to be able to give them good advice based on reality. So Yeah, for sure. A, like what has the process been like of trying to talk to them about their fears and their concerns about how to make judgments?

20:17 And B, has that changed as you have become more comfortable using AI in your personal and professional life? I'm gonna answer the second part first. I think the answer is definitely yes. Yeah. Okay. It's going to change like the conversations I think I'm gonna have this coming school year, which is right around the corner for me.

20:34 I think those conversations with both students and also with my colleagues, frankly, are are naturally going to change because of the experience that I've had using the the tools really heavily over the course of this summer. ~ as well as, you know, the the longer view of of thinking back. I mean, ChatGPT came out in 2022, which is not that long in the Yeah. Yeah.

20:54 grand scheme of things, but is a very long time in the world of technology, of course. ~ in the in the context of students' lives, though, somebody entering ninth grade, ChatGPT existed their entire middle school. school career and at a time where they probably didn't have the best judgment or discernment as you talked about in terms of how or when to use those tools.

21:13 And Yeah. so it's gonna be really interesting to see what those conversations look like this year. ~ The other thing though is that I, you know, I have kids as well. My kids are a little older. And they've definitely shaped my thinking a lot as well, which is that they are all like all of them effectively like extremely anti-AI, especially in terms of generative AI.

21:33 Like there's a real generational divide there. Yeah. Yeah. And, you know, it we've had some really good, really thought-provoking conversations ~ that have kind of checked my thinking, but also helped me kind of clarify my thinking in ways that I explain, you know, ~ why I might choose to use these tools or when I might choose not to use those tools or or what things I might use them for or not use them for.

21:56 You know, so the the thing that I think is has been really important in the conversations that I've led with students so far, a lot of that's been in the context of this advanced topics in programming class that I've co-taught and then also kinda like, you know, helped with basically. So one of my colleagues is one of our computer science teachers.

22:14 He's right next door. We actually co taught that class one or two years, I think. And then I think the last two years it's been he's been teaching it, but we overlap a lot because my engineering capstone class is at the same time That's effectively his capstone class, the advanced topics class. And so oftentimes we'll do like Fridays together, talk about current topics. AI cool.

22:33 is ~ always ~ an option as far as current topics because it's it's changing so rapidly. And what we've tried to do is ~ not lead students to any particular, you know, ~ conclusion, of course, about AI usage, tools, ethics, whatever, but just to get them engaged in the process of thinking about it instead of just sort of blindly using things.

22:54 You know, hopefully that's the goal of all educators, right? Is to is to open students' eyes to perspectives they hadn't considered and give them broader context and all that kind of thing. ~ but yeah, they are definitely on that cusp of, you know, either going to college or maybe even getting an internship or whatever it might be. ~ So I think we have a an obligation to talk to them about the way that the world is shaped right now.

23:17 And the way that Yeah. the world is shaped with technology and engineering and software and robotics is these are almost not an optional thing anymore for so many industries and so many companies. Yeah. and that's not a value judgment on my part. That's just trying to look objectively at how I think things are out there in industry. ~ That's the way it seems. so I think we owe it to have those kinds of honest conversations with students about what they might expect or not expect about helping them come to those their own individual conclusions about when to use or not use tools.

23:54 ~ but also to to give them some framing for their thinking. That's been a big part of what I've been working to develop for myself over the summer. And one of the things, that kind of framework of trying to help them, you know, slot their thinking into stuff as well. And so for myself, like You know, in the in the learning that I've done, whether it's formal learning through like I've taken, you know, some of the Claude anthropic training specifically, ~ you know, I've done a a a number of professional events, ~ like development events and conferences this summer.

24:24 ~ you know, thinking about AI as a thinking partner, thinking about it as a tool to do some kind of tedious work, and thinking about it as a a way to build something. Those are the three kind of main categories that I see. And Okay. I think if if I Able to help students understand when they might or might not use AI and how it fits into one of those categories, or maybe other categories categories that they come up with, I think it's gonna be a lot more powerful than just I don't want to write this essay, I'm gonna paste my prompt into whatever your tool of choice is and get an output and then paste it into that Google Doc and turn it in. You know, that fits into that sort of second category,

25:06 but ~ it's gonna be important for them to learn that that, even though it's like doing a tedious task for them, what they perceive as that it it's easy in the short term and it's hard in the long term, right? And and that is the thing, that perspective is the thing that that students of whatever age often struggle with. ~ you know, needing to put in the work now and and to build up those intellectual muscles.

25:29 Like I You know, I can't give you the experience of having written essays without you writing essays. And of Yeah, yeah. you learning how to write an essay and like make a convincing argument or do whatever. Not I don't assign a lot of essays as an engineering teacher, but you know, my colleagues who are who are working on on those kinds of prompts, you know, that would I think would be a big part of their experience.

25:49 But on the other hand If you've got a problem that is sort of AI-shaped and you can build a tool that will let you solve that thing. And here, of course, I'm thinking s very specifically in terms of the context of like a robotics team or a engineering capstone project or whatever it is, I I I think we we well, for myself at least, I feel like I have an obligation to give them guidance in terms of what tools might work best, what tools are appropriate, the the trade-offs that they're gonna give, and give them a really easy low stakes way to to take a a run through that whole loop of prompting and building and testing and iterating on on all that stuff.

26:29 It's the thinking partner one where I'm I'm just not as sure yet with students. I don't know what that's going to look like because I really don't know how it's obviously it's gonna be different for every student, but I think that's a place where they're really some of them are really stuck on like I'm not talking to a machine.

26:48 I'm not I'm not doing it. Yeah. Well, it's Yeah. it's interesting to me. I I love that you gave us kind of those three sections to work with. and it's interesting to me that the more we categorize and delineate what AI means and within AI, what me working with AI or having competence or capability with AI means, the easier it is to A, use it appropriately for each of those different contexts or those different kinds of delineations.

27:19 ~ B reason with objections and figure out which are valid and which maybe can be kind of diminished a little. and I already forget C was, but like I you have those three. And one of the things I've been thinking a lot is ~ when when Tighten talks to people and they say, do you guys use AI? The first thing I have to say is that already means two things.

27:39 One of them is do we a use AI as a part of our development process? And two is are we going to be able to integrate AI into your web application that we build for you? Right. There's already two completely separate things. And for a Totally. student who's coming, you know, into college or into the workforce, you've got at least two aspects.

27:55 And one of them is, do you understand how to use AI as a practical tool that's actually happening? And the other one is, have you learned how to have a healthy relationship with AI and your learning process? So is at bare minimum not keeping you from learning what you should be learning because you're delegating your thinking off to AI.

28:12 And people have been calling people who know how to use it as a tool, like AI native, which is fine, great. You know how the tools. But especially when I talk to teachers, I talk to the principal I had the podcast earlier. The bigger concern with a lot of teachers is not about making sure they are trained in using AI well. It's that they are trained in not using AI poorly in a way that harms their education.

28:34 And as I'm hearing you Yeah, yeah. kind of talk about that, a lot of that the I had never thought about that because I tell people all the time, use AI for the things you don't want to do. I'm not gonna give that advice to my son because I know he doesn't want to write an essay or he doesn't wanna Yeah, very different context there, right?

28:47 That's such an interesting delineation. Yeah. Yeah. I mean, I and I guess one kind of delineation there is is it that you don't want to do it and it's a rote task, but it provides value to you in your growth and your learning? Or is it a rote task that literally is just rote, like loading your piece of, you know, your CAD into whatever? D is then are we growing from doing that or Yeah. Yeah, like n are we just wasting time?

29:09 exactly. No one is learning anything. Like literally none of my students is learning anything after the first time or two. They have opened that, you know, PDF in Inkscape or Illustrator and changed line weight. Like, okay. There are vectors. They have different colors and line weights. Cool. Learning over. Now the next 99 times you Yes.

29:27 have to do that, you're not learning anything new. You're just, you know, bridging the gap between two kind of weird pieces of software. It's you're you're not gaining anything there. ~ but of course what Yeah, go I it makes me think No go ahead. ahead. No, I was just gonna say, you know, the the thing where it's really difficult though is I think like I mentioned earlier, that students by and large have not yet developed the sort of self discipline and awareness Yeah, exactly.

29:49 of you know, many of them just want the short term reward, which is reasonable. Like there are plenty of times I just want the short term reward of just I want to check this thing off my list so I can do something that I prefer, whatever that thing is. but it is really tough to ~ to help them understand, like, no doing the whatever it is might seem rote to you right now, and maybe it is, and maybe it's boring, but it's gonna lead to a better result in the end.

30:15 Yeah. It's the corniest thing I've ever thought of, but I just think of like the wax on wax off from, you know, karate kid as like what he thought he was doing was this a very rote thing. And it turned out to be a thing that he had to do the thing in order to gain the skill that was actually of interest to him down the road. And asking youths, asking 14 year olds to be responsible for the, you know, the responsible decision making about whi which is the thing that you should offload at AI and which is the thing you shouldn't, sounds really unreasonable.

30:45 Do you think that that is something that is gonna come purely from teachers saying this assignment can be used AI, this assignment can be used A AI? Or is it further back and they need to learn the skills of how to make those decisions so they can apply them universally? Like who or do we even know? Like who's teaching them these skills?

31:02 Where should it come from? Are parents responsible for this? Yeah. You know? Yeah, I mean I think I think any any like two line answer would be so reductive to just be useless, Yeah. you know what I mean? Like it's it is one of those things where whether AI is in the picture or not, that's gonna be the case I think in education.

31:20 it is it is exceptionally hard to help people understand why they're doing something that they perceive as boring, that has value if they have not figured it out themselves. But interestingly, like the group of people in my experience that have the most acceptance of that are folks who do sports or or music.

31:44 Those are the ones where you're like, you know, it's probably not that fun practicing your scales, but when you get to play this piece or, you know, perform this this ballet, or you get to, you know, take your your at bats or whatever, like the reason that you do all of that that hard work that is just repetitive and you know, building your muscle memory and building up your endurance and your grit and all those things, like that's what it leads to, right?

32:08 ~ and so there there are definitely experiences that I think we can draw on to help people understand that. The Hard part is whether they value the thing that we're preparing them for or not, I think. Right. Like the person who's like waiting in the wings to go and dance their solo is pretty personally invested in that.

32:25 The the kid who's like, really do I have to learn all of the long division rules? Maybe not so much, you know. That makes a lot of sense. as I I I want to talk about the other parts of your life as well, but I'm just so fascinated in this teaching aspect. ~ we've talked about kind of like two angles of can helping students, and one of them is ensuring that the students are not using AI in a way that would keep them from learning, which I think is pretty clear, especially to most parents. But you also mentioned the fact that I keep hearing, which is that ~ I think you would call them Gen Z and Gen Alpha, young the young Gen Zs, the older at Gen Alphas.

33:00 Are very, very, very anti-AA AI generally. It's not universal, but generally. Sure. Yeah. have you found yourself or do you anticipate finding yourself now that you've had your summer of AI, needing to actually convince them there's value to them? And if so, what's the convincing? Is it, hey, this is a tool that's helpful to you, or is it ~ the world has changed and you have to get comfortable with AI and in order to make it in the world?

33:24 And like, what does those conversations look like? Yeah, I mean I think it I think it is gonna be some of both of those things. ~ I don't have you know, I'm I'm relatively new into working with these tools. ~ I I definitely don't claim to have a fixed perspective, and I think having that in in this case would be probably difficult anyway, or problematic, because things are evolving so fast in some ways.

33:49 ~ but I think From my perspective at least, I don't feel the need to convince anyone that they should or shouldn't be doing something. I do think, like I said earlier, we have some responsibility as educators to to help them understand the way that the world is right now and where those tools are used and not used.

34:10 what benefits or trade-offs they might get by using some of those things in various situations. And I but I think it's going to be very different if you are a ~ you know, a software developer versus a graphic designer. And I think about somebody, ~ so my youngest daughter is just finishing up her degree in experience design.

34:31 And so, you know, she's spent a lot of years now learning how to use tools like Illustrator and Figma and learning how to evaluate signage and wayfinding and exhibit design, like all the things that she's doing around experience design and the idea that ~ You know, the Claude design tab is gonna replace you is like that's a that's a brutal thing. And I don't think it's true, Yeah. Yeah. It's still scary.

34:57 but it there's also parts of it that feel true Yeah. or feel pretty scary. And it has come so fast that it really reshapes your thinking about what your future might look like. And I think that's a lot of where the ~ backlash from the like really younger generation has come is like the rest of the world is kind of on fire right now.

35:17 And now the thing I thought I was going to do when I grew up is also completely changed. Like that's a lot to wrap your head around. And it's not Yeah. not really fair. But, you know, we don't we don't get to decide what's fair, unfortunately. We just have to try and help prepare kids the best that we can.

35:31 you know, so I think that's where a lot of the the really hard conversations are probably going to come. It's probably going to be some combination of helping them understand the reality of the world, helping them form their opinions, and then helping them evaluate if there are tools that are going to be useful for them in their personal learning or their personal sort of practice, their set of tools that they're comfortable with and so on.

35:53 I I think it's going to be a really interesting year. You know, check back with me in a year, I guess, and we'll see what this school year looked Yeah. Let's do it. like. Cause I really I ~ yeah these are all open questions for me really. And I really appreciate that you keep framing things. And I I even said, like, are you gonna talk them into it as if them using AI is better than not using AI, which I don't believe. But them getting jobs is something I want, them having opportunities, something I want.

36:19 And I really appreciate how consistently attached you are to the framing of ~ give them the resources and the guidance and the safe space to figure out what they believe, to figure out their own opinions, to give them tools to choose to use or not, right? ~ I love that because I don't know if that's always how people look at high schoolers.

36:38 I think it sounds more common for college, right? Like you're teaching Sure. them how to think or whatever, but it's very cool to think about it from from even giving high schoolers the honor or respect to believe that they have independence and they have agency in their own lives. So thank you. Yeah, for sure. I mean I mean I I I say this coming from a position like of privilege that I really recognize, which is the students that are in my classes Mostly are students that have elected to be there and are taking my class.

37:04 Like I think I would have a very different opinion if I was a ninth grade biology teacher or a tenth grade history teacher. Like everybody's gonna take those classes whether they like it or not. You know, the students are in my engineering program by and large have opted to to be there unless they somehow got randomly placed in that first year class, which I've had and have had, you know, all kinds of different mixed results.

37:23 Some people really fall in love with it and some people are like, I will never do this ever. So let me get through this class and we'll be on with our lives. ~ You just never know. But no, I mean I think that that definitely plays into it, right? Is when you have people who are invested in w in what they're doing, those kind of conversations and that kind of you know, I'm not sure how you said it exactly, but that kind of level of ~ you know, respect or independence I think naturally comes along with that.

37:49 Or hopefully it does at least. Yeah, I love that. ~ I could talk to you about this topic the whole time, but we're so close to time and I haven't got to talk about your other roles. So I I want I think That's all right. the the sexiest and most fun is electronic musician. And Yeah. I I do have to be honest with you.

38:02 You and I met in a context of music, and I have loved your electronic music since we first met. I don't know how many decades ago it was when we first met. wow. Yeah. Pro yeah. Yeah. And so I would love to hear because electronic music is electronic, right? And so like you're much closer to AI than piano or guitar or something like that, in in some theories at least. ~ and I imagine, you know, from some of the things that I know about your work that there is way pre-AI generative work happening.

38:30 Even just as simple as like an arpeggiator in some ways is, you know, generative-ish, right? Sure. Yeah. as you think about being a creative, obviously you're not, you know, it's not your day job, but I know that there a lot of that fear of we're getting replaced. I mean, there's there's entire musical artists and the websites devoted to Make all the electronic music you want exactly the way you want it so you don't have to worry about, you know, human artists creating it the imperfectly.

38:57 Like what has it felt like ~ as a creator? and have you been actually interacting with AI in your creation process at all? Yeah, it's a it's a great question. And again, I have a different kind of privileged position there because it's not my day job, because I don't rely on any income from that. I'm sure over the course of my like career it's been a net negative in terms of like the amount of gear and software licenses versus like the amount Bike equipment and exactly.

39:26 I've made off of releases or shows or whatever. Uh-huh. Like it's probably a net negative. But ~ Yeah, you know, it's you know, one of the things that I'd heard quite a while ago is like I don't want AI because I want it to write art, ~ you know, to write music and make art. Like I want it to do the boring stuff in my life.

39:45 So I have time to, you know, make music and make art and that stuff. Exactly. Yep. And so for me, that's largely been kind of where it's at. Now, there is a lot of drudgery around making art still. I I I am going to need to update my website when I, you know, put up this new release and I'm gonna need to prep these tracks for Bandcamp and for the digital distribution services and make different you know, versions of the cover art and all that stuff.

40:11 I don't have any specific plans to use AI for that, but I could see a world in which some of the business like like management kind of stuff that gets wrapped around that could be stuff that I offload. ~ I don't have any specific plans. ~ And for me, in a lot of ways the generative stuff has always been kind of intellectually interesting, but never something that I've decided I want to pursue myself.

40:38 Like I there's a there's like I said, I respect it and I think it's interesting, but to me the joy of like having a minute to sit down in my studio or with a laptop or a piece of gear or wherever the wherever it happens to be and just like playing with it and treating sound like it's really plastic and moldable and you know, trying out rhythm stuff.

40:59 Like that's the joy for me. I don't it's just not a place where I feel like I want ~ I don't I don't want outside opinions yet. You know what I mean? Like I wanna I wanna Yeah. Yeah, totally. be I mean not to say that I'm not influenced like everybody of course is influenced by stuff, but I want what comes out to be stuff that like my personal human brain has filtered through my own sensibilities and like get gotten out into the world in some way.

41:24 one thing that I'm I'm curious to check in with you in a year about as well is ~ I don't know, but I can imagine as you're talking about three and one of the kind of through a lines of you got these these the the the laser printing and then and y you got the three printing that you're doing on your own. And then I often think of ~ electronic musicians and looking at patch bays and all these things as being very similar.

41:43 yeah, yeah. for sure. Yeah. Yeah, It's just lots of these and so I I wonder absolutely. Yeah. if if a year down the road you don't end up saying You know, there was a synth that you you are you mentioned ~ your business your your musical partner doing that already, right? Like that he that he translates and that's that's For sure. Yeah. Yeah.

41:57 again where I'm just like, I don't want AI to write music, but are there parts of the music creation process that it could allow me to do things I couldn't have done before? And I don't have the answer there because I'm I'm not even doing music right now, Right, right, right. Yeah. but I'm curious whether next time you sit down to do something and you're like, ~ this again, you know, or and maybe yeah. Yeah, yeah.

42:15 there isn't anything like that. And it's just but yeah, yeah. Well, no, I mean that to me comes to that sort of third category of like, can I get AI to help me build some tools? And then that part becomes very interesting. Uh-huh. Yeah. Cause that is that is a big part of where ~ you know, some of the creative technologist stuff that I've done has been so ~ a group of friends and I ~ formed this thing called North Coast Modular Collective. So dealing with Euroract Modular Synths, ~ specifically around a lot of like DIY and education focused stuff.

42:45 We've we've kind of taken that's kind of been on the back burner for a couple of years as people have gotten, you know, busy with other parts of their lives basically, but like, you know I don't do a lot of graphic design, but I do a little graphic design and it's really fun to design a front panel for this synth. But then there is some tediousness of like, we've got to figure out how to make this thing do this thing.

43:05 I'm like looking back on some of the development work that we did, like we were working with a lot of the you know, basically Arduino compatible things ~ to do some custom synth modules and that development process was fun. I got to do some pair programming with friends and we, you know, front panel design and prototyping and that stuff's all fun.

43:25 But I look back now and I'm like, if I pulled that repo into Claude code and was like, what are three or four things we could do to make this better, faster, more efficient, that might be interesting. And the danger is like there are a million of those ideas that could go out there. But that sort of like let me make a tool.

43:44 And I mean, and you know, Ableton Live, the software that I use for making music, just came out with this thing called extensions, which is not a plugin that you could like download and just load on an instrument or audio track, but it's literally changing how the program works. So it's here's this API, Mm.

44:01 you can fire up a little like Node.js server and like have this live connection to the thing and write a thing, and you right-click and a contextual menu pops up and your tool is there now. And then when you're ready with it, you just like bundle it up and now it can run inside your main software that you use for making music and now it's got some new capability.

44:20 And so it's funny because I I like took a little like toe over the line there huh. Yeah. and was like so and and this was a place where I I threw Claude at this and it Like you were talking about like developing the instincts of how and when to use AI. And so I realized pretty quickly, like, ~ one thing I need to start doing is not saying help me build a tool that does this, because it will always be super happy to help you build a tool to do whatever.

44:48 But I think that next thing is like, is there already a tool that does this? Do the research task of finding out Yeah. like here's the problem I'm trying to solve or here's the thing I want, the capability I want. what stack of tack or tool or plugin or whatever is already out there that does that thing that I just haven't heard of because I haven't looked into it and like what would be the best one.

45:09 So I kind of threw A research task at Claude to be like, okay, this new Ableton extension thing, what are the really highly rated ones that exist out there that people have made? What does the development path look like? What would be some good ideas based on these couple of things that I'm personally interested in? Like in ways that I use Ableton Live and the related tools. And it gave me a pretty decent report after I did some back and forth. And I was like, okay.

45:36 That is not a tool. Like, that's not a project I'm taking on right now. Like it would be interesting and fun, but I am not like I have no shortage of projects that I could be doing in my life. I have three projects in my garage workshop right now that need to be finished that are physical. Like I need to, you know, finish building this thing.

45:52 ~ I am not going down this particular path right now, but I think that's as useful as anything, like knowing. what the problem space is before you just invest a bunch of time into it is is as useful as anything else, or more maybe more useful since time is such a limited quantity, right? So yeah. Good.

46:11 ~ I will link this in the show notes, but I gave a talk at Laracon US this year talking about how a lot of people who treat ~ Claude code as or its equivalents as a code generation machine means we are no longer intellectually engaged with what we're doing, which means we lose out on the opportunities to learn.

46:28 So I gave a talk about like what it looks like to work together. I wouldn't say as a necessarily as a thought partner, but to work together is something where you're learning as you go. And I love that your idea of like First of all, you learned a lot out of that. And just replacing every single time you are saying, go build this thing with with something that engages you more.

46:48 Go research this thing, go teach me about this thing or whatever. I Yeah. feel like is gonna be a healthier long term thing. So we're not just slowly offloading all of our, you know, our all of our mental effort off to something else and suddenly we're just not engaged at all. For sure. And I mean, for me, you know, tying back to one of the things we talked about a minute ago, I am ~ So I've been primarily using Claude and Claude Code and Claude Design, but I've also engaged ~ with some of the Gemini stuff.

47:16 I've played with anti-gravity a little bit and what is now Notebook, you know, used to be Notebook LM from Google, as well as some education-specific AI tools like Magic School. And for me personally, coming into this summer, I knew I was about to take a training using these six-axis robotic arms. Like this is part of this new advanced manufacturing course we're going to teaching.

47:35 It's really exciting. And I'd done plenty of mobile robotics in education, but never any robotic. ARM stuff. And I was like, I need to learn some things about this. So I ended up going to notebook and I think I didn't even necessarily paste in sources. I think I was like, find 10 high-quality sources on how robot arm kinematics work.

47:55 Like that's the system of how arms move through space. I don't like I know a few words, but I don't really know this stuff. Help me learn some more about this. And And then I added on a couple of additional sources, I think, based, you know, after I saw the list that it created or or pulled. And that was honestly like one of the coolest learning experiences that I had over the summer was I went, all right, make me this five-minute podcast from it.

48:22 And I did. And it's like there are edges. There are places where it's like, Mm-hmm. ~ that's a little that's a little awkward listening to these two robot voices Yeah. talk to each other. Like they're kind of that like little bit of the ~ you know, a little bit of the gosh, what's the ~ uncanny The uncanny valley or I was like creepy valley? That's not right.

48:39 What is the word? The there's a little of the uncanny valley thing Yeah. huh. and there's some tells for sure and stuff, but I mean legitimately I learned some things about how robot arm kinematics work from listening to that That's so cool. five minute podcast. And then I made it how, you know, make the 20-minute version.

48:53 I learned some more things. But the thing the thing that I'm keep tossing back and forth in my head is like, if I give the five minute podcast to high schoolers in this new robotics engineering class that I'm that I'm running this year, are they going to enjoy it? Are they going to like be freaked out by it?

49:11 Are they going to ignore the learning that they might have because it was AI generated? Are they going to question the content and like Yeah. whether or not it's hallucinating? Even though it's, you know, it's all sort of grounded in an external ~ you know, documents, it's it's it's not it's not just like make up some things about robot arnkant. It's like here are Right.

49:30 these sources from Stanford and like all the other all the other, you know, you know, s you know, things that it was based on. But that's like a really interesting one. And we're gonna hit that like really quickly within the next month. Yeah. ~ the the sort of ~ the pledge that I've made to myself is that ~ Our district, I think, has lagged a little bit in terms of having a coherent, consistent policy or like direction around AI usage, which is a little disappointing.

49:57 But for myself, my kind of pledge to myself is like if I'm using anything AI generated with students, I'm gonna disclose it and I'm gonna let them know why I made it and why I think it's gonna be useful. But the first couple of times I have no idea what that's gonna be like. It's it's really interesting. Like I I'm kind of like eager and also a little bit, you know, there's some trepidation there.

50:18 Terrified. Yeah, trepidation, yeah. Yeah, a little bit. Like just to to see how that to see how that reaction is gonna go. Cause I just don't know. I'm sure I'm going to learn so much over the course Yeah. ~ I cannot wait to hear about it. of this school year, basically. Yeah. ~ I'm less disappointed about the fact that we're out of time now that I'm like, well, obviously I just have to and I I'm sorry, but you're gonna you're coming back on the podcast at the either at the end Okay.

50:38 of the school year or even maybe maybe just in December, to just kind of like weigh in for us about how things went. Sure. Yeah, for sure. So I'm so sorry to cut us short, but we you know, we still have to get to our speed round. ~ you are our first ever speed round, so thank you for your willingness to do this.

50:51 ~ so we're gonna jump over to those. So I'm going to sh Share my screen and hopefully you see a little timer here. Yeah, I got it. Yep, I got it. You see this? Okay. So for the listeners, we're gonna end every single episode this season with a speed round of five questions. ~ and those five questions, the guests have a minute or less to answer, and we'll just kind of see how this goes.

51:13 So your first question is what is your AI tool stack of choice? And that could be individual AI agent or the whole stack the whole way up the chain. Yeah, for sure. for myself, the tool that I've really spent the most time working with is Claude. ~ ~ And a lot of organiz a lot of things organized into projects.

51:33 so chats are definitely organized by project with some supporting documents. ~ using Claude Code a lot more to build tools. ~ I've been building a number of simulations that I'm going to use with engineering students, and that's been super exciting. I've committed more code to GitHub in the last three months than I have probably ever.

51:50 Yeah. but then sort of supplementing that with ~ with Gemini, ~ specifically Notebook LM, and then some education specifics. specific tools like Magic School, which is a tool that our district has has paid for and has a a district wide license for. So that allows us to do some things that we couldn't do with the sort of third party tools, so to speak.

52:11 I could ask you questions and these, but I'm not allowed to. So we're gonna get to the next one. Have you ever cussed out an AI? I don't know if I've ever cussed one out, but I came very close the other day to posting one of those snarky things on threads. Like if I hear the word load bearing again, I might throw this laptop. I might throw Belton suspenders. Yeah. Uhhuh.

52:32 this laptop into the street. and ~ laptops and gates were the two things or load bearing and gates were the two things that it kept ~ getting stuck on and I I don't know if it's a fable or a Opus V thing rather or or what. ~ hopefully load bearing gets just wiped from the face of the earth at some point.

52:50 Yep, I'm with you on that one. All right, your next question. ~ what is one pro usage of AI that more people should be using? Well, for myself, ~ this is the thing that really sparked its usage for me, which is using it to build tools to solve the problems that you specifically have. And that could be for you personally, it could be for whatever group of people you serve.

53:13 For me, it's students. ~ so building little tools that help me, help other educators, help students do a particular thing as opposed to going and like throwing a problem at an AI tool and Yeah. letting it solve it for you, letting it build a tool for you. Love that. ~ next question. Is AI going to ruin the world?

53:36 Well, I I don't know if it's going to ruin the world, ~ but there are certainly pieces of it that I think are ruining the world. ~ and so You know, obviously the maybe the big obvious one is is data centers and the environmental impacts that they have. ~ my sincere hope is that the AI bubble bursts a bit, frankly, and that we get companies a little bit more restrained in terms of how they are expanding and kind of trying to eat the world.

54:03 ~ my hope also is that we get a better technology stack that allows people to run some of these things locally instead of having Mm-hmm. to rely on these massive ~ data centers. for all of that compute and kind of spin things up locally when they need it as opposed to just having these sort of always-on facilities.

54:21 I don't claim to be an expert, but that's my that's my sort of naive take. I'm with you on that. Okay. Well, if anybody's interested in having those conversations or learning more about you and what you do, how can they stay up to date on what you're working on? Where are you on the internet? Yeah, appropriately enough, I've been using Claude pretty heavily over the course of this summer to work on building a new version of my site.

54:45 It may or may not be up by the time this episode posts, but it's billvanloo.com. and ~ you know, so you can kinda see a bit more about my work there. ~ I'm somewhat active on Instagram, so Instagram.com slash Bill Van loo and LinkedIn as well, which I don't remember what the URL is off the top of my head, but I'm sure it'll be in the show notes, Put it all in the show notes. Yeah, right? It'll be all in the show notes.

55:04 exactly. Yes. Amazing. Yeah. Yeah, those are the those are sort of the big three. Bill, I am so grateful to you for hanging out today. I'm so grateful for you to being the first guest of the new season as we figured out some of the friction before the show and during. So ~ Ha ha ha. and I'm also just grateful for your friendship and the way you're leading and shepherding kids ~ at at a really important time in their lives.

55:23 So thank you so much for hanging out. I really appreciate it, Matt. And it's been so fun to reconnect and I really appreciate the work that you're doing here. I try to always find that what feels like a like a reasonable sort of middle ground, ~ as opposed to staking out some end of the spectrum. And so I really appreciate the perspective that you're that you're giving and and bringing guests on to share as well.

55:45 Well, thank you so much. And thank you for the rest of you for hanging out with us. We will see you all next time. All right, thanks again.