balenaPodcast episode 03: Open hardware and empowering humanity with technology
Our guest, Jason Kridner, Co-founder of BeagleBoard, joined Alex Marinos, CEO and Founder of balena, to discuss overcoming supply chain issues and artificial barriers to technology, the future of open, modular hardware, and how technology should empower (and not limit) humanity.
I don't know when we get into it, but it's been a while since we caught up face to face, or even, even just something like a video chat that I get. Um, I think we were at Embedded World? Was it Embedded World? Like in ‘18? ‘19?
When even was that? Do people get together physically in the same space?
I know, right? It's it's um, I think it was at ‘18 or ‘19. I don't know. Like the beginning of there. I was actually at Embedded World 2020.
I mean, yeah, yeah. It's um...
Nobody else was, I was there alone. It was just me and like people from China, because like, there's like everybody, like people from China came [to Embedded World 2020] and everybody said, “whoa, people are coming from China? We're out of here!”. And of course, I had just gotten back from China, so…
I see. I see. Yeah, no, I mean, I have no idea what the world's gonna look like going forward. Um, I think everywhere, everything is like up in the air and I think we're just gonna, you know, I don't know. Maybe we needed something to just like shuffle the deck a little bit and rethink a little bit. Who knows?
I've stopped trying to figure out something good. That's gonna come out of all the bad things that have happened. I just, you know, how do we deal with it? Right. That's my mindset. These days. I used to really try to find the bright side and I didn't take it. You know, how to take advantage, how to see how the world's going to be better because of something. And I've stopped, man. I've been beaten up too many times on things to think that there's any, um, there's, there's any better purpose for any of it, right. It's just, you know, how do we cope and move on?
Okay. So, uh, this isn't how I, I had thought the conversation would go, but I, I have, uh, I have, uh, have something I could maybe try to convince you, uh, might be a good thing that could come out of this.
Okay. So I was thinking about this, right. And you know, like there's a whole supply chain disruption, which we can totally talk about and you know, who knows what the next variant’s going to come and what's going to happen with that. And like all of the uncertainty about all this stuff. Right. And when like, things are gonna go back to normal. So I was just kind of trying to process that. And I was like, you know what, maybe we're thinking about it wrong. Right?
Maybe there's not like a return to normal or whatever.
Hmmmm, I think you're onto something there.
I was thinking about it from a hardware perspective. Right. Not trying to go to like broader, but I was like, what if the hardware we have today is kind of like a species that evolved under resource abundance. Right. We have like thousands and thousands of components with deep supply chains that you can get like, um, like that. And, you know, we built what we built.
But now like with all the disruptions, et cetera, et cetera, all of those assumptions are getting overturned. So it was like, what is, you know, what's actually going to come is we're going to have to lean down the depth of the supply chain we rely on. And that could mean doing more things with your PCB. That could mean doing more things in software, using more microcontrollers, uh, fewer dedicated ICS, um, more standard components. Um, more things like is that fair, you know, like standardizing the interfaces between things so that you can just switch them quickly, uh, you know, more modularization. So you can maybe like your wifi module can like be popped off and you can put another one on, like, I just like wrote about like 10 things. Uh, as I kind of thinking about that.
So much about, about everything that you just said. I think those are excellent things to dive into. Right. I mean, I'm a supply chain. The supply chain has always been a challenge, right? And Beagle[Board] has always been really supply chain focused right. In terms of like, you know, what we can do for the community. So, um, so that's absolutely been like, um, you know, thoughts, like how do we provide tools, but like, we're, um, like our, our demand, right, you know, with other shortages, our demand has more than quadrupled, right? It's through the roof. Right. Um, you know, and we, we, we thought we did a ton of planning ahead, right. With like, you know, um, uh, having whatever you ordered 52 weeks ahead for all sorts of stuff. And, um, the, uh, you know, things that you would expect to be able to get things that you had on order things that like, we're supposed to meet three times.
Right. You know, that it's like, this is, this is not like I ordered this last month. Right. It's, you know, this was a year ago. And, and so we're struggling to, to kind of just like, um, you know, meet expectations and when, when we could have like grown 4x easily, four or five more, um, and most of that's based on other people's struggles. Right?
I mean, it's just like, they look for, well, what is available Beagle’s available? Um, so let's go there and well is available, but only so much. And, and, and, but, but, but, but you talk about the world's dependence on, on different things. Right? You know, how do you, how do you build something that's really flexible enough for all the types of problems that people want to go and solve? Right? You talked about having, you know, modularity. Right? So there's, um, one of the, one of our, you know, I don't, I don't want to call it skunkworks projects because we don't really do things that quietly, we just don't really, like, sometimes we do little projects on the side, um, that we don't really publicize a lot, but like, trying, like there was this, there was this really cool thing: Project Ara. Have you heard of project Ara? (ALEX: Yes.) So right.
You maybe want to say a little bit for any listeners who may not...
Absolutely. So well, so you know about it, but, um, right. It was this idea of just um, modular mobile phones, right. So like a cell phone instead of just throwing it away. Right. Because the processor becomes obsolete. You don't have enough memory, the camera's old. Right. All those different things. Like you talked about supply chain, right. It's like, well, what can I, what camera can I get a hold of today? What, what, what processor can I get a hold of today? Right. Just like treating them all as more like disposable cogs, right. That you need to kind of put together a whole machine. Right. A display,right, um, you know, all, all the, all those different elements and to, to break them down and make all the different ones kind of work with each other. Right. And, you know, that's, that's a lot of the, the, the, the magic of, of Linux, but a lot of that didn't like feed itself forward to how hardware would actually talk to each other.
So project aura, right. Took kind of this, this power of like, what Linux is this monolithic, all encompassing, everything supported under the sun kernel. Right. I run on everything from, you know, on the, the, the, the stock exchange, right, to your Android phone, to right. Your, your thermostat. Right. You know, it, it can go all the way down to something like that. And, and it's all this one monolithic one software base.
And, and so there's the idea of bringing this, this, take this one software based. So when I'll make it, where hardware is just as interchangeable, right. So it can talk to all the different hardware through this thing called project Ara introduced this thing called GreyBus, where it didn't matter what the fiscal interconnect was, you know, they could just, you know, kind of instantiate those things. There's a low level of physical hardware interfaces and have them talk to each other.
Um, and so we've been, we've been thinking a lot about project aura and, um, um, you know, sorry, there's so many different, different ways that this kind of kind of impacts our conversation. Right. But, but, um, you know, with, with something like that, you could take all the individual components and swap them out.
But what is that, that processing element, right. That one's such a complicated beast, right. You talk about, um, or even the sensors, there's so many different little elements that even if you solve that problem of, I've got all the software that makes everything talk to each other, and I've got like ways for them to talk to each other. I've actually got to get those things. Right. And, and, you know, kind of what are those, those magic elements that people need to have. They're always available through our disposal and not just in, not just for the select few. Right. I mean, cause that's, that's, that's one thing is that, you know, we can get certain things in my hands that I'm just not going to be able to put in the hands of, of people, people in third-world countries, um, you know, people that, I mean, we deal with expert restrictions and right. There's such a political divide between countries right now and it’s growing.
Right. Right. You might not be able to in another moment as it happens.
Exactly! It's like, you know, it's like, because, because we're in like this, um, you know, you have these great manufacturing plants in Taiwan, but then you got to take things packaged in the Philippines. And, we're very globally interdependent, um, economy in order to make any of those different little things. But yet people throw up like just these crazy artificial barriers that there's no reason that we, as people should, should like create. Right. And, and, and we're creating, we're creating problems that we didn't anticipate. Right. Like, um, we're, we're creating our own problems that we don't anticipate. And, um, and I don't know, I'm, I'm, um, I said, I've stopped. I want to hear more about how you're looking at like, where the good stuff is that comes from this because, um, you know, I see a lot of problems to solve and that's great, but we'll go and solve those problems. But I, I, I, I feel like most of these are self-created problems.
I mean, you know, this could go all sorts of ways. I, I honestly think that we have overextended what the industrial revolution could do. Uh, we were like the guy with the steam engine, just everything looks like, uh, you know, we have a hammer, everything looks like a nail, right?
So we’re talking about the same thing by creating these re re usable cogs, right? Like everything is a computing problem that needs a computer to go solve it.
The thing with modularization though, is that you can customize, right. You can make the thing that fits for you. It's not like the one size fits all approach. And I mean, you guys at BeagleBoard have been doing that. Right. So you've got the AI, uh, addition, the, the, the, the blue, which is more for robotics. Right. So I see that sort of momentum, but at the same time, you can't be making a model for every use case under the sun. So, yeah. That's kind of the limits that we're hitting regardless of the supply chain disruption.
Well, I mean, but, but it gets right back to that point because like, one of our identity points is that we're open hardware. Right. And for us open hardware, it means that other people can get a hold of the components and the supplies go in and make those things. And like, you know, that's, that's the biggest to be somewhat the challenge for us. Like we have to redo our designs, not because we can't make them because other people can't make them. So it's because they can't get a hold of the components. Right. And because, um, you know, it's not our vision of the world that, that, that we, um, I mean, we certainly wanna, you know, take advantage of whatever supply chain and things that we have that would just satisfy our community's needs. But, but in the end, we really want to be enablers for other people to be inventive and to create their own solutions.
And, you know, because I mean, I think that everyone really around, um, you know, around Beagle really believes in the, you know, the multiplying, the lift that comes from, like, when you, when everybody comes up and write and be inventive, like I chat with some, some, um, I don't, yeah, I shouldn't drop names, but, um, there's some folks about the, the, the idea of like fostering an artisan culture, right? Like, um, like, like, uh, what does an artisan class or culture look like in today's society? Right. So we want to see more people learn about the things that might seem mysterious to them about electronics and, and, you know, um, software, like there's just different types of, of software, but like, this seems to be where the places are, where there's just so much growth and expansive of human knowledge that nobody can really wrap their heads around it.
Right. So, so how do you, how do you go in and get enough of it that you can go and build on it and make more, that you're really not just kind of subject to other people's knowledge. Right. You can actually build up enough of a knowledge base to kind of, um, grow and like, and how do you make a living at that?
Not just, it's not enough just to go in and make something as a hobby, but if you can't get it, if you can't get paid for it, right, then you, there's a limit to how much time you're actually going to be able to do. And there's a limit to how much time, how much solutions, if you can actually make feasible for the other people around you. Right. And, and, and, and, and even to the point of like, well, you don't want to make something that they don't understand. Right. I mean, we're surrounded by things we don't understand. And, and, um, I think for, for you and I, and maybe a little bit less, so, but I think that it's not too hard to find a little bit of sympathy for how mysterious things are for, for other, for other people. Right. I mean, right.
Yeah, yeah. A hundred percent. I mean, I don't find it hard at all to find, but then, because I tried to set up my home for IOT when my son was born a couple of years ago and I was pulling my hair out. Right. I've got a PhD in computing and I was like, I have, this is so stupid...
But you, but you, but somebody made decisions like that, and you don't know what decisions they made and what justifications they did for making those decisions. And, and, but, you know, because of this, this supply chain thing, right. The power of multiplicative purchasing power, you're stuck with their solution. Right. Rather than your own. Right. Rather than the one that you'd understand.
And, and, and, and some people are stuck with that for everything in their lives. And I am so sick of this thing [JASON holds up phone], I cannot stand how much this thing does not do what I would want it to do. Right. I can't build it. You know, these things are a thousand dollars now, and I can't build that for a thousand dollars. Right.
Yeah, exactly. Right. If you try to replicate, I mean, you, can't, no matter how much money you have, you could buy all the components, but you'd make something, make it twice as thick as, you know, last 20 minutes and like spits hot air everywhere or something.
And maybe you'd be able to make phone calls. Right. You know,
I have seen actually, there's a, you know, there's like a single-board computer, uh, phone. Um, and it basically is just about at that level, but exactly right. It's it's, and this is really my word. A lot of, I think he goes a lot with the time and maybe the maker movement and the 3d printing community as well, sort of pushing that direction. But like, there's more to go where, like, we need to take control back of like, making things ourselves, like not have it, like centrally designed somewhere, you know, and, you know, wait for the, uh, the, the airdrop to, to arrive with the goodies.
Well, I mean, even like there, there's so many things that like, you know, there there's, there's like for printed circuit board assembly, right. There's essentially like, like, I, I don't use a 3d printer anymore. I use desk stuff like 3d hubs and, um, Shapeways and, and, you know, it's like, it's, I, yeah, I would, I be less inclined to use those things if I didn't know that I could 3d print it myself, I would, but I'm not going to actually spend the time to 3d print it. Right.
And I don't have a pick and place machine and a reflow oven to go and make myself my own, um, printed circuit boards. And even if I did, I probably wouldn't do it. You know, I use the circuit hubs and the, you know, the, um, that there's, uh, I'm blanking on a number of the different ones that I've, that I've used. Right. But I, I send those off to, you know, online. Right. And, you know, just like you'd order from Amazon, I order my printed circuit boards from people that, that, that, that, you know, assemble computer parts and assemble printed circuit boards.
But, um, you know, that's not a, you can't assume that that always, that click of the button is going to work. Right. Because if they can't get the parts, they're not going to get you to the boards.
Oh, I'll show you something cool. Just completely randomly, just because you mentioned it, but, uh, I've been playing with this idea of, uh, copper plating, um, um, uh, conductive filament. So you can make something, this it's going to be seen. I did this in my house. Like I basically just 3d printed on, on conductive filament and then put it in a copper plating rig. And like, I've made circuits that work, but, you know, I hear it was just fun because I just had it there. But, uh, um, yeah, and the more people that have that we can pass it, the, of what we have available, that's kind of my, where my mind is going, because this sort of, you know, deeper and deeper and deeper and deeper supply chains kind of, it almost feels like a Jenga tower that was bound to fall at some point. But like this thing, you know, the, the virus came and just kind of knocked it over maybe earlier than it would have otherwise, but I just can't understand how it, you know, where this was going to, like, you know, at some point it would have hit a limit somehow.
Yeah. But, you know, it's like, um, some time ago there was a capacitor storage. I don't know if you, if you remember the capacitor.
I don’t know, please. I want to learn.
Back in, I don't remember. I don't remember exactly what was driving the capacitor shortage. It wasn't all that different from today's, you know, semiconductor shortage. But, um, what I was going to go ahead to was, um, like I know how to build a capacitor, right? I can take two metal sheets and I can put it insulator between it. Right. And I can roll it up, you know, but that's pretty different than the, you know, um, the, the, you know, a 1206 in these, these tiny little, these tiny little packages of, you know, of, you know, very controlled, uh, capacitance, you know, things that it's like, knowing, knowing how to make a capacitor and having a hundred capacitors for putting down on a circuit board, um, you know, when you need them...
Yeah. I mean, there's, um, I don't know, like there's a, uh, there, there are pros and cons of globalization and centralization, you know, having like a full-vertical stuff. Um, it just, the quicker we can, um, both incentivize, um, like building those, those total vertical solutions so that there is, um, that opportunity for continuity supply chain, as well as eliminating the barriers to just working with everybody nice. Right?
And, um, um, and we're, we're not really doing either one right now. We're not, we're, we're not enabling people to build the real vertical solutions because it's like, yeah. I mean, we're doing a lot in the maker movement to kind of like solve the problems of things that we can do at the, you know, hundreds, hundreds of dollars, thousands of dollars, tens of thousands of dollars. But the, the, some of these things like required, like for a modern semiconductor fabrication facility, it takes $10 billion to go manufacture one of those. And it's great that TSMC is building one in the United States. That's great, but it doesn't solve this great disparity between the manufacturing capabilities of different countries, nor does it dissolve, nor does it does, nor does it solve the need to kind of deal with the the way we treat other countries as, as somehow more or less entitled to the things in the world.
The thing that's in the way, the capacity, right. Like, you know, I said, well, “that's kind of mine.” I mean, we've seen that with, with vaccines as well, the same way.
Completely, right? Right? It's like, this is highlight as a priority, so I'm going to make sure that my people get it right. So, and that's, yeah. Um, I I'm, it's not that I don't understand it. I just don't agree with it.
We're on the same wavelength-- the thing, um, that's on my mind is that you can't avoid that. Right. So you, the only thing you can do is make sure there is as widespread production capacity as possible. So that's, yeah. But an issue, because if it is an issue, then somebody will turn off.
You know, I, I used to be very optimistic that, you know, like there's like Ricardo's Law of Comparative Advantage, and, you know, everybody was specialized in their own thing and we'll be beautiful, but the idea there is economics is so strong that the politics would stay out of it.
I thought it was odd that we actually started to see some shifts where that was starting, you know, in some ways that was becoming true. I mean, there are corporations whose interests outweigh those of you know, national interests, right. There are, but, you know, but even those, right. Um, even that much power concentrated into a corporate interest, they become essentially like a national interest. Right. They, they, they, um, you know, it's, you know, you, you always have to keep the people oppressed by the, this, this one entity, you know, happy. Right. If you don't keep them happy, right. Then you can't stay in power. Right. And, um, and that happens the same for corporations. Right.
If, you know, if Kings tell their, their, their subjects, right. This is the way it shall be. And as long as they can live with that, you know, that, that they won't revolt. There needs to be a bit of a technological revolution around some of these, these, these powers. And I don't mean that, um, you know, cause there's a lot of people that want to tear down like, um, patent law, for example. Right, right. And, and, and tear down, um, you know, protections that are meant to have inventive people, you know, um, you know, be able to benefit from, from there. Right?
So you, you can't, you can't tear down the means of production. Right. You can't, but, but, but Ayn Rand is not right. Just because like it's like believable. Right. You know, it's like, um, it doesn't make it right to, to not respect the needs of, um, the people, um, oppressed by these limitations, even if they don't understand them. And that's, that's the part that I think is the more fundamental is like making people understand, um, the limitations that they're, they're put on by not being, have, have access to certain levels of technology.
Um, and, you know, having everything be in, you know, like the, you know, the folks up in Portland, just falling just a little bit behind, but having everything in Taiwan and everything in, um, you know, in mainland China, it's, it's not against, it's not anti-Sinocism. Right. It's just, you, you need to have multiple sources, right. You need to have some redundancy.
I actually saw if, um, if anybody's seen the, um, the Tesla battery date, um, I actually saw sort of their road, the roadmap of their describing, I think, I don't think they said it explicitly, but you can sort of read through the lines.
Basically they were building supply chains in every major location. So Europe, Asia, uh, United States, um, and who knows if they're going to add more zones, but in that way, the, the, the, the capacity was going the other way. So they're replicating what they have in the U S in China, um, to work with Chinese sort of bottom up, you know, they partnered up with CATL for batteries.
Um, but, uh, it's not a, you know, pro some country or anti some other country it's that I think honestly, like, uh, we have the saying in Greek, you know, “good accounting makes for good friendships” and everybody's got their own, there's no reason to kind of fight, uh, for, you know, to get somebody else's and then maybe we can exchange information, the not sort of, uh, blackmail each other on, like, I've got the thing you need, if you give me the thing I need, or like that kind of tensions.
I dunno, that's kind of where my mind is going with all the situation, but like, it's, it's, you know, I, I can't say it's crystallized for me, but I, I just can't stop, like, thinking that, you know, if we, if every country had like, or country whatever zone, right. You can, it could be smaller than country. Um, had like some approximation of like, you know, the bottom-up capability also, like sometimes you get conversations happening between expertise that would not be happening if they were across different oceans. Right?
Like to, to sort of like, like software and hardware, like you guys are, are, are working on, uh, where, you know, you're, you're playing on that boundary, right. Of what is software, what is hardware, et cetera, et cetera. If these people are living, you know, in different, like, you know, uh, continents it’s much harder to negotiate that boundary on the, on the architecture then if they are living next to each other. Um, but that's, you know, again, super abstract and, and, and sort of, you know, I guess like you have been sort of, you know, uh, with, with the whole pandemic, just my mind is trying to figure out what, uh, you know, where this whole thing is going and how we might sort of absorb it and become better through it. But I just, I guess I'm just an incurable optimist.
You know, I I'm, I'm an optimist as well, believe it or not, it's just that, um, you know, um, you know, I don't think we're always moving forward. Right. And, and, and that's okay. Right. But, um, if we, if we move, if we move backwards, some it's not the worst thing in the world. If we can just, you know, and, and maybe, and maybe this is, this is what you see as kind of an upside, right? If we can learn from it. Right?
But I think it's, uh, it's our choice to make an upside. I don't think it's a given that we're going to learn from our mistakes and, and learn how to do things, um, and how to do things better. I don't think that's a given, I think we have to work for it, um, and,
The maker movement all over again!
And, and, and, you know, what does Maker Movement 2.0, look like? Right. Um, you know, I, the, the, you know, Dale did wonderful things and I think, you know, you know, people can have some frustration with them for not like not helping us, like, directly get into what the next phase is. Right. You know, the, you know, let's identify the maker movement. Right. You know, uh, fostered along, but, but how do we, how do we, um, go in and, and th the, the shared interest of the individual and the, the, the corporation, right?
And it's, it's not, um, it's not one without the other. Right. You know, there's, there's, there's capability in the, in the corporation and the, the assets that they have that are, you know, billions of dollars and what those give to the society. I don't think the answer is to, you know, it's not to be all communal about it. Right? I think that they, they, those, those corporations have have rights to those things, but how do we best enable them to enable us, right. As, as individuals, right. How do we, um, how we mutually benefit from investments that people have made before us, without just being solely focused on undermining them. Right. I mean, the, um, you know, TSMC is not a flash of brilliance, right. It is. It's just like, it's, they've built an amazing organization there. Right. And that the world depends on, um, you know, it's, I don't know, it's, it's, it's all encompassing, but how do we, how do we break that problem up in a way that we use as, um, individuals can, can contribute to solving it?
One way to think about, not the way you said, it kind of made me think, like, you know, the maker or the individual, or the experiment, or like, even maybe some startups, I think of them as divergence. Right. You try all sorts of things, right. Things that a corporation would never try. Right. Because they don't make sense because if they're not obvious, because, because, because.
Um, the large companies that is the convergence, right? They, they, if something pops up, that's really good, I'll get like, um, you know, absorbed and, and, and, and, uh, but the, I mean, the problem of course is that, that means that certain structures keep going, right. Like, they're not, maybe something could be good, but it doesn't fit and then get thrown on the sidelines. So there's, it's not like a perfect model, but, um, there, there, there, there are ways that these two worlds can coexist.
Um, there's also the, uh, so, uh, this is something I've been thinking about a lot, so there's the thing called. Um, so, uh, I think it's called Ronald Coase? Coase is definitely his last name. Um, he has this Theory of the Firm, right? Which the, the, the question he asked them in his sort of seminal, uh, work is, you know, if markets are so great, why do we need corporations? If you think about it, right? Like if we, if everybody could just be like a sole trader, essentially, like, and we could be in one massive market, you know, what's the point of a large corporation?
His answer was transaction costs, right? It takes, it's very complicated to, um, coordinate, you know, 10,000 little businesses to produce a CPU. It's much better to do what is essentially, you know, a black box of like socialism with centralized control, you know, within a market economy to build your CPU.
Um, but I think as we make these technologies and the maker movement does all of these things, um, you know, technologies are coming in from all sorts of places, um, what the individual can do increases, right? Like we, even, what, you know, we, one person can do today was unthinkable 10 years ago. Um, we've got, you know, small startups, even far smaller than SpaceX going to orbit. And, you know, that was a state actor job, like 20 minutes ago.
I know I know. Just a blink of an eye and all of a sudden private industry. And, and there is a degree of democratization that happens that, right. That the, the, the, the one you cube satellites. Right. Right. I mean, I can't, there's so many, there, there are a lot of Beagles in space.
A lot of beagles in space. Um, so cool. And, sorry, go ahead. Good. I bet that there's um, but there's a, um, like some of these things, like th th th th the way that people shift, like aggregate buy, right. So like, um, um, like designing that there, there there's, um, like some of the, some of these large, um, asset organizations have shifted their models to some degree. And I think they have a lot of learning. I think this is something that would really matter to push forward more, but like, there's basically all the large semiconductor companies that, um, you know, a lot of them have things called shuttle services, right. So a shuttle service it's, you know, it's very expensive to make a Silicon mask, right. So the set of radicals and things that you use to do the, the manufacturing of a, of a piece of Silicon, um, it's very expensive to go and do the photo lithography to create these, these very large, right.
Some of them are the size of rooms and have features, I don't know, really small, you know? And, um, and it costs a lot to go to go make these things well, there's limited supply of who can go make these things. Um, but there's even a more limited there, there's also a very limited supply of people that can actually do anything with it to actually get went to Silicon.
But the good news is that you can do like this kind of group buy, and you only have to buy, you know, this big of a space on the reticle. And you share that space with all these other people, right? Much like, you know, the, the cube sets to get to space, right. I mean, for $10,000, you can launch a piece of hardware and a space these days, um, right. 10 centimeters by 10 centimeters, by 10 centimeters, you get a block, you know, there's some interface for, for power and for some, some data collection, and you can put something in space and, um, and, and, and run experiments and collect data and, and all these other things.
Well, the same thing for making semiconductors, you could use shuttle services now, and there's, there's people that are even sponsoring to the point where if you design a chip, you can get a free shuttle service, right. There, there are people that are doing sponsored shuttle wafers every now and then, um, you know,
Speaking of which, um, you guys have been, uh, sort of, uh, experimenting with the RISC-V chips right?
Absolutely. We think it's really important to move risk five forward. Um, you know, all of those supply chain things right, are, are huge inhibitors, right? So, uh, there are a large number of RISC-V uh, microcontrollers that have been fabricated. Part of our identity is Linux, right? Very, very capable Linux systems. Um, you know, we, we did the, the, the Beagle Five-star Light project, um, which unfortunately we had to, to, to, to reach an end on, um, but we were doing a lot more with RISC-V and Beagle V, right?
So we do have things in the works that I'm not quite ready to talk about, but we are very motivated to, um, to move RISC-V forward for our community. Right? That's the thing our community tells us is they want an open instruction set architecture. They want, you know, to open as, as, as much as they can, right. RISC-V itself is not, does not mean it's an open source hardware design in the Silicon itself. Right. It's just an open source instruction set architecture, um, and there will be closed source CPU's and open-source CPU's, um, both of which are a lot of interest to us in that we're gonna work with our community to try to, you know, give that group buying power sort of thing to, um, to the embedded Linux community.
Actually, uh, if I remember correctly, the last time we spoke, um, the BeagleBone has, you know, it's sort of tip of the iceberg, which is the BeagleBones you guys produce. Right. If I remember correctly, he told me that was like a far, far greater sort of implied, uh, capacity that for people who use your designs.
Yeah I mentioned something about that earlier, right? So, um, you know, you know, I talk about trying to, to, to kind of enable a, uh, a bit of an artisan class, right? I mean, our intention is right. That people, we do open hardware and we don't, uh, you know, we don't approve these, the BeagleBones themselves for use in designs, but if you want to validate it and use it, like we're not going to call the police on you or not, you know?
Um, but, but you know, a lot of times people reach out to people in the embedded Linux community to go and create something custom around it because they're open hardware, right. There's first of all, there's a number of compatible boards where people that will do, you know, customizations will actually release some kind of standard products, um, you know, Seeed and SanCloud, um, kind of stand out and not work in the legal world, you know, as folks that are willing to do custom solutions, um, around BeagleBones that, that are already doing something that's available off the shelf.
Um, but then, you know, other folks will, will, will help you do like other full custom designs. And, most of the time, like if it's big enough, right, it's probably going to be done in-house anyway. Right. Somebody's going to have some in-house board capability.
And so those are the people that bug us for, you know, fixing the fact that our micro SD card cage went end-of-life. Right. And, you know, and, um, even though that, you know, we could still keep getting it, nobody else could because they weren't doing enough volume to justify going and getting those SD card cages. So like, we, it took us a long time, but we finally got through full FCC and CE certification and everything for a new micro SD card cage design, right. From, from folks. So now people can, can, can use that. Um, but you know, we want this ecosystem to have kind of all sorts of jumping on and jumping off spots. Right. Because, you know, I think we'll only hurt it by trying to control it.
And again, right. That's the power of divergence and convergence, like people try different things. Has there been a situation where somebody tried something that you guys thought was really interesting and wanted to sort of, uh, you know, merge it in or, uh, somehow learn or get ideas from your community about what the next sort of iteration is going to be?
Um, there's a number of small, uh, small points to that. Even the pocket beagle, right. That came from, uh, a project that, um, um, Michael Welling did. Um, and I think some of the confusion points get to be around, like, what is BeagleBoard because we need to try to keep the BeagleBoard identity, you know, somewhat separate if it's an official BeagleBoard product.
Um, a number of people do things that aren't nicely branded as very separate activities that we would like them to be. Um, you know, we'd like it to, for people to be clear about that, but, um, yeah, Michael Welling did the original, um, you know, uh, he called it a PocketBone design, right. And the neck, it became the, the PocketBeagle. Um, so that's more, I'm like more of like a direct board creation.
Um, there's been, uh, well, even the, the BeagleBone blue, right. That actually came out of a university of California San Diego project. Um, for all of their mechatronics courses, they were coming up, they were building a BeagleBone Black and an, a Cape, um, right where they were, they were teaching, um, you know, mechatronics like controlling motors, controlling, you know, um, they were green quad, copters and rovers and, and everything off of this. And we just combined it all into one to try to help with the overall cost. Right.
You get it into one board, you get it out to more people, you get into distribution. It goes out to lots of places, it becomes readily available. It becomes a cheaper solution for everybody. Um, and, um, so there's been a number of those places where it's kinda, um, uh, come back into, you know, kind of a, a beagle standard product, but I'd say most of it's more, more subtle than that. Most of it's more, um, like guiding an influential, um, things. Right. You know, and, and people, you know, letting us know where they're unhappy.
I always have more projects to show if you want to go into, to, to Beagle stuff like that. Yeah. He's got stuff sitting on my desk that we're, we're working on because we're always working on new stuff. And some of it's guided by the community. Some of it's, um, more my, my personal or local people's kind of imagination, but, um...
If you've got something on your desk that you can show us and we were not going to...
I always have stuff I don't, I don't mind showing people also from my desk when, when they, when they start chatting with me. So, um, so this is, this is, uh, this is a project we're working on. This is going to be a fun one this year. This is when we talk about the project art stuff. This is a prototype for a thing called the beagle connect. So this has a, some gigahertz radio that'll work a kilometer away, um, from the gateway. Um, and so it's got the um, Mikrobus connector on it.
So Mikrobus is, uh, I almost say that's an open standard. It's a, it's a, a standard coming from MC electronica, but it's freely licensed. Like, so anybody can go and put one on there if you put the Mikrobus logo on it, right. There's a little bit of, um, restrictions in terms of what they want to see on the silk screen, but they don't charge you anything to go and use it.
Um, and they're kind of a, um, you know, a nice kind of vertical solution and that they've got a thousand different boards that you can add on. Right. And nobody else in the industry, I think has a thousand different things, you know, adhering to, to kind of one standard, right. But you just, you kind of plug in wherever your sensor board is on this, and you can put this a thousand meters, a thousand meters away from your gateway. Um, and it's nice and low power.
But the magic of this is that we're turning the software from Project Ara, it's running GreyBus. So the peripherals on this actually show up on the Linux gateway, right? So this is a I2C bus a UART bus, GPIO um, SPI, analog, right? All the buses on this thing, um, show up as peripherals on the Linux gateway.
Um, so we don't need to write drivers on this. This just comes to the fixed firmware load. Um, and whenever you hook up to it, right, it just deals with it on the, the, um, on the gateway and the you know, the MikroElectronika folks have been nice enough to, we've worked through a little bit of a standard with them on how to do the identification. So these can eventually be plug and play right today. You'd have to like tell the system what you have connected to it, but, um, you know, or working towards, um, is that it will, um, it will self-identify. So you just plug it in the, and it just starts collecting the data.
I could just put a bunch of them...
Exactly. You want weather sensors all around your farm. Right. You know, you check the water levels. Um, you know, you wanna check the, um, you know, what I've been doing a fair bit of as air quality monitoring, um, with these. And so, yeah, whatever you want to throw it around the building, throw it around the farm. Um, you don't have to run wires to at, you need to get some power to it, or you can run off of a battery. So that's, that's, that's a, that's a, that's a cool one. I think the software is really the magic to it. We're about to innovate on it. We, we we're, we're not, we've been limiting our talking about it a little bit. I talk about things all the time, but I've been discouraged from talking about things too much and we don't have a release date.
That's all right. Yep. Keep them guessing. Uh, I mean, we, yeah. We've, uh, since, since, uh, yeah, I mean, we've been, we've been, uh, you've been, you've inspired us to start playing the similar technology, so yeah, it's, it's, it's really, uh, cool.
And, um, I'm hoping I hear more about you leveraging that.
Well, the, for, for us, it comes back to the problem that you were touching on before, right. If you make a thing, like how do you go from idea to production at scale? Like, how do you, like, um, let's say somebody up their basement and they have a cool idea. How do they turn that into like a company that pays their salary and maybe, you know, could become huge?
If, if it was a software idea, right, like I could just launch it on Amazon and, um, or, you know, some other cloud and, you know, I don't have to necessarily exit my basement. Right. If it's a hardware thing, like let's talk about logistics, uh, certifications, you know, it feels to me like harder today is where software was in the nineties, right. Where you basically needed permission from like a number of gatekeepers to even get started.
Um, so, you know, balena is always thinking about that problem and like, how do we make it easier? How do we lower the barrier? So a big part is it's very close to what you're saying is just that the distance element is not even in our radar, like, what we want is for people to just be able to throw together the right components and for all of that, to be immediately visible on the kernel level and, uh, above, um, when I’m able to have modular electronics, even in the same sort of package, uh, not, not package like a Silicon, but like, sort of, you know, case, let's say, let's just put, you know, like the PC is where you can was, I guess, back in the day they got a bunch of cards or whatever I got an idea is not, it's kind of lost almost.
And, and, um, and, and not like that, you can't do it. It's that you hit the boundaries about kernel modules and a recognition of the higher levels of the software stack. And if you are like a gray beard, uh, you know, uh, electronic engineer that's, these problems are solved. Right?
My beard’s a little greater than yours is.
Yeah. Or if you're, um, you know, some, you know, 20-something with a cool idea. Um, yeah, these are real barriers. Right. And, and that's where my mind is at a lot. And that's why I, um, yeah, I keep thinking, like, you know, we'd need to be as easy as web design. Right. And web design is really easy, but this stuff is like, gets, you know, it's kind of like, uh, you kind of walk around and it's cool. And like, you just fall into a hole, like before, you know, it, you're not looking at like kernel source codes or something, and we just have to resolve those issues, make it easy to make something that is like reasonably custom. I'm not talking about completely custom, but like, if a thousand modules is incredible, right? Like, can we get that to be available easily on a single-board computer?
Um, and people can mix and match and make the piece of hardware that's right. For them that, you know, because, so one thing I'll tell you is whenever we hear a customer of balena saying, we're going to develop our own custom board, I, you know, like the rolling, uh, clocks,
...like you just put it into add another year to it? ALEX:
Two, probably right? It's like, you've got to avoid that. Right. So if you, if you can fit within the resolution of the modules that exist in order to avoid buying stuff that you don't need, right. Because people will go custom a lot of the times to add something and sometimes to remove something, right. Like, but if you had smaller modules, you could approximate the, the, the boundary you want, uh, without building your own custom thing.
Um, and then you want to make it programmable, uh, which is again the same problem. So, so really all of these things come down to having, I mean, the kernel module really is a pain and you know, what else is a pain with kernel modules upgrading the kernel remotely, right? Because let's say you got it all working. And then we send a, was that an update? Um, the, the, the, the kernel module interface, uh, for Linux is a very interesting part of it.
So the, yeah, the, the API standard for like the kernel is there is no API standards. Um, you just work within the version that you've got, if it works within the kernel, that's the API standard, right?
There, there is no API within the kernel as a monolithic beast, probably.
So that like the right way to work with the kernel, right, is if you, if you want to leverage an interface, make sure that whatever your code is that leverages the interface is actually in the kernel itself, because then when somebody breaks it, they're responsible for fixing it.
But you were getting into a point that, that to me is fairly fundamental, which I believe in paying for, for atoms and the electrons are free.
And, um, and for the, the electrons like really should be free, but you have to remember, I'm also balena customer, right? So, and, and you know that, but, but why am a balena customer? I'm telling you about all these, um, you know, electrons being, being free, and we're just paying for atoms. So the fact is, is you've got atoms and server space and it's like, okay, well, it's not that because, um, you know, I have plenty of like servers, web servers that I can, you know, that are already up and running and doing different things that I could get leverage that, but organizing those electrons and all the right ways to make that hardware just work, right.
What you've got put together in the cloud services is a real value. Right? And that's kind of a lot of what we see with, with a lot of the stuff with, with, with Beagle, right? Is that, you know, we, we work with the upstream Linux community as much as possible. They, of course, you know, they'll, they'll always think we can do more. Um, but you know, we, we, we do what we can and we try to make sure that there's good upstream support and that there's, you know, the, the, all the, the kind of the softer bits. Right?
And it's, it's, it's value added hardware, right. It's software value added hardware, right. We're yeah. I mean, ultimately it's the software that makes the hardware so valuable, but, you know, I, I believe in paying for the hardware. Right. And that's what I do with, I may be a little bit different than what most of your customers think about, but that's kind of the way I think about it with balena. Like, I'm paying for your server time because you've done the hard effort of making that hardware value with the software. And that's, that's a lot about the way I think about the beagles, right.
I mean, what makes the beagles super awesome is all the software that runs on them. And that's something I learned as a child. Actually, my mom taught me that, right. It's like that, like, what can, what computer do we buy? It's like, well, you buy the computer that runs the software you want. And, and, and that's something I learned as a, you know, an eight year old from, from, from my, from my mom. Right. And, and that, that has stayed true for me, you know, to this point. Right. It's, you know, the software is what makes the hardware valuable, right? Any, you, you pay for the hardware that runs the software.
And I, I see, I want the software to be open source just because it's open source doesn't mean you don't pay for it. Right. You know, you, you, there's, there's there's effort that goes into organizing all those electrons and making them all work in the, in the, in the, in the right way. And I'm perfectly okay for paying for those atoms.
I mean, the one way I've thought about the whole sort of “balena equation,” uh, just because you, you kind of touched on that. And I think it's an interesting way to think about it sometimes, uh, you know, pricing, conversations come up with customers. Right?
And I mean, pricing conversations are notoriously difficult for a thing that isn't “atoms.” Um, the way I, I often explain it is like, okay, let's say, you know, I'm on your side and I, to replicate balena for you, like, you'd need a team of, I dunno, very, very, very generously, like, you know, four people, like, I think it's more like 40, but let's say four. Um, right.
It's more than four.
... for a year or two. Right. Um, like whatever, the running rate for a full stack, a full, full, uh, fully sort of accounted for engineer is, um, for a couple of years, and that you need to assume that we'll succeed.
Let's take out the risk equation that this whole project will collapse if it is there. Okay. So, and let's say they succeed, right. Just like all of it goes through about, okay. And you need to keep paying them. Of course, because they, each one of them knows where, you know, how each thing works, you know, four or five, six, whatever software engineers, at balena is enough to pay for, you know, I don't know, 10 tens of thousands of devices. Right.
And if you get to tens of thousand devices, you're going to need more engineers because that thing is going to be scaling. Now you have more problems. So I actually see it as a timeshare on the balena team, rather than the servers or the software.
Um, you are getting, you know, these minds that are struggling with these problems day in, day out, they've seen things you would not believe they've solved, you know, things that you will have no idea even exist. Um, anyone. Right. And, you know, like how do you get the HCP on a random person's house obvious, or completely impossible to figure out it took us a couple of years. Right. It's salt. It's just, it just works now. But, um, yeah.
And, and, and I know that I could personally go and solve each of those individual problems, but just because I could, and just because the software is out there to do it, the time it would actually take me to do it is tremendously more expensive than just leveraging your solution.
And this is really what connects it with what you guys are doing. Because the, the, to me, the, the future that I want to see is not where we build platforms to control is where we build platforms to enable. Right. So being alone isn't a beautiful platform to enable people's creativity. We want Belinda to be the same thing. We want to absorb all of the complexity and we want it. We want it to look trivial. Right? Like sometimes again, it blows up in our faces because people are like, ah, why would you pay for you for this? It's like, okay, go do it and come back in six months.
But, um, the, the, you know, same thing as, as the view, like, it's like a guitar, right? Like you don't think about all the complexity that goes into making it, like, if you're good at it, it just, your brain sort of absorbs it as a body part. Right. It's just, you're thinking guitar sounds right. But you want to produce, and it just, they just happen. And I think that's a great platform that's successful, uh, like, like what you guys are building people, you know, it becomes a mental model for hardware, right? Like, they're like, okay, I'm gonna start with a BeagleBone, I'm gonna add, you know, this or that. Especially if the Beagle Connect stuff works, I'm sure there's going to have the same sort of, uh, characteristic for people.
Um, and, and, and that's how we move people forward. Right? We, we, we create pathways for them to channel their creativity, to whatever medium that each thing represents. And, and, and all of this stuff needs to work. Right. You need the hardware, the software, you need the connectivity. Um, and, and you know, who does, and then you, you start, you starting right. Then, then you're like at the starting line to actually, you know, uh, make something valuable, somebody.
So there's so much, um, yeah, so, so much that a hardware startup faces that, you know, anything we can take off their plate. Uh, anything you can take off their plate, uh, I think is totally worthwhile and, and just takes humanity forward by my account. But it's just a daunting, daunting task to, to, to try to deal with.
What it means to take it forward though, I think deserves some, some definition. And, um, I think there's a lot of moving forward that we've been doing. It's not really moving forward, right? As a, as a, as a technological society. Right?
I think that what we've done to the open internet and what we've done to like walled gardens with the Facebooks and the, um, you know, these things of the world, right. I mean, it's, it's really moving, moving backwards. And, um, and, uh, so I think it, you know, when we do things to try to enable, right, we gotta make sure that we're, we're actually enabling and not creating the same sort of like, um, walled gardens.
I mean, I envisioned technology working more in our daily lives like appliances. And I think that, um, you know, balena does a decent job of respecting that as well, because like, you know, you, you worry about making it too simple, right? And validly. So, because people don't appreciate the value, but that's exactly what you should be doing because it should just work. And you shouldn't have to think about what it's doing under the covers, or it's not doing its job, right. It's supposed to provide that, that degree of abstraction and hiding from the details in order to just work.
And what I care about is like, as somebody that wants to know more is that there's all the different possibilities to peel back the layers of the onion and kind of understand how it really is, is working. And that's, that's, that's a sensibility we, we try to achieve at, at, at Beagle. And we're, we're still working on some of the, the sides of, you know, it just works, but I think we've done a pretty good job at enabling other people to make things that just work and they’re appliances.
Right. So we have, you know, PCR machines, right. Um, that, you know, they don't think about, oh, there's a beagle bone inside of them. Right. They just, they they're, they're doing their analysis to find out if there's, there's, COVID in a sample, right. Um, you know, people building laser cutters and 3d printers and all these other things-- they, they just, they like work like what they are. And then, you know, people shouldn't be limited and faced with technology, right? To me [JASON holds up phone], this is the worst experience of technology, because it tries to shove everything into one. And so this becomes a constant distraction device, right. It doesn't do what I do when I want to do it. Right. It tells me when I should be doing things. So what the hell??
Well, this is actually touching back on something you said before, um, about the Linux kernel. Uh, so, so you say like, you know, if you want it to be taken care of, you have to put it in the kernel. Right. Um, but in my mind that it has to be, you know, that takes time and it you'd have to be working within a limited domain of uh, peripherals.
And that makes sense, ish, for a server, the desktop, a mobile phone, even in mobile phones, Android has had to really like bend the system. But for IOT and edge computing it’s just, unthinkable to me that we would be putting everything in the kernel. Like that's just not going to flow.
So the, like when we, you know, the, the thousand different boards that, um, micro electronica has today, we took 150 of them. And the code's already in the kernel, all it needed was the configuration information. And the next hundred are small feature fixes right now, the next seven 50, that's going to take some crowdsourcing. Right.
And, but, but what we've done is we've provided a template for which that crowdsourcing can be done upon. And I do think that the, the, the, the Linux kernel can be made perfectly suitable for IOT. Has it been geared for IOT to this point? No, it really hasn't. Right. I agree with you, right. You're, you know, there's a lot of IOT problems that you're addressing outside the kernel that simply don't have a mechanism for dealing with inside, inside the kernel Um, there, there are a lot of things that are in the kernel that you're having to deal with daily, for kernel modules support, and like kind of dealing with things out of tree.
Um, but I think that those are solvable problems. I think we actually can get those things into this, this single knowledge base. You know, my, my personal tastes with, uh, with Linux is I feel it's a little bit of spaghetti code and, and, and a little bit hard to read, but guess what? I know that my opinion doesn't matter because there are way smarter people than me. And, and, and there's way more knowledge in that code base, um, than I can extract and, and build upon, um, that just really makes, um, my own personal sensibilities mute, just moot, sorry, move, not mute.
Um, and, and I think that, um, um, I mean, I, I, I just, the Linux kernel to me is just this, you know, Godsend a history of computing, right? It tells you why all these things are there because of the way that the Colonel gets maintained. And it's not the code base, it's so magical, it's the maintainer ship and the way that the community works with each other and the spirit of the community that does make this thing absolutely magical.
And honestly, to me, if, um, the, the way in which you've approached, uh, the Beagle Connect, um, problem and solution, um, it would, you know, it, it really comes into a very, in a sweet spot for, for the kind of friction we're seeing. So, you know, it'll be super exciting to see this, um, go forward. But, um, as we've been talking, you kind of mentioned, you know, like your earlier interactions with computers and I can't help, but like what, like when, or if you, do you want us to sort of tell us, like, okay, sort of where here it's 2021, but like, the BeagleBone is like a mainstay, like everybody knows the BeagleBone, right. Like, so how did what's, what's the story there? Like what's, how did all get started? Like go as far back as you want?
Well, I go back to my start with computers. That's, that's honestly, um, when I got involved with computers, that was the seventies. Um, um, and, um, you know, w you, you had a floppy disc, right? You didn't have hard drives. I mean, okay. Honestly, I started with cassette tape, but let's just start with the floppy disc and make things simpler.
And, and, um, like my mom, you know, she would take her floppy disk with her important work stuff, and she'd go, and she put it in the safe, right. She'd have multiple copies. Right. And she put it in the safe and she said, I could do anything to the computer except open it up that took a few years. Um, and, and, and I could do anything I want, and it was, I found it empowering. Right. You keep using that word around, around balena.
Right. And, and, and that's what you wanted to that's, you know, we, we want technology as a whole, right. As it, as it lives and breathes to be empowering for people. And I felt, I felt personally empowered. Right? I, I, you know, I was just printing my name across the screen and loops, and it's like, “it's saying my name on that screen. Ah!”
You know, and I don't know what it takes for the younger generations to kind of feel that same sense of, of empowerment. You mentioned you have like a, a two-year-old, is that right? Yep. Um, you know, I've got a five-year-old now, but Beagle's older than that. But, um, you know, I spend that my time thinking constantly about what is it, what does it mean to do something that would make them feel empowered? Like I did as a, as a, as a kid, right.
I could do it type anything I wanted into that tag. And there was no limitations. I didn't have to worry about just deleting the family photo album or, you know, like, you know, uh, we got a modem later. You weren't worried about what was going to happen to me online because I was spending too much time on the computer and, and, and all these things.
I was able to just, just explore what I could do through having a machine do things for me, rather than, than having to do them than myself. And, you know, I started applying things to all sorts of things in my life. Right. How do I automate things? How to simplify things in my life by making a computer do them for me, right. They love doing repetitive tasks, they don't care. Um, right. And so I love just finding repetitive tasks that I can just make go away because the computer will do them for me over and over and over again.
And, um, so, um, you know, Beagle was, was, was, was born of trying to, to, to provide that same trajectory of empowerment. And, you know, we're in a world where, you know, there's Halo. Uh, there's like just, you know, uh, you know, the what'd, you see what computers can do these days, it's not approachable.
Especially for, you know, there's so much to know, you're just not gonna know it overnight. Right. But, but, um, but making the, making lights, blink, making things happen tangible, um, I think is still a little bit, uh, enigmatic for, for, for people that, and, and, and you can make that mystery go away, but actually writing the line of code that makes a motor move or makes a light turn on. Right. And, and you understand that there are electrons flowing in the directions, uh, you know, in certain directions of the, of, uh, of a wire.
Um, and you know, it was really the, the, the change in the cell phone industry that brought Beagle around, um, because, um, the, the, the chip sets were no longer being designed by third parties. Right. You know, folks like Apple designed their own, their own, their own chip sets now. Right.
You know, the, the, there's, there's this, this, this fundamental shift and at TI who was the world's leading manufacturer of cell phone chips processors, um, at the time we did Beagle was really looking for new ways to apply, um, processing technology. And, you know, we thought by, well, I thought that there's this, you know, there's this great Linux community, you know, they're always looking for veteran solutions for making, uh, things with embedded Linux that they can actually get a hold of and, um, that they can actually change. And they're actually empowered by not just used as a, um, a momentary cog. Right.
And, and I thought we could go and address their needs. And, um, and, um, there was this great community called, um, um, um, Maemo, um, I don't know if you've ever heard of that.
But then the Nokia tablet stuff, you know, if the N 900, it actually came out sometimes sooner Beagle might not have existed? Um, but, but like the 900 was way too slow. It was way, way behind, um, and like, like that, um, addressing the needs of that community, um, by giving them, uh, you know, uh, a single-board computing solution where they could go and address all the software needs of the emerging embedded mobile market, um, was, was, was where it was born.
Um, and then, um, you know, Gerald and I, you know, had some influence over the, um, sub-arctic check the chip that ultimately got used, um, although not as much as we wanted, but when we got to launch that, that ship, um, at TI for the BeagleBone, um, and, um, um, so that, that one we actually got to do when the chip launched, um, we launched the original BeagleBone as well. Um, and, um, you know, um, Gerald always, uh, uh, uh, know my requirement was that it fit inside of an Altoids mint tin? Because that's, uh, uh, there's this kind of meme of, of, um, of survival tools, um, being put into an Altoids mint tin. Are you familiar with that?
I think so. Yeah. Yeah. Like, um, yeah, like, sure. I think, like a carving knife, and whatever else...
Yeah. Inside of an Altoids mint tin. So for, for, for me, the BeagleBone was the survival tool for the 21st century. And so that's where it got its, um, you know, a characteristic outline right there. It actually fits inside of a, of a, of a mint tin.
Um, and, um, you know, and that chip was done more specifically, you know, broader embedded solutions that are these things called peer use, where we can do real-time microcontroller tasks at the same time as, um, as the Linux solutions and stuff.
So I, maybe I made that a little bit too long, but, um, you know, there there's, uh, you know, it's, it's been about embracing the needs of the embedded Linux community all the way from, from my inspiration has been about what does it take to make people feel empowered to make it appliances rather than simply, you know, um, um, you know, be like building something for other people's platforms, right?
And like there's a lot of fun that can be had through making stuff on the web. Um, and I think that like you, like when you can inspire somebody to do stuff with, with, with web technology great. Right. Um, but I think when you can make something tangible and mobile and physical, right, there's a lot more opportunity, um, to make that, uh, inspiring and empowering and enabling.
I mean, the, there is something very, very real at the, at the earliest days of balena, we were cold Rulemotion of the day at the time, and we haven't any idea about what we were gonna do, but we were like an agency. Right. And we had different projects and some of them are pure software. And there was this one project that was digital signage. And we have like real screens on the streets, uh, that we were working on it. They weren't like that, like, you know, uh, low, they were actually on a recycling bins. So, which is a whole other story.
Um, that company, that, this story, that, that team that was working on the screens, there was so much more attached to what they were doing right. For, for, for reasons like hard to process, but it was, it was a physical thing.
And that like just the human brain, I think, works on a very different level. Like you may understand as a developer, right. That a piece of software that's in a data center is just as real as you give them a piece of electronics that you see right in front of you, but somewhere deeper in your brain, something else is really confused about that point. It's just like, if you see it in front of you, you just feel so much more real. So I, I completely subscribe to your view on this, and it’s something that we all, um, I really kind of, this is where I get so attracted to the maker community, the 3d printing community. Like we need things we can touch, like, you know, which is weird for a software person to, to say. And I'm even like my, my personal stuff goes into like modeling and abstractions and all that stuff. But like at the end of the day, you know, it's very different if you can, if he can smell it and, you know, like he can just, you know, just really interact with it on, on a, on a, on a physical level.
I think so. Um, and, and I don't know that it's always, um, and maybe that's always been true. I think there's always some element of people that can be inspired just by the imagination. Right. They can flip through a book and, and, and kind of like see things in the mind’s eye.
But I think that there are always going to be a number of people that I'm just like instilled in so they can touch it until they can feel it. And I know for me, like a computer, wasn't like a, you know, something, you know, completely empowered until I could start taking apart and building my own. Right. You know, putting processors down, putting memory down and making things, uh, twiddle. Right. Um, fortunately I already had kind of the, the bite, right. I was already, um, you know, Intuit, um, and invest in. Um, but I, I, I knew I wouldn't be complete until I could go and make something of my own and making more stuff all the time.
[JASON shows board] This one is a, this is the we're working on an AI 64 board. So this is, um, about, let me see, like, what's the 64 bit board, and it's about four times as much a neural net processing as the BeagleBone AI. And so this is, this is for, for late this year, this is a fun one.
Very cool. Thank you for that.
Yeah. The, the cool thing about this one is that we're actually going to do a robotics kit directly with it so that we can, um, like actually, you know, I've, I've, um, I don't know if you've seen this thing called Donkey Car, but I really like it.
I really like what they've done to make kind of, um, neural net stuff, uh, accessible. Right. I mean, I took neural nets, all sorts of neural net classes when I was in college. And, you know, I never thought that stuff would be like, it was like interesting theoretically, but I like, I, you know, the, the exponential growth of computing power, wasn't something that like, you, you can kind of understand it, but like, to, to like that, that, that vision of like, seeing it actually like happening, right.
To see where we can train models like we do now, it's so mind blowing for me just that we have the computing horsepower now to train these models. Um, like I, I, you know, I've always, I, for, for longest time, I've been, you know, excited about what you could possibly do with the neural net, but I never, never knew when we would get to the point where we could actually train these models effectively.
And, and now we can, you know, have models that we can train by driving around a little racetrack and make ourselves an autonomous driving car with this thing, this project called Donkey Car. And I, I love it. Right. So we're actually going to build a little Donkey Car around this, right. So that we can like make, make it where, um, you know, people can, can build a little bit of an intuition.
I think that, um, you were still a little bit, um, I, and AI is, is, is a wonderful technology. It's going to do wonderful things for humanity and civilization, but could also do horrible, horrible things for humanity and civilization because people don't understand it or they apply it incorrectly. And the way that we encode human biases into these things.
So I think it's really important that we go out and educate people on what AI or neural nets really are, um, and, and help them build an intuition for them. Um, you know, as many different people as possible, right? So that we can actually speak intelligently amongst ourselves about it because, um, there's huge risks to society as well as huge opportunity.
Yeah. Yeah. And, and, and I mean, that's kind of connects back to what we were talking about at the beginning, right. Where, um, you have all these trends that are sort of coming together in ways that, uh, you know, are very hard to maybe even sort of think through how they're all going to, uh, connect, but ultimately it feels like something, you know, very different going to be possible at the, on the other side of that.
Uh, even, I mean, what you just described, right? You've got the robotics, uh, board that you've worked on in the AI board. And if I understand correctly, like these two things are gonna start, uh, maybe ideas from both are coming together in a way, um, you...
Know, like you talk about the, the, the, the convergence divergence, right. We even amongst our own projects, right. There's convergence and divergence. So...
Yes, but these are like, they used to be two different, uh, trends, right. AI on the one side robotics on the other, all of a sudden, you know, you have them coming together with like progress in Silicon and another place, you know? Um, so it's, it's, it's fascinating to me how they coming together and I, I don't have a good capacity to sort of project how all of these things going to come together. Blockchain is another one, that's great.
You know, it's computing of course, um, that, you know, it, we all kind of feel like the, all these things are gonna change the world hopefully for the better, but if you ask like, exactly how, you're going to get far different… divergence.
Yeah. I just don't think it's a given that they changed the world for a better, I, I think that, um, you know, technologies and incredibly empowering, and I think that the way we've reduced the risk for it making the world for the worst is to put in as many different people's hands as, as possible, um, such that, um, you know, th that you can get the right inspiration and then, and, and understanding, um, so that the people that can kind of learn to deal with the technology as it exists, right, can understand that and, and help us cope as a society.
And I know that's, that's probably a little verbose, right. I think it's just as simple as, um, if, if we're going to change the world with technology, then the more people that understand it, the better. And, um, and, and, and just that, right. I think it just put it out and have it be a black box and to have, you know, us kind of being, you know, Gandalf and all the rest of the world is hobbitses, um, doesn't, doesn't, doesn't do the hobbits any good. Um, so let's, let's, um, let's, let's, let's put it in as many peoples’ hands, so they can actually understand it.
You you're, you're actually phrasing this in a way. That's kind of very interesting because essentially if I understand what you're saying is like, you know, this, this is happening, right? Like the, the, your new iPhone has the neural chip in there, all of this technology. Right. But somebody needs to take the fire, you know, from the gods to be everyday, you know, the Prometheus's sort of movement, right. Like opening it up to everybody else, because that's all and good if it's happening in, you know, in, in like a closed silo, but the potential of what it can do, and also sort of the, the counterbalance, right, like of, of what can be done is so much greater. If, you know, more people are equipped with the same, the same tools, then having them be sort of, you know, a very limited set of companies or teams or whatever, they can have access to them.
And, and, and just the fact that, you know, in everyday conversation, you can have somebody kind of, we've, we've lost authority right there in our society, right. There is no authority anymore.
There's no one that people feel that they can entrust with the, the, the, the rights to make decisions for them. And everybody thinks that they're the, the, the leading expert on all the decisions that impact their life and, and reality is they probably don't have any clue, um, of how the things that they ask for and push for it are actually really going to impact them. You know, you see a lot of people making decisions and pushing for choices, um, that are in their detriment, right. And like direct detriment, like personally harmful to their jobs, their health, you know, their, their, um, you know, their freedom.
And I don't know that I can, can, can fix that, but I, I, you know, I sure hope that by giving, um, you know, closer proximity to two people with answers, right. Maybe it's not them, they get the answer, but maybe at least it's their neighbor. Maybe it's somebody that they're talking to. Maybe there's some sort of close proximity of somebody that actually understands this stuff around them, or if they don't understand themselves that they're not, um, just easily duped by people telling them just flatly wrong information.
And that's actually another fantastic point, right. Because you can see different contexts in different cultures, um, and how different technologies apply to them. Right. And, and, um, I've been fortunate enough to live in like three different countries that are very distinct. So, uh, I grew up in Greece. I spent my kind of student years and the UK, and now I’m in Seattle. And, and, and it's very helpful to sort of think of how, you know, certain technology will get applied. And I know that that's like a tiny sampling right. Of all the cultures in the world, but the, the, the spectrum that I've personally experienced. Um, and then, you know, just multiplies...
Going from one to three perspectives is already like, so much more enlightening. Right.
But the interesting thing is that you can sort of see how, you know, a lot of times, and I think I'm forgetting it even because now I live in the U.S. so a lot of things here are U.S. centric, and we forget how it seen, uh, you know, everywhere else. And it's obvious that it's, U.S.-centric like, it's the, it doesn't lose the U.S.-centric-ness when it gets broadcasted over, like, I don't know, CNN in Greece.
It's still looks stupidly U.S.-centric, but the U.S. tries to, like, it's almost like an acts as if the world doesn't exist. But, um, the, the, you know, there's all sorts of micro context basically in cultures and people, and, and, and use cases really. I mean, just to, to bring it back to technology that all these, uh, things that we're making to fit into, but we'll never see them with a one size fits all, uh, top-down approach. Right.
And this is why I'm so excited about kind of the bottom up there, the divergence, like, I feel like that's what we're, we're, we're lacking a little bit, uh, we'll get the chance to converge later, but I think for now, like we just need to get it, you know, more and more people involved and, and trying, and making and building and experimenting and just pushing boundaries. Um, that's kind of, to me, to me just feels the most urgent, uh, of the kind of relevant things we could be doing.
I won't disagree with that. I would agree that I think we don't know ultimately whether or not it's going to help, but...
Yeah, th th th there's always the question right. Is, is, you know, something really, you know, scary, gonna get a pop out of that, like, you know, a divergence, but, um, it's, to me, I don't know, I, I'm starting to get convinced of, of, of, of the, like, being a little bit more optimistic, I guess, because I just get this feeling that we've been taught that people need to be controlled.
Like, this is kind of like the systems of control versus systems of enablement again. But like, um, I think there's been this story. We tell ourselves a little bit, maybe we're afraid of each other. Maybe I don't know what it is, but like, I feel like maybe we just need to take a step back and say, look, why don't we just enable each other and, and, and, and see where, what we can build rather than just spending, you know, being constantly sort of fearful. And it's not like when we're constantly fearful, we're getting better for it. You know, like, it's it, there's still things that go really wrong. And then you just have a centralized system that goes really wrong.
And then what do you get?
Yeah, to me, to me, you're kind of clicking on some of the fundamental cause. Like I, you know, I'm not a fundamental believer in bad guys and good guys. I just don't think they exist, but whether you believe they do exist or don't exist, you know, if you say that there's, there's, there's, there's bad guys. There's, there's bad actors whose intention to use technology is nefarious fundamentally, right. That they don't have a good purpose. Right. Their purposes is fundamental. And if I were, I guess the, the, the, the issue I have with like simply trying to like, okay, we're going to deny everyone in order to try to keep those people out of it. What, what do you end up doing is that you end up not building up this broad understanding of the risks, because it doesn't make the technology go away.
That's, that's the thing that just like, it's like mind blowing to me. It's like, it doesn't go away just because you say you want it to, it's like somebody thought of these things. I'm not, I'm not getting things into my hands that are somehow like, you know, locked up into like some secret vault, somewhere in, you know, some place, uh, not naming any of those things.
This, this is stuff that, yeah. I mean, do I help change the access model? Yeah. Somewhat, but guess what, those bad actors could have already done this right. That they could have already gotten around all this. And, and I'm not saying that I don't worry, um, personally and responsibly worry about somebody using technology that people Beagle creates right. To do something nefarious I do, because, you know, whether I believe in good people or bad people back, you know, or not, people do bad things.
And, um, and, and, and so I worry about that, but I think it's more fundamental for me to try to make sure that, that people are familiar with what is out there and what the tools are, and, and, uh, build a kind of a broad even baseline so that we learn how to deal with these things.
Um, you know, facial recognition and like objects that you can barely see. Right? Um, you know, one of our, um, I think one of, one of the, kind of the community's nice, uh, thought leaders, um, looking at, um, attack vectors, right? You know, kind of envision this, you know, a quad copter that would go in and land on, um, on different buildings and, and, and attack into their, their, their, their wifi systems. Right. And, you know, within use motion sensors, as people came up to it and they could fly away and they could talk to it from kilometers away through, through mesh networks. Right?
It's not that, um, they're, they're, you know, people don't think of these things in order to create them. They think of them because the fact is other people. Right. And if you, if you, I don't think about the fact that other people could, that were sufficiently motivated to go and create these things, then you're not going to deal with the situation, and, and, and put the security features and other things in place that you need to actually move forward. Right.
So, um, you know, I I'm, I'm, I'm not willing to, to, to sign off on, well, if I didn't do it, somebody else would know, I need to figure out the right way to do it. And it's not okay for me to say, if I didn't do it, somebody else would, um, I need to figure out the right way. But, um, um, and, and so I think we, we try to come about it thoughtfully, but I think we, we, it, it, it's not going to help just to try to lock in a closet.
And, and, and that's really a fascinating point because it speaks to something else that's very central to how we work at balena, but also like, just very sort of near and dear to my heart, which is kind of feedback loops, right?
Like, it's not about just what you give or what you see. It's also like you are there to see when maybe something goes wrong to, uh, champion fixing, you know, or, or, or plugging something. So, so it's, you know, you can think about it, like as a, as a, as a yes or no answer of like, do we enable X or Y, or, or not, or whatever. Um, but at the same time, you can think of it as a process where you say, we, you know, we will enable things and we will see how it goes. And we will try to make sure that we are, you know, uh, mindful basically of what we are seeing, process that feedback, take that into account, uh, and keep going.
And, and that's all that, that's the best we can, we can do, because as you say, he can't put the genie back in the bottle, um, and you, and you, can't sort of hope that people don't notice the genie exists. So the best you can do is channel that to a good direction than that in a bad direction.
Um, honestly the democratization is, uh, is, uh, it's the concept that talks about is great. I don't think it like democracy is the right, is the right word for it. But like the concept of like broadening access to technology, um, is sometimes the paradigms that we think are right turn out to be wrong. And when that happens, you want the people who are right, who might be your, you know, your opposition to be ready to sort of, uh, shift gears rather than, you know, just kind of double down, like, you know, Apple doesn't do it. Nobody will. And that's, that's that sort of thing that can go up to a point and then you're stuck, right? Like we, we, this has happened how many times-- Microsoft was, could do no wrong and then Google could do no wrong. And then IBM, once upon a time could do no wrong...
I think we've not managed to say an important word, which is accountability. And, and, you know, like the, you know, you're talking about different governing models, like what democratization, like there's democratization, like in terms of hardware, access and stuff, but it's still kind of a, uh, a similar thought, right. You're democratizing power in some way. Right.
And, um, like even classical Greek democracy, right. The citizens were representatives, right? I mean, they were the few, right. Um, or not everybody was, uh, uh, you know, the classical Greek, you know, democracy, not everybody was a citizen, but the citizens were there, they're responsible. Right. And they were accountable and in some way, I mean there are failings always, but, but, but I think that, um, and I I'm like, fundamentally when we come to technology and, and, and large corporations, right. I think we have to think about how these systems address accountability for when things do go wrong. Um, and, you know, uh, like there, there's a, there's a right degree of accountability. Um, and, and I think we should strive to find it. I think we've swung a little bit far to the point of, of almost no accountability.
Yeah. I think the big thing I'm noticing, and it goes back to sort of organizational design, is that, um, to meet you're building an organization, power and responsibility should never get, uh, on, unhooked from each other. Right. Uh, decoupled. Um, but that happens all the time and everywhere. Right. So, to me, like power should be totally like fine to seek and to acquire and to wield so long as the responsibility of that power is, is well, uh, you know, uh, connected because the, the real problems start when you have decoupled those things.
And there's certain people that are always accountable for everything, but there's other people that are not the ones who make the decisions. And then, um, you know, it's very, very, it should never be to have, you know, power without responsibility. But like, to me, that's kind of the problem we need to resolve otherwise we will be stuck.
Well, yeah. I mean, it's fascinating how you've, you've sort of taken the conversation because it's, it's really the theme that we were hoping would emerge from this podcast that, uh, we are acquiring the power of the gods and the one side. Right. All of the new capabilities and giving them to everybody, but we need also the, the love and wisdom of the gods, uh, you know, in that kind of poetic frame to overlay on that technology so that we don't actually end up blowing ourselves up. Right.
Like, so how do we bring these two worlds together? Because right now we're doing pretty good on that and capability, but, um, in terms of adding, uh, you know, channeling that let's say, uh, we are, you know, less good at and getting worse, maybe? I don't know. But like, yeah, we like, that feels like a, you know, like the bodybuilders, like I've worked on the upper body and the legs are like skinny?
Yeah. And, you know, we, you know, with, with working with Linux, right. We're building on the shoulders of giants and, um, you know, it's important to kind of go and look at what it took to kind of build all that stuff up in the first place and make sure that we can kind of rebuild all that again, if we, if we needed to.
And it, when it comes to like overall, you know, tech technology and society, right. I think that, you know, depending too heavily on anybody, who's just kind of built a, uh, a monolithic solution that you can figure out how to rebuild it on your own. You're probably putting your, your, your eggs in the wrong basket. Right. Um, it's not that you don't value all the effort that they've, that they've put forward, but, um, if you're just simply banking on them doing it all right. And, you know, like, oh, I trust them. Right. You know, you know, how, how accountable are they really, you know, um, when something goes wrong, you know what, what's your, what's your recourse, that's your way of dealing with that. And, um, and those are, those can be really hard questions in business. Um, and, you know, you know, and you don't, you don't need an answer until you do, right. You know, until something goes wrong and, um It's worth figuring out early.
Okay. I think this is a fantastic sentiment too, to leave everybody weighed in their heads with, so, uh, thanks, uh, Jason for, uh, for joining us and being so generous with your time. Uh, I, I really like how this conversation sort of evolved, so I'm sure we could have talked about all sorts of other things, but this is just, thank you. Thank you so much for all the time and, uh, thoughts.
Uh, I'm, I'm, I'm happy to share it. So, you know, if anybody's looking for me, they can go to, um, to beagleboard.org/about, and they can find contact info for me. And, um, you know, I try to make myself available. Right. That's, that's part of my side of being accountable to my community is to make myself available. So...
So, so are you, are you, do you spend any time on any social networks that people can find you or is, is this the, be the BeagleBone is the entryway to...
So IRC, I try to spend some time on, um, you know, Twitter to far less extent, Facebook, to the, uh, least extent. Um, you can, you can kind of find me on those, but each of those links will be on the, um, the about page.
Perfect. Thank you so much.
--- END EPISODE ---
by Andrew NhemAndrew is the Content Strategy Lead / Product Builder at balena, and enjoys tinkering on web content, building stuff, music, hydroponics, and homeschooling