Limor Fried // @adafruit

OPENING QUOTE: I think it's actually better to try to teach and present electronics engineering from the top down. There's a problem you're trying to solve and what skills and technologies do you need to solve it?

Brian: That’s Limor Fried, founder of Adafruit. And this is The ReadME Podcast, a GitHub podcast that takes a peek behind the curtain at some of the most impactful open source projects and the developers who make them happen. I am bdougie aka Brian Douglas…

Kathy: And I am Kathy Korevek.

Brian: In every episode, Kathy and I invite a maintainer or open source developer into our studio to explore their work, their story, and where the two meet. 

Kathy: In this episode, we speak with Limor who seeded the idea for Adafruit when she was a student at MIT. As you will hear in this interview, she decided to make her own MP3 player and this triggered an interest from those around her and beyond to make their own electronics as well. It all started with empowering people with the tools and allowing their creativity and knowledge to lead the way. Seeing that there was a desire for people to make their own electronics and wanting to satisfy that, Limor started Adafruit, an open source hardware company based in downtown New York City. Adafruit provides the tools for people all over the world to make anything electronic they can imagine. This all started in 2005 and by 2014, Adafruit had a revenue of $33 million dollars. And today, it has 100 employees. 

In this conversation, Limor talks about the birth of Adafruit, its philosophy, and how open source plays into it all.

Kathy: Limor, welcome.

Limor: Hi.

Kathy: I feel like you're kind of a legend.

Limor: Yes.

Kathy: I feel so privileged to talk to you.

Limor: It's awesome to be here. I'm hanging out on GitHub.

Brian: Tell us about your first earliest computer interaction memory? It's a good way to start the conversation.

Limor: Okay. Wow. My first computer interaction memory. Well I don't know if this is my very, very, very first because I was pretty young and I was pretty lucky that my dad is a professor and so he used computers. He was a professor who would use computers in math analysis. And so we had a home computer pretty early, basically like ‘84, ‘85, which was pretty soon. But we had a Mac 512K, and these were adorable little computers. And I liked playing games on them, people would swap disks around and I would use ResEdit. And I would like change all the icons. This is what we did before Fortnite. We used ResEdit to fuck with programs. So that was kind of fun.

I was actually talking to somebody yesterday about... They were like, what's your first program experience? And technically it was with this program called Adventure Game Toolkit, which was a text adventure programming system. Again, before Fortnite, all we had was text adventures. And I was like, oh, I want to write a text adventure. And so there's this tool called Adventure Game Toolkit and the biggest challenge for me was I didn't understand that you wrote code in a text file and then applied the compiler to it. Which seems kind of obvious now maybe, but at the time you're like, well, there's this thing that makes a program. And I'm like how do I... I tried double clicking it. I didn't realize you have to run it on a file. Weeks, weeks it took me. But eventually I figured it out. That was the hardest part.

Kathy: I feel like people might now, I mean, I don't know. They might not understand what a text game is, which is kind of sad because I grew up playing text games too.

Limor: Do they not know what text games are? I mean, I'm thinking there must be something online that's similar. I mean there's probably websites where you type things in, there's text games on a website.

Kathy: It's like a “choose your own adventure.”

Limor: Yeah. Anyways, there you go. That's your first question answered. Done.

Brian: All right. Move it right along. Also just a shout out, I watched the Netflix documentary during COVID about the history of video games and they talk about that. And I think John Carmack, I think, was in that documentary.

Kathy: I saw this one, yeah it was really good.

Brian: I don't remember the name, but if anybody wants to go and watch Netflix and learn about text-based adventure games, there you go.

Limor: Yeah. Yeah. InfoCom was... I grew up in Boston and InfoCom was the company that did these games. Very, very popular, very successful. Did Zork and other ones and they then pivoted to doing database design and totally destroyed their company. Very impressive.

Kathy: Limor, so I think a lot of people probably, especially folks who follow you… they know about Adafruit, I hope. But I'm wondering if you can just tell us a little bit about the inception, the creation, the birth of Adafruit.

Limor: Right. So I had to write this thesis and I really didn't want to do it, like really badly. And so I was like, what can I do to not write this paper? I know, I'm going to build my own MP3 player. That's where I was at. I was kind of looking online and I knew a little bit about electronics, I was studying computer engineering at school. And so I was like, well, I want to make things with it. And there was a group of people who were all making like LEDs and blinky things. And I found some MP3 player chips and so I decided to make my own MP3 player based on some tutorials online. And so I did that and that was very successful for avoiding writing my thesis. And then I continued to avoid my thesis by doing other projects. They just kept coming because procrastination can make you very prolific.

And eventually I did have to write this paper and I graduated and I was making stuff like synthesizers and little pocket games, DIY game boy type things and was posting them online on blogs. And folks were really into them and they wanted to build them. And the way electronics kind of works is that basically the price for buying one set of parts is like $30, but buying enough parts to make a hundred is only 10 times as much. It's very nonlinear. If you buy enough parts to buy a hundred kits worth, it's not that much more than just buying parts for one, because it's just high setup cost, especially for a circuit board.

And so folks who were emailing me and like, hey I want to buy a kit of these parts. Will you sell me a kit? And I actually was working at the time at a small startup, but for fun I was like, yeah, sure. I guess I'll do that. Let's do a run of a hundred kits. Now they would call it a group buy, but I don't think we had that word yet. It was basically a group buy and I had a little PayPal button and people would send me 20 bucks and I'd send them a kit. And then I just kind of kept doing that. And I was sort of making new kits every once in a while. And it's fun because I get to make new things constantly. I'm not stuck with one thing that I have to... Well I have to support stuff, but I'm not like making one thing and iterating on it as much as kind of like doing new, interesting things, whatever sort of attracts my attention.

Kathy: That's awesome.

Limor: But now I have a hundred employees.

Kathy: Okay, so you've opened the doors in a lot of ways, kind of being a woman pioneer in this male dominated industry, making hardware development more accessible to hobbyists. Can you discuss some of the challenges that you faced along that path?

Limor: Yeah. It actually turns out, when you run GCC if you're a woman, you have to add extra dashes to the flags. No, not really.

Kathy: [Laughter]

Limor: Writing code and doing electronics, it doesn't actually make any difference. It sucks to write code and debug things, even if you're a lady, but here's something kind of neat. If you're a woman or if you're under-represented in your hobby or profession, you're going to be underestimated a lot, which is very powerful. Because they never see you coming.

So I've tried to use that as my special power, because people are like, “Oh, there's no way she's going to do this. There's no way she'll design this and there's just no way they'll put me out of business.” And then I put them out of business.

Kathy: And then you're the first woman on the cover of Wired magazine.

Limor: To take them out. Yeah. Why not?

Kathy: That's awesome. I love that perspective.

Limor: Yeah. So I'm using this as a technique. It's like you look so harmless. Yes.

Brian: It's a Testament to open source too as well. And how you're able to one, put together kits for things you were interested in at the time and maybe other folks are interested in. But also being able to grow community around this too as well. So have you seen... Adafruit wasn't the first project company to put together kits for hardware computers, but are you seeing a trend in folks really adopting hardware computing through, I guess, through interacting with Adafruit?

Limor: Well, so my thing is I personally find engineering and making electronics and all this stuff very interesting. I don't know what's wrong with me. I personally like it. I enjoy it. I think that trying to convince people that electronics and computing is interesting on its own terms is not a good idea. I really don't like how every book that's about “I'm going to learn electronics.” I'm reading a book and page one is like, “this is an electron.” And I'm like, why the fuck are you teaching about electrons? Nobody cares. Nobody cares what an electron is. You're never going to see one. You're never going to interact with one, you deal with them as an abstraction, but it doesn't actually matter. Who cares, which way it flows. Positive, negative, negative, doesn't matter. Right?

And I think that there's this obsession with teaching from the bottom up in a lot of ways, because I don't know, we learned that in school or something. You learn an atom and then you learn a molecule and then you learn like bio and then you learn cell structure. And then before you know it, you're like the mitochondria are the powerhouse of the cell or whatever. And then it goes on and on, it goes up in complexity. Whereas I think it's actually better to try to teach and present electronics engineering from the top down. There's a problem you're trying to solve and what skills and technologies do you need to solve it?

And I think if you do it that way, then people are like, “Oh, I'm in a four H group and I have a chicken coop that is constantly getting raided by local foxes.” Even though I'm not an engineer, I want to design something that's a motion sensor that will detect when a warm thing is near my chicken coop and sends me an SMS message. But if you had to start from electrons, you'll never get there and you'll hate it. And you'll probably quit. So what I try to do is say here's all these cool things that you may want to do and if you learn engineering to do it, amazing, wonderful, I'm glad. But nobody would ever say, you're not allowed to read a book until you've taken a class in literature.

I actually don't know how to segment a sentence. You know in school you learn a noun and an adverb. I actually don't know what an adverb is right now. If you asked me to point at an adverb I actually don't know what it is. I know adjectives and nouns and verbs but I don't know all those other things. They never had clauses. But I can read books and understand them and enjoy them. So I think take out the tweezers, remove the detail and enjoy the larger scope of what it is, is my idea. I don't even know if that was your question, but that's my answer and you're stuck with it.

Brian: No, that's a great answer because drawing it back to, you mentioned about text-based adventure games, which also shout out, High Score is the documentary on Netflix.

Limor: Okay, good. I know you're Googling that frantically like, what is this? Okay.

Brian: But, you also mentioned Fortnite in comparison to text-based adventure games. No one needs to go back and play the original games. It will be fun and you could probably find them and hang out with those. But also if you want to play Fortnite, play Fortnite.

Limor: Absolutely. And I think there's people who probably do cool things where they program keyboards to automatically create forts or do machine learning on their PCs so they can automatically detect things in the space. You can do cool things with computation and electronics and sensors, whatever your game may be. You actually have a lot of people who do cosplay. And they want to have costumes that look like their characters. You can't cosplay as text adventure. I mean, you probably could, but it'd be weird because you just have a lamp and a gru and what does a gru even look like? I don't know. And a piece of paper with a map on it. I guess you could cosplay.

Kathy: It'd be pretty obscure.

Limor: It'd be really obscure. I mean it'd be cool at Dragon Con cause everyone would be like, that's so obscure. That's the third character in...

Kathy: Yeah, I mean the people who would get it, would get it right. And they would be so stoked.

Limor: So stoked. So people like to do cosplay, a lot of cosplay is they have to do a lot of electronics now, they have to have LEDs because everything's glowing and like making noises, but they're not learning electronics and then doing cosplay. They're doing cosplay and they're like, I really want to make this glowing sword that when I poke something with it, it makes the noise that the sword makes in the game.

Kathy: You know, speaking of things that are glowing and things that are making noise, all that kind of stuff. So you've got this big warehouse in New York City where everybody works in your company and I'm trying to picture it in my head. All of these parts, all of these light bulbs, all of these things coming to life in this warehouse. I've never seen a photo of it actually. Can you describe what it looks like to us?

Limor: Yeah. So it looks like every photo you've ever seen of Alltrax Manufacturing. No, the main engine is this pick and place line. And the pick and place line is something that basically takes circuit boards, which are these custom CAD designed things that look a lot like what my background is on Zoom, which you guys can't see, because you're hearing my voice. But it's got traces on it and when people think of circuit board style, that's what it is except it's designed to actually connect traces and components together in the right order. And I design those and I send those out and they get manufactured at a factory that does that. It's a chemical process. We can't do it, you need to basically have a setup to do it. And you need to be doing it 24 hours so you can recycle the chemicals, which again, we wouldn't do.

And then we had the circuit boards and these components, the components come on reels, they're just spun around on spools basically, of components. And then the pick and place, you load it in and then it has these little pickup nozzles and it can pick up the components off of the spools and place them into the right location on a circuit board, that then goes through the oven and it kind of melts the components into place. And then we have to test them and program them. But that's kind of the engine and everything else is sort of the support network for the machine line.

Kathy: Okay. So when you walk into your office, what's the feeling you get as the creator of all of this as the head of this company, when you walk into that building?

Limor: It's pretty cool. I mean, we've got some awesome people. I really like the folks that we've hired because they are people from New York, we look like New York, we're New Yorkers. We just hired a bunch of people because we're finally getting back into business. Business is starting to pick up a little bit again. And these are people who are like, I have kids and I want health insurance and I want a job in New York and I want something that's reliable and something that gives me benefits and stability. And that's what we are. We like being this member of a community and having all these New Yorkers from all around the boroughs and some New Jersey, some Connecticut of course. All come here, it's just... we try to make it a really good place for them to work and a really fulfilling place. Where we're actually making something that helps people.

During Covid we were making face shields and we were helping make bridge ventilator components. And we were helping people with... Actually one of our recent projects that became very popular is we have projects to make CO2 sensors that you can put around schools and offices and buildings, and they show you the CO2 levels, which is a very good way of analyzing your airflow. Because actual actual air flow analysis is a total pain in the ass to do, you need fans and sensors. But this thing it just tells you, “Do you have enough fresh air? And if you don't, open up the window.” And so those are actually a popular project for people to make now for their classrooms or for their offices or for their homes.

So actually what we're doing is really helpful to people and we're able to move very quickly for what the engineering community needs or wants.

Kathy: I want to ask you a question about the repositories, but we can come back to that too.

Limor: Ask it now, we're here.

Kathy: Well, before we get there, Brian, do you want to ask this question about... This next one, about keeping the company local.

Brian: Yeah. Yeah. You had mentioned that the company looks like New York, you have representation of all the boroughs, Connecticut, New Jersey as well.

Limor: Yeah, “bridge and tunnel” also count. We love them.

Brian: Excellent. Well, I think everybody listening from there appreciates that and the shout out. But can you talk about the value and why you value the company that looks like New York and its representation of folks from the local community?

Limor: Yeah. So what's really neat about the work that we do is... And I'm going to be just like, a little bit mean to the city. So I'm really sorry for anybody listening from the city.

Kathy: That's okay. I feel like if you're from New York, it's okay.

Limor: No, literally from the city government.

Kathy: Oh, okay.

Limor: So the city government, they were always kind of like, “Oh, don't you need to have Ivy League engineer-trained people to do this stuff?” Hint, hint, we should open up campuses and get Cornell to move here. And I was like, no, we have people who haven't even graduated high school working for us. They have a GED and it's like, I can train them to do electronics manufacturing. If you have good documentation and you are willing to work with people, folks from New York, they're so hardworking and they're so focused. They're just so there. They want to live their lives. They almost always have something else going on. They have families they're raising or they're doing theater, or they love photography or they're volunteering in some of the nonprofits.

Almost everyone who works for us has something else going on. And so Adafruit is... For those people I want Adafruit to be a stable place that they can come in, they can do the work. And then when they're done, they go home and they don't have 1500 GitHub repos to get emails from. They're moving on to their next thing. So I think for me, it's worked out. We've always hired from within. So our CFO started in kitting and our COO is a painter who came in and did facilities work and moved up into being a COO.

Everyone who works for us, nobody was hired externally as, “come in and be the manager.” These are all people who, they worked at restaurants and they hated it. Or they worked for Urban Outfitters and they eventually hated it. Or they worked any other entry-level retail or service job. And then eventually they're like, I'm kind of tired of that. I want a job that I can manage people and learn these skills, but I'm not dealing with customers also yelling at me at the same time. And it's really worked. We have very, very little turnover, people who work with us, they work with us for like 10 years.

Kathy: That's really cool. I mean I was listening to, on Friday afternoon, I was like, oh, I'm going to check out Twitter while I'm on my walk today. And I see it and there's the notification that Brian Douglas is speaking right now in Twitter spaces. And it happened to be this really great conversation, all about getting people involved in tech and in coding and there were four guys and some of them I hadn't heard of, but mostly from Atlanta. And they were all talking about different ways that they have seen, and that they're doing with their companies and in their open source projects, to get people involved in coding.

And one of the coolest things is that just through open source and maybe now we can talk about your 1500 repos, but through open source there's this automatic, almost democratization of code. And I wonder if you can talk to that a little bit. Adafruit is a hardware company and I feel like not everybody knows that you maintain 1500 repositories and I mean, maybe not maintain them, but you have all of this code out there.

Limor: I absolutely get emails for each one.

Kathy: I'm sure you do.

Limor: So I think it's like take it, steal it, the whole history of computers is all about theft. And property is theft, theft is property. So if you see some code you like, take it. I mean, yes once in a while, if you're a very big company, you'll end up going to court over it.

But for the most part individuals, they don't, and you should just take whatever you see. Because in my opinion, and I know people, they're like, “Oh, we need the copyright to maintain the GPL.” And that's great. And there's some people who are more [inaudible] or copy that for whatever. But in my opinion, code sucks so much. Okay. Coding is so bad. It's horrible. You're typing numbers into a computer.

Kathy: That's the dirty secret everybody.

Limor: Okay. No, I mean, it's true. It's really terrible. You can get into a zone and you can be addicted to it, but it's actually not good for you in my opinion. I don't think so. I feel really bad after typing numbers into a computer for like seven or eight hours. I don't know about you, right?

Kathy: No, I do too. I think we can admit to that. It's the reward afterwards. It's like seeing the thing being made-

Limor: It's amazing. It's so worth it.

Kathy: It makes you black out all the rest of the pain.

Limor: Right, but it's terrible. And so any code that somebody can take or any design they can take and then they use it, would be great and then good for them. I've actually just evolved past the thinking of people using my code as like, oh, they took something. I just don't really care anymore.

Oh, you stole my rabid opossum. Why am I angry? Okay, if you gave it a good home, congratulations, it bit me.

Kathy: It's really interesting because the artist community is a lot like this too. There's this concept of “there's no original art.” And my sister is an artist in New York and so she's going to listen to this and totally be like, “I can't believe you said that.” But it's the borrowing of ideas.

Limor: Do you know what the definition of art is?

Kathy: What?

Limor: It's anything that's in an art gallery.

Kathy: [Laughter] Yeah. Yep, exactly. But I love that. And what you're doing is an art and I think open source is an art. And if you think about it in terms of, you're constantly borrowing, stealing, you're then pushing this thing forward. You're taking it, making it your own, turning it into something.

Limor: Yeah. There's too much obsession about trying to be original. And I think it's actually really toxic and it keeps people from getting to be able to express them... because even okay, you can't step in the same river twice, because you're a different person and it's a different river. Somebody reimplementing the exact same project. There's no point in trying to pretend that anybody implementing or even compiling or running the same program is going to have the same experience as another person. So you should just let them. You should give people the tools to do what they want, and so that they can have their own self-expression and success in doing it. And so there's two parts of that for me.

So every time I do a podcast at the end I'm always like, you aren't going to play this because I didn't talk about anything that you wanted. There's two parts about coding and technology that I think are essential. And at no point, is it the CPU, the amount of ram or what microcontroller you used? It's whether there is a pool of code or designs that people can look at and learn from, because we're mimicking animals. We look at how other people did stuff and we mimic it. And then we do something a little bit different and you have to have a community that doesn't suck. That lets people have those experiences. Because if you have a community where immediately it's like, “That's lame, you just copied that person,” they're never going to get past the duplication process into the synthesis process. Because you have to do duplication, you have to mimic before you can create. Humans are like this.

Okay. So somebody's going to be like, “I'm an evolutionary biologist. I disagree.” I don't care. You're not on this podcast. I am. Right. So you have to mimic first and then you can create. This is true.

Kathy: I love that.

Limor: But so many people, they screw you up before you even get past that mimicking state. Because for some reason, mimicking states are like, there's something people get really attacked by them. And they get defensive and they're like, you're just copying, it's cloning. You're not doing anything original. When like, who cares?

What we really need is more tutorials on how to blink an LED. That's what we need. Even though there are 50,000 of them. We need more. Because there's always more beginners that haven't experienced it than people who have.

Kathy: Speaking of beginners, what's the one piece of advice you wish someone would have told you when you were getting involved in electronics?

Limor: Who gives a fuck, just do whatever you want.

Brian: That is too good.

Kathy: That's good.

Brian: So much truth in that.

Limor: Decide what you want and then do it. I think actually that’s the hardest part is people don't know what they want, but if you can figure out what you want, you're well ahead of most people. And then you can actually get to that stage.

Kathy: As a kid, did you like to make things?

Limor: I really liked to sew things, I used to make my own clothes a lot. Which were really ugly, but that's okay. It was the 90s and big pants were in. I don't know what to tell you. The bigger the pants, the better. Just huge. Just get lost in them. I mean we put drugs in them basically was, I think, the answer.

Kathy: That was the reason why you had pants. I don't remember. I grew up in the nineties. I don't remember big pants, but I grew up in Alaska where we wore farm clothes.

Limor: Big furs. I don't know.

Kathy: Yeah, big furs.

Limor: I don't know. I don't know what you wear in Alaska. I'm assuming lots of layers.

Kathy: Yeah. You wear lots of layers. Exactly. I love that you used to make a lot of your own clothes. There's a lot of actually... I do a lot of knitting. There's a lot of math involved and I feel like a lot of almost, very engineering focused decisions you have to make when you actually do make your own clothes.

Limor: Yeah, we had a robotic knitter that we hacked a while ago. I can't believe we did it, we just-

Kathy: Okay, you're speaking my language.

Limor: Yeah so we programmed it. It was like the brother KH something, something. And it had these diskettes that you would program it. But of course these diskettes had been demagnetized 30 years ago. So somebody figured out the protocol that you could use to connect to it. And so we basically had this cable with a little bit of hardware and we did very little wiring. And then we wrote a Python program and you could basically load a BNP into the printer and have it print out custom bitmaps.

And that was cool. So we started to do QR codes or like video games, somebody made a video game that as you knit the video game would play.

Kathy: Would come out. Yeah. That's so cool. So I have-

Limor: Yeah, each stage your character would move with the knitting machine.

Kathy: Whoa, you got a lot farther than I did. I have a Brother Electroknit and it was the same thing. I had a little Arduino board and-

Limor: Yeah, that was a project. You probably used our guides for that.

Kathy: I probably use the same projects. Yeah. So somebody in Italy had this open source board that I bought and then I was able to hook it up to my computer, made a little JavaScript application and was actually able to run and print my own designs.

The designs I printed were like pluses and minuses. It wasn't anything close to what you were doing.

Limor: Right. But I have a question for you.

Kathy: Yeah.

Limor: What's an electron?

Kathy: What's an electron?

Limor: Yeah. You have no idea.

Kathy: Yeah.

Limor: Who cares?

Brian: Doesn't matter.

Limor: It doesn't matter!

Kathy: Yeah.

Limor: It's completely irrelevant to you being able to build a project that uses engineering and electronics. You don't need it.

Kathy: Yeah. You don't need it. Well, I mean, technically I needed Electron the application

Limor: The structure. Yeah. That's different and that's not my problem. I think that's Google's fault or something.

Kathy: Yeah.

Limor: Whatever.

Kathy: Yeah.

Limor: All right. Cool.

Brian: Excellent.

Kathy: Cool. Fun to geek out with you.

Brian: Pleasure, Limor.

Brian: So good to have Limor Fried on the ReadMe Podcast. To learn more about Limor and her company Adafruit, please go to adafruit.com.

I am Brian Douglas, and I am a developer advocate here at GitHub.

Kathy: And I am Kathy Korevek, I work in product. The ReadME Podcast is a GitHub podcast that dives into the challenges our guests faced and how they overcame those hurdles. In sharing these stories, we hope to provide a spotlight on what you don’t always see in the lines of code, and what it took to build the technology that inspires us all. 

Brian: It’s been really great spending time with you. The ReadME Podcast is part of the ReadME Project at GitHub, a space that amplifies the voices of the developer community: The maintainers, leaders, and the teams whose contributions move the world forward every day. Visit github.com/readme to learn more.

Our theme music has been produced on GitHub by Dan Gorelick with Tidal Cycles. Additional music from Rhae Royal and Blue Dot Sessions. 

The ReadME Podcast is produced by Sound Made Public for GitHub. 

Please subscribe, share, and follow GitHub on Twitter for updates on the podcast and all-things GitHub. Thanks for listening!