Transcript
Neale 00:10
Hi, I am Neale and welcome to the next episode of another Bright Spark Podcast. Our guest today is a founder of Tech Explorations. It’s Peter Delmaris. Peter, welcome. Welcome to the podcast.
Peter 00:21
Good to meet you, Neale. Thanks for having me on your podcast. Very excited.
Neale 00:25
Broadcasting from Australia as well. Where are you based, Peter?
Peter 00:28
Yeah, so I’m in Sydney right now. I’m about, say, 40 minutes out of Sydney CBD up on a mountain. It’s winter here down under. Rainy cold day, so it’s going to be in and talking about technology.
Neale 00:44
Yeah, fantastic. Thanks for joining us. So, Tech Explorations, you set up in 2013. It’s a YouTube platform. Why did you decide to set up the company?
Peter 00:57
Well, yeah, we got back like over 10 years now. I can’t believe how time has gone so quickly. So the idea behind Tech Explorations was basically to scratch in each gig. I like to learn about stuff. But in addition to that, I’m also an educator. So I spent about 15 years prior to studying Tech Explorations as a lecturer at various universities here in Sydney. At some point, I got tired of lecturing the same stuff to my students. Technology was just going so quickly past me. I felt like I was falling behind. And like academia is a great place to be if you are lucky enough to have, I guess, a research lab so you can be at the forefront. But for the rest of us, it’s like, okay, I’m doing the same subject for the sixth time.
Neale 01:56
No one can hear you sort of trapped in a time bubble type thing.
Peter 01:59
Exactly. It was pretty, pretty annoying. I felt like I was trapped, as he said. So I decided to start Tech Explorations at about the same time as, you know, the makeup movement started by getting traction. I got my first Arduino a few months prior to actually deciding to leave academia and start an online education career. And I wanted to combine the two education and technology exploration. So I created Tech Explorations as the vehicle. So what Tech from the beginning is just I decided to record my lectures and my lessons instead of doing them live. The reason for that is twofold. Basically, one, I decided to never repeat myself ever again after the university experience. It’s just one set of you could play, replay, rewind, whatever you want, but I’m not going to say the same thing again. And I think we should say as well, isn’t it? It’s so broad, the subject, you know, yeah. If I want to continue exploring, I can’t go back, you know, covering the same stuff again. So, so it was a big incentive. And the second one was reach, obviously, studying from scratch. I could not rely as a business to do the, the coaching thing that it was quite common back then, where you’d say, I’m a coach, I’m going to teach you mathematics, electronic, whatever it is, science for your school exams. So yeah, that was not viable for me. If I did that, I would have to repeat the same stuff again and again. So violates rule number one. So that’s how it started. Now, back then as well, I was lucky in the, in terms of timing.
Peter 04:00
So timing was good at about that time, as I was exploring ways to basically promote my content. It was, of course, YouTube. The problem with YouTube was that the, the pay was kind of iffy, like I wasn’t sure if as a YouTuber, I’d be able to pay the bills. But other online education platforms starting popping up. So there’s quite a lot of competition, still is, but one kind of stood out and it was Udemy. So I was one of the first people to actually join Udemy and go on it full time. So that’s why I said, I was lucky with the timing, but I think I had a good formula in terms of the kind of content that I put out, kind of resonated with makers, which was also good timing. It’s the beginning of the maker revolution, as I call it. So I have to admit, like it was not planned, but when I look at it back now, like look 15 years back, I feel very lucky for being at the right place, at the right time. And just to put one more thing that happened at the right time, like I’ve got very sick of university. There were a few things that happened there. I’m out. And there’s other things happened that are good, good timing. So, yeah, it worked out good.
Neale 05:28
Our engineers, a few engineers certainly watch your channel and you’ve got quite a decent following. You know, they use it as like a reference point and, you know, they really enjoyed it, specifically our M.D. Neil, he was the one who asked us to reach out to chat to you. What’s your community like? I know you said you like to record things and do it. Is it quite interactive? What’s the feedback been like? They found it useful?
Peter 05:54
Yeah. Yeah. So it’s complicated. So I’ve got a kind of my community is a little segmented, if that’s the right word, because I’ve got you to me where it’s a large community there for about I think lost kind of about one hundred and forty thousand students. I’ve got my own community at Tech Expirations, which is about depending on how you counted, say about twenty thousand people. And then there’s on YouTube, which I think at last counts, maybe sixteen and a half to seventy thousand people. It’s like I’m starting to put some more emphasis there now. So these are three essentially separated communities. I interact with all of them, but due to time constraints and in order to keep my sanity under control, I’m careful how to do that. Of course. In the beginning, when I was the only community that I really had was the the Gidimi community. It was growing, especially the first three or four years. I spent about maybe three hours every single day interacting with students, replying to questions, you know, help them along. But eventually I reached a burnout state state where I would I would feel kind of stressed and I would get cold sweat thinking that I’d have to go to reply to students. So it’s a lot of pressure.
Neale 07:27
You feel obliged. They’ve come to part of your community. You set this up. So I can well imagine you feel an obligation to look after them as a duty of care there. So it’s understandable. But you have to manage your time as well, to get that balance right.
Peter 07:43
So the problem was not really talking to students, but at the same time, I had to do all these other things like maintain courses and create new courses. And it was always kind of a roll of the dice every morning where I’d go to, I would clear my student questions first in order to allow uninterrupted time for the rest of the day. And I would never know what might pop up. A question could be answered in say two or three minutes. But I was always afraid of those other questions that could reveal some problem with my content, like an incorrect solution to a question that would require potentially rerecording an entire lecture, which means two days gone. So it would make things very difficult to manage. And I’ve got to say that in the early days, I was a little, you know, not really knowing what I was exactly doing. So I had, I was making a lot of mistakes, many of them technical, like audio, visual, yeah, like audio, visual problems. Like, I remember one of my first videos, entire course, I recorded it on a table with no mouse pad. And the, the microphone was mounted like hard mounted on the table. So every time I moved the mouse, the sound would just propagate. I didn’t realize how bad it was. So stuff like that. And students initially were kind of used to crappy audio from other instructors, but the students were becoming more used to high quality stuff. They started complaining. So that took like a month to record all those lectures. How many emails were you getting then?
Neale 09:25
You said you would, you’d spend a lot of time, you know, going through replies to students problems. What sort of numbers were you looking at?
Peter 09:30
So I was looking at statistics at the end of each year. So you didn’t, it would provide some statistics and it was like two, 3000 questions in a year that I would answer. So each question would, again, could be either maybe a quick, yeah, good job. I’m doing the right thing. Or I have to redo this whole section all over again because I totally messed up. So anyway, back to now. So I’ve got a more segmented approach now and more controlled approach. So I’ve got my own community where students that want to pay a little bit extra can join that community and have basically one-on-one interaction with me and other members of the community. So the kind of segments, everyone else goes into what I’ve got a maker club. The maker club is free for everyone. It’s kind of a more open-ended type conversations. We talk about technology, we talk about our own projects, but we don’t expect that PETA is going to solve every single question and answer every single question that comes in. So usually, uh, other members will chip in and help. And it’s just amazing because like, uh, I, when someone joins, I explained through an automation automatic message that, uh, most members of the community are actually much knowledgeable in the kind of thing that they will be asking because we’ve got so many different specialists.
Peter 10:58
So we’ve got people, for example, that are really, really good programmers and I’m a crappie programmer, like I’m okay, I guess, but not when I’m comparing myself to some particular strength. Like I see myself as a jack of all trades. So I’m really passionate about the educational aspect, but I’m good enough to be able to put something relatively complicated into simple words for a beginner to understand, and of course, take the beginner into intermediate and, um, um, in some topics, more advanced stages, but there’s just so many other people that have been doing specialized work on, on, on things that I could never imagine, like today, I was talking to someone who is, um, very good at on, um, uh, Colonel cross compilation topics on the Raspberry Pi. So he will, he will compile or recompile some sort of Linux operating system, uh, with some kind of peculiar kind of drivers for the Raspberry Pi on a windows 11 computer, which to me, like, I don’t know, where do we start, uh, to learn stuff like that, but, um, I was talking to him today. Some, someone asked a question. I knew he would be able to answer it. So I connected them together and, uh, the answer was forthcoming within a few minutes. Now, but it was a great position to be in as well to have that community, to have those people that you can lean on.
Neale 12:27
It’s interesting what turns engineers on, isn’t it? Because I mentioned before, we’ve got, um, the guys who work for us, a T shape. So they just like you have a wide knowledge of a lot of things, but then have specialisms or that’s high powered hardware or FPGA or whatever that may be. But for certain people, they just, they just get, um, obsessed is the wrong word, but they get really into the specialism and sort of deep dive into it. And it’s, it’s, it’s a passion more than anything else. They, they want to know how it works and what makes it tick. When did you, I mean, you’re, you’re obviously an engineer by heart. When did you first, uh, when did you first get the bug? When did you start wanting to know how things work?
Peter 13:10
Oh, okay. Uh, yeah, I was a kid, obviously. Like I was, um, uh, for as long as I remember myself, I used to be interested in technology, so I grew up in the eighties, um, when I actually became conscious. It was before that memory, uh, but great time as well with the spectrum and Amstrad computers and home computers. Yeah, um, I started becoming interested in all the technical stuff, like, um, the first digital watch with an LCD screen kind of fascinated me. Um, so, uh, the HS, the HS video recorders, um, cassette deck players, uh, audio systems, the computers came a little bit later. So in the house, I would, um, kind of open up things to see what’s inside. Um, I was very careful not to loosen the screws because no one knew, uh, things were so expensive back then. Like to open up, um, VHS video recorder was like the jewel of the house. Like, uh, it would be like a three or four salaries.
Peter 14:23
So I was very careful, but I had to know what’s inside. So I would carefully open it up. I had a screwdriver. It was my most precious tool and have a look inside, touch nothing. Just look later on computers came though. That was the big one. Like once computers came, uh, my first computer was an Apple two E, which I purchased, I believe in, um, or my, my dad purchased that for me in, uh, 1985 or six, something like that. Uh, then programming came, um, you know, back then computers, uh, had an open architecture, it was just so easy to open up the lid, have a look at all the slots. Imagine what else can go in. I would hack the, the speaker to the speaker and the Apple two E was kind of dingy, so I would put a bigger one in. Um, and then programming, uh, programming was a revelation. So, uh, back then you didn’t really have a fast way to load a program. My Apple two E didn’t come with any kind of media, like a disc drive or cassette type, so everything had to be typed in. Oh, sorry, it did have a disk, it did have a disk, but programs back then were not really circulating in an easy, like you couldn’t download it from GitHub, so you’d have to type it in.
Neale 15:45
Yeah, yeah, we had an Amstrad, that was the first computer in our house, and it came with a book and you could literally type a program like Pong or some really basic games. Remember, my dad’s sitting there for hours, just typing over two fingers, it’s like this.
Peter 16:02
You’d go to shops, like bookstores or convenience store, and we’d buy a magazine that contained computer code, like printed on the magazine pages, so I did that for a long time, and that’s how I learned how to type, and I learned basic, that was my first language. So yeah, after that, it was quite, for me, it was obvious that I’d go into engineering of one kind of the other, and that’s what I did.
Neale 16:37
And why did you choose the academic route then, rather than sort of go into industry?
Peter 16:43
Yeah, I guess, yeah, so again, there was lack of planning, I guess, I went, so I did my engineering, then I thought, I don’t know enough, I need more, so I went for a master’s degree, and there was, in a way, a big mistake. In the sense that once you decide to do postgraduate degrees, then kind of you’re with one foot in academia, it really depends on your environment. So I was lucky because I studied my master’s at the University of Technology in Sydney, and I had a pretty good team there, professors and my colleagues were really nice people, and that kind of influenced me to do two things, one was to continue doing postgraduate degrees, so then I did another one, and then I did a PhD, because I was working with, back in 2001, I think, with what’s called intelligent agents, so these are little programs that back then, they were developed by IBM, and the idea was that the program with its state, so its variables, its data, whatever it might contain, would package itself, and it, as an entity, would jump to another computer and then unpack itself and continue its operation.
Peter 18:13
So I did that for my master’s, and it kind of was very exciting for me, because there was very interesting research, so then I decided to stay and do more stuff like that, and you know, once you’re in a research degree, then the logical next step is to teach, is how you pay for your fees. I need to pay this off now, where’s the nearest job? Yeah, so you can do like two hours of tutorials and get like a couple of hundred dollars, and you know, that’s enough of the day, yeah. That’s how it started, when it got a lot worse, I’ve heard. If you were,
Neale 18:51
So you’ve taught, so you must have offered student advice, what, if you were, if there’s young engineers watching this, or people who’ve, you know, got a passion for it, they want to sort of learn more, would you prescribe, go another educational route, would you suggest a more sort of hands-on approach, like an apprenticeship, would you be your, what would you be your advice?
Peter 19:11
Yeah, look, today things are very, very different because of the internet and online education. So today, for people that are interested in engineering, I’d say you definitely get your bachelor’s degree. It could be a four or even a five-year bachelor’s degree. You try to keep it as short as possible, but very focused. So the first couple of years, make sure you do as many subjects as you can in university just to get exposure, unless you already know exactly what you want to do. In my experience, most students want to do engineering. They know they’re good at something, but they don’t have a good picture of what is available to them. So the idea is the first couple of years. Good experiences. Talk to people, participate in groups online, offline. Like if your university has got some sort of makers club, make sure you join that club because others like you will be joining, you’ll be able to talk to them. Talk to your professors. Take advantage of opportunities to travel. So if you can go to an exciting conference, whether it’s an industry conference or an academic conference, industry would be better actually. But try to travel. If your university somehow can fund that travel, it is possible. If you write a paper, you’ve got some good topics to present from some sort of original research can, but no, depends on the university. So spend the first few years exploring, right? That’s important. Try before your mind. Exactly. Now, once you know what you want to do, focus. The technological landscape is so broad that you either going to end up with, as a generalist, there is a lot of value in that. But I think, you know, young minds are very good, but some are very good at focusing. So what you said about obsession earlier, I think it’s the first few years of being exposed into something new that is exciting, that gives you that obsession. And I find that being younger facilitates that better than being older, because once you get a little older, then your mind can start broadening its scope, that’s when you become more of a, not a generalist, but someone who understands more and how things connect.
Peter 21:40
But I think it’s important to just get the depth in one technology. So it could be, you know, signal processing, it could be robotics, it could be information technology of some sort, programming, of course, but I get that focus. So once you have reached that determination that you are very, very interesting in something specific, then look at where you can get the most exposure to whatever you’ve chosen. That could be a company, you know, many companies are far ahead in the game in particular technologies than academic institutions. So try to work in any capacity, even free in such a place that you identify, it could be a university. I’m not saying like I’m just agnostic. Now the point is that if you know what you’re looking for, whether it’s a university or a private institution or a public university, it doesn’t really matter. It could even be an online course. So just be open-minded in order to not miss out on opportunities that otherwise would be possible. And that is again, that is something that I think is specific to the times that we live now, because 20 years ago when we didn’t really have these opportunities, especially in terms of internet and communication and online courses and such opportunities, you would have to go to a very good university or maybe a handful of companies. But today things are a lot more complicated. So you need to use artificial intelligence, you need to use search engines very well and your network of friends and people that know that stuff. So yeah, a little bit convoluted as an advice, but just to recap, broad spectrum in the beginning and then narrow spectrum as the next step.
Neale 23:39
You just mentioned AI, we talk about quite a lot on this podcast. How have you found its impact on the industry so far? So we’ve talked to video game designers, we’ve talked to copywriters, they seem to be using it as a beneficial tool to sort of help the creative process more than anything else. Where do you see its place in electronic engineering?
Peter 24:03
So obviously there’s the good, the bad and the ugly. The productivity gains, the creativity gains that come from a responsible use of AI is just outside the chart. It’s just unbelievable. For people that know their craft, so for people that are already engineers, that know how to make things, how to design machines, software, how to think in an engineering way, AI is good. It is a capability, productivity multiplier like we’ve never seen before. for students. So people that are now training to become engineers, you got to be very, very careful because I’ve seen it. I’m also like in the academic board of a couple of universities here in Sydney. And this is one discussion that we have quite frequently as we are trying to deal with how do we allow our students to use artificial intelligence without crippling that ability to think for themselves. So it’s like in our generation, we kind of had something similar when Google and search engines came out because we were able to now not go to the university library to find stuff. And we had a library in the pocket.
Peter 25:42
We were able to access high quality content much quicker. So that kind of accelerated our research time. I guess as a researcher myself, I was able to now to find what I was looking at in a couple of minutes and I didn’t have to walk through to the library. Well, yeah, exactly. It’s an hour walk or and then you got to walk around the library, find a book, open it, write it down. And the book was lent to someone else. It’s just horrible. I don’t know how people did anything back then. So, but what we experienced compared to what students today, you know, have to deal with is just an order of magnitude smaller in terms of impact because like it saved us time back then search engines saved us time. Today, artificial intelligence um, act says your, your second brain, which in many things is smarter than you and it knows more than you. So we have to, uh, we have to be very careful not to make young people dependent on artificial intelligence, uh, in a way that it substitutes their ability to think. So whatever the question is, I’m going to ask ChatGPT.
Peter 27:02
So that’s the ugly case. So the good is for anyone who is already a capable, not just engineer, capable marketer, uh, capable professional in whatever capacity, it’s a multiplier of capability. Uh, and, um, unfortunately for others, it is going to be, uh, it’s got to be very dangerous. I can see it like, I can see situations on it. Do you think? Yeah, people rely and they don’t train the muscle. Okay. The brain is a muscle. If you don’t train it, um, essentially you lose it and you become dependent on the AI. So we like, that’s a whole different conversation. And as, as an educator, I think about it a lot and there are some strategies that we can use. Um, but this is a whole different, I don’t know if you want to get into this, but yeah, no, no, no way.
Neale 27:58
I mean, it’s, it’s a, it’s an open conversation. I think it’s really interesting. I think that I suppose that if you think of you’re an engineer, um, you need to obviously sort of learn the language or, or learn what your specialist name is in, and then you have to make mistakes. And that’s the only way that you learn. You don’t learn a lot from, you know, achievement, but if you’re reliant on, you know, opening up chat GPT or whatever the platform is and saying, I need this code to be able to, you know, scrape this data or be able to put it into this processor or, or whether that is an equator for you. And then if it doesn’t work for whatever reason, you can’t diagnose where the, where the issues are because you’ve not written in the first place.
Peter 28:43
Exactly. It’s like you lost. Speaking about, for example, technologies like the Arduino. When I started using chatGPT, it was 3.5 a couple of years ago, it was okay. It would create some code, like definitely a variation of blink. It would do PWM. It would do stuff that were quite okay. But it was at an elementary stage. Now I can get it to write code that it would use ESP32, connect to a Wi-Fi network, go to if this and that, send some data, get the response back, write a file or store data in a database, and then send me an email notification.
Peter 29:29
And it can do that flawlessly. So I feel that if I didn’t know what Arduino code looked like and how it worked, and if I started like that, I would probably never learn how to program an Arduino on my own, because I wouldn’t have to. So that the impetus to learn wouldn’t be there. So got to be very careful about that. And I guess in the industry, when you have new graduates in whatever discipline that might be, it could be hospitality, economics, technology, you get new graduates. So the next few years, when they’re going to start graduating from the current crop of students, one of the things that hiring managers would have to do would be to figure out how much these students actually know versus how much they think they know, because they’ve got the support of an AI behind them. Now the good universities, I’m not too worried about good universities. They have processes in place to ensure that the quality is maintained and the graduates will be top tier, no matter what kind of technology they use. But you can’t be sure for everyone. So yeah, that’s something that will start becoming important in the next couple of years.
Neale 30:51
Yeah, it’s an interesting issue, isn’t it? I mean, when I went to university, the big issue at the time was because that’s how I graduated in 2003-2004. So I had a Windows computer. I don’t think I had access to the internet at home, but certainly university we did. And the problem the universities were facing there were, of course, people getting hold of papers, using existing content and just copy and paste jobs. So it’s kind of like the next level from that, I suppose.
Peter 31:19
Yeah, I work with AI now every single day. I feel that I have a team behind me doing important work. And I feel my productivity has gone off the roof because of that. But I’m very careful. So every single bit of code that comes out of it, I test it, I read it, make sure that I understand it, I can modify it. I only get it to do things that I understand, that I would be able to do myself. It would just take me like three or four days instead of five minutes. That’s a type of work that I get AI to do for me. But I’m still trying to establish some kind of operational parameters. How do I have some barriers in place? And it’s a moving goalpost because it just constantly changes the capabilities. And now you have multimodal artificial intelligence. I have to get used to talking to the computer because it does speed things up as well. And it speaks back to me. And I feel like, because look, the voice quality that comes back from these systems now is just perfect. I was playing with Google AI Playground and the audio, like the voice that comes back from the computer with human kind of inflection. It will make a mistake, it will apologize, or I make a mistake and I say, oops. And it will respond like a human.
Neale 32:55
They’ll stop swearing next. That’ll be the next step.
Peter 32:58
You never know.
Neale 33:00
What was the most interesting problem that you’ve been asked from one of your students who are in the community that’s come across your desk?
Peter 33:07
That was a real challenge. The most interesting problem, because right now I’m working with PCB or one of my courses is kick out like a pro. I’ve got a couple of course actually there. a question that does come up not frequently but you know once a student reaches a certain level is they want to create more complicated PCBs. So initially in my courses on keycard I’d show people how to create something like a Raspberry Pi hat with a few sensors and buttons or a shield for an Arduino or like a clone ESP32 device and those things are quite easy to build because they are low speed and they don’t require too many components. And that means that you don’t have to worry about all these intricacies that PCB designers have to worry about when they design things that operate in much larger speeds and contain more components in a smaller amount of space. So I won’t give you a single question but the nature of these questions have been pretty much consistent is how do we design higher speed circuits that are also using smaller components like components that you can’t really solve that with your hand you need some kind of robotic device to operate. So that takes you to the level of basically designing a computer which is what my last keycard like a pro course did to an extent and what my next one is going to do to the full extent. So I rank student questions in a way based on what my interests are. So if students ask me things that I’m myself very interested about then they go up their rankings so they go into my this probably going to be a good course kind of list. So that’s the top one. So my next course is going to be about designing a computer essentially with high speed interfaces lots of mini miniaturization integrated circuits on it ram communications disk interfaces and things like that. So yeah maybe it’s good. So I’ve never done something like this. That’s another thing I should disclose right now. So I’ve never done something like this. I’ve done up to the level below that but you know that’s always the first time the principles documented and just like you know you’re talking about the ingis earlier sorry ignis ignis earlier and that you have a peer review process and I do the exact same thing. So I have a peer review process especially for high stakes content where an error can cause catastrophic consequences to my course and reputation. I cross check things so this is going to be the same thing where I will do the design learn do the design cross check it with third parties for you know design qualities fix errors then that’s going to become content for my next course. So yeah I mean it’s probably not a fully satisfying answer because I don’t get groundbreaking questions but questions like that okay they always make me think about you know this student asks an intelligent questions because they understood stuff that are important and they are ready for the next thing and I find that those questions are really good.
Peter 37:17
So they show initiative and they show that they know they know what they’re talking about and the questions relate that. It shows the passion as well doesn’t it that’s you know they’re ready for the next step they want it and they’re excited for it.
Neale 37:29
That’s fantastic. All right Peter this has been really enjoyable thank you so much for your time.
Peter 37:36
Thank you thank you for inviting me.