Welcome to the Electromaker Meet a Maker podcast, episode 4! Senior Editor Moe Long sits down with Mike Darby, aka 314Reactor, to talk maker communities, film, and more. The brains behind such projects as a "Terminator 2"-inspired Raspbinator skull, The Scorpion "Star Trek"-inspired robot, and Windows 98 wristwatch, Mike has loads of awesome knowledge to drop.
Electromaker: I know about your maker projects, but what do you do for work?
Mike Darby: I work as a Q/A Engineer for a small company in England. I just test software pretty much.
Electromaker: So did you get into the maker space after being hands-on with software, or did you get into software after being hands-on with hardware? Or was it a little bit of both?
Mike: I think I've always been into software and hardware ever since I was very, very young, so it kind of just let into my career. And then when the Raspberry Pi and stuff came out I got into that.
Electromaker: Can you talk a little bit about your maker origin story then?
Mike: Yeah, I first got a Raspberry Pi back in 2013, and I got a little board with it...[with] little LEDs and a little buzzer on it. It was before the HATs were released, so you just had these experimental boards or you made your own. And I played with Python for a bit, and I kind of ran out of ideas for a while what to do with it. Then suddenly, around 2014, I don't know what happened, but just had hundreds of ideas flooding in and in and in, and I just kind of couldn't stop making stuff.
Electromaker: What initially sparked your interest in the Raspberry Pi?
Mike: I heard about it in 2011 roundabouts and thought that's pretty cool. The way it was advertised was a little credit card-sized computer, so I thought that'd be cool to get Linux on and stuff, but then when I actually got it I realized the more hardware application side of it and I was like "Oh, ok, so there's a lot to do here." Not just running Linux and using it as a PC, but actually using it as a platform to build pretty much anything on. Ever since I was a kid I wanted to build stuff like that.
Electromaker: You mentioned the Raspberry Pi. What technology, hardware or software, do you enjoy working with the most and why?
Mike: In terms of making my own stuff, it's just everything. It's a really tough question because in terms of what I do, messing around with computers, I love everything from graphics cards to Raspberry Pis, but the most interesting stuff is probably AI (artificial intelligence) because you can now get TensorFlow on the Raspberry Pi. And I really want to get to grips with that. It's the more forward-thinking, experimental stuff that I really like.
Electromaker: The AI part really intrigues me. One of my favorite projects that you did was your Artificial Life project. Can you talk a little about what some of your favorite projects that you've completed are, and how you come up with project ideas?
Mike: My own favorite projects...the Raspinator, because there's a lot more I can do with that, the more I learn. And it's also influenced by "Terminator," because I love those films. And that's where I get a lot of my ideas from. "Terminator" 1 and 2, not the others obviously. "Terminator 2" is just a film that resonates with me and it just gets more and more relevant every year. Also the Marvel films as well. I look at the technology in them and want to make it real. And one day I just want to have an Iron Man suit, so that's how I get a lot of my ideas. Just trying to take something that's abstract or from a movie and trying to make it real no matter how impossible it seems.
Electromaker: So what are some of your favorite films that either portray tech in a really good or terrible way?
Mike: Well, I'd say in a good way it's probably the Marvel films. In terms of bad though, "Terminator 2" because it's not just bad technology but how we reacted to it. Skynet became self-aware and then we reacted by trying to unplug it, and we reacted by trying to unplug it so it kind of defended itself. That's a kind of mixed bag of it had a good application but the way we reacted to it was bad so it's kind of a grey area. There's stuff like "Altered Carbon" which is a really good technology, it can be put to a really good use, but in terms of application a lot of people kind of abuse the technology. You get a lot of high up people using it, and lesser fortunate people tyring to abuse it.
Electromaker: Those were great examples. And even "Terminator 2" trying to unplug Skynet calls back to HAL from "2001: A Space Odyssey" through trying to fight back...So your projects are often more difficult or moderate. What's a challenge you've encountered on a recent project and how did you overcome that?
Mikey D: Probably trying to get Raspinator to come up with decent responses with the chat box program because it's so difficult to get something to respond in a remotely human way. And the way I did it was go back to basics of how anyone sort of learns. I mean, you observe the world around you, interactions with other people. And when someone interacts with you, if you don't necessarily first understand it, you react back and then you see their reaction come back. And then you know, it's up to you after this. So I kind of programmed it to take your reactions against its reactions, and then flip that around it is another response come in, because it can use a human response back if that makes any sense.
So it kind of just learns from examples like its heard stuff before. So if you say something to that it said before, it will have a response to it from your own responses to it. That's how I sort of overcome that if that makes any sense.
Electromaker: And chat bots especially seem to be dominant of late. So do you have any sort of thoughts on how those will continue to evolve?
Mikey D: I imagine with soort of deep learning stuff, machine learning stuff, they'll probably end up getting sort of plugged into wide arrays of languages and language models. And probably will be more sort of mass learning sort of stuff as well as...you get two times tablet, ones that will like mass learning where they learn a whole load of interactions. And then singular ones that learn from individual interactions. And know both develop quite different because you don't want to have unique personalities from talking individual people. And then you have the ones that just learn from everything on the internet. And that we sort of get scarily smart sort of chatbots. I think that's probably where they'll go a lot of deep learning a lot of TensorFlow kind of stuff like that. I think that's probably where it's going to go.
Electromaker: You've used a lot of different technology from kind of more accessible hardware such as Raspberry Pi's to chatbots, which the average or beginner maker probably doesn't use. Do you have any sort of technology they haven't been able to get hands-on with yet that you really like to?
Mikey D: Probably the more sort of complex machine learning stuff like TensorFlow, I haven't tried that yet. I need to get on that. Get learning that because that's really complicated. And I'm quite a slow learner release that's going to get around good.
But yeah, that's one technology, I really want to get on board with is TensorFlow, because that opens up a lot of opportunities to get some really good sort of hands-on with machine learning and see see what you do with that possibly put that in the next Raspinator as well.
Electromaker: And I'm glad you mentioned the next version. Because you made the windows 98 wristwatch and then you just debuted 2.0...which I'm really excited about. Windows XP was one of my favorite operating system so I might have tried to replicate that...but how do you pick and choose which of your projects you update?
Mikey D: I think I kind of look over them and then I look over them and go I can improve that one this way or a new technology has come out and I can approve it. Like the artificial life project, there's been a new higher resolution unicorn HAT and there's also a new LED cubes that are on Pimoroni, now and I really want to make like a 3d version and then like a super HD version with the massive unicorn HAT we've got. So I kind of look at the technology perspective. Also, while I'm making them...I kind of have an idea beforehand of what I want at the end of the process. I basically aim for, like a self-aware skull and then kind of built my way to that. And then as I'm building I'm like, well, I've gotten this far, this is like stage one, I can wrap this up, learn what I can from that. And then for stage two down the line I've learned a lot more I can implement that and then as I go from other projects I learned other technologies and then that feeds into the next project and so on so forth. So it's kind of got an idea in my head and kind of work toward that slowly.
Electromaker: That's a really awesome iterative process...So you tend to have a lot of projects that are very like public facing? Do you have any particularly interesting or unique interactions that you've had with fellow makers or fans?
Mikey D: Um yeah there was one person that took my my Nerf gun ammo counter project and they improved it and like did loads of 3d printed parts and stuff and then put on their own weapon and made it really cool. I thought that was really cool because...someone's taking my idea and because I wanted them ideas out there people improve and learn from and then figure you know, it feeds into the overall open-source world out there and if someone look at my idea and "go that's cool, I want to improve it and make my own thing," that it was really amazing and I was like that's so awesome that somebody actually thought my rubbish thing was so cool that they want to improve it...
And then on the more...not the negative side, but some people make very valid comments that some of my stuff isn't so well made I could 3d print a bit more which is it's true I agree with them on that because a lot of my stuff is very prototype been in doesn't look very aesthetically pleasing often.
Electromaker: The way you phrase that was really fascinating and it's something that I really love about open source communities is the way someone will improve upon something someone else is created, whether it's hardware, whether it's code, and rather than being upset about that, it brings joy. So that's what I really like about the maker community...And you're talking about someone kind of improving on something you'd created so on the flip side of that, what makers do you follow on a regular basis and inspire you?
There's someone who works for Hackster, I think her name is Alex Glow. I have seen a lot of her projects, she's quite inspirational. Just one of those people who's very clever, far cleverer than me, they can just make these really cool things.
And I generally, I can't think of any names but I see stuff people have made like the portable GameBoy stuff and other such like really common projects inspire me and I sort of go "yeah, I want to try my hand at that," or I made a portable Pi recently. Like a little pocket thing that was inspired by a bunch of other projects. So yeah, it's pretty much the sea of it that's out there I just kind of get inspiration from but often my stuff isn't as good as that stuff
Electromaker: You're so humble!
Mikey D: I try to be because...I think I'm very, very lucky I mean just sitting out in the interview someone right now he thinks my stuff is cool is amazing. Because it's like wow, someone thinks my stuff is cool. So yeah, I try and keep very level headed on it. Because there's always more to learn as well. That's that's the other thing.
Electromaker: Absolutely. So what inspired Would you give to someone first starting out in the DIY space?
Mike Darby: I would say just pick a simple board like a Raspberry Pi, or Arduino. And then just pick it up and just stop building stuff. It doesn't matter whether it's just a little LED flushing or anything, really just do anything your imagination can come up with. And then there are loads of great communities out there. Like Hackster, Electromaker. Loads of little maker communities on Reddit, where can push your stuff and people can respond to it. And then yeah, just keep making and thinking about stuff. So I think like I did people may run into a sort of area where they can't think of anything to do, but eventually if you just...let your imagination flow, eventually you'll just think the opportunities are limitless.
Electromaker: Definitely. And that's kind of the beauty of a lot of the projects that you see. Even something as simple as blinking an LED is immensely satisfying if you've never done that before and then it kind of gets you excited about moving on to bigger projects. It's also really nice when you see people in the comment section being supportive of basic maker endeavors...We talked a little bit about movies earlier what is one film that you would recommend every di wire and maker watch and why?
Mikey D: That's a good one. It would probably have to be Iron Man because it just shows [Tony Stark's] ingenuity is so inspirational the way we can just sort of mock up one suit in a cave while under surveillance. And then he sort of just busts his way up and just sort of wins the game makes a Mach 1, Mach 2, Mach 3...He goes from the Mach 1, Mach 2, Mach 3, that kind of iterative process is so inspirational to me. So yeah, definitely, I'd say Iron Man because, Tony Stark, he's just cool.
Electromaker: And even though it's sci-fi in "Iron Man 3," he's got the bots and then of course in "Ultron," [it deals with AI as well] even though I didn't think "Ultron" was the best of the "Avengers" films or the MCU, it was still fun. So I think that kind of focus on AI as well as the iterative process was actually kind of intelligent and I think the average Marvel fan might not recognize.
Thanks Mike! Keep being awesome and inspiring us!