#105 – An Interview with Chris Anderson - Deambulatory Daedal Drones

Chris Gammell: This is the Amp Hour Podcast, recorded July 23rd, 2012. Episode 105, with guest Chris Anderson, Deambulatory Dadal Drones.

Dave Jones: Welcome to the Amp Hour. I'm Dave Jones from the EEV blog.

Chris Gammell: And I'm Chris Gammell of Chris Gammell's Analog Life.

Chris Anderson: I'm Chris Anderson from DIY Drones, Wired, Longtail, and a bunch of other things. So many things.

Dave Jones: A bunch of other things. Thanks for joining us, Chris. This is going to be confusing. We've got two Chrises, so we'll do our best here.

Chris Anderson: If that's the most confusing thing we've got tonight, we'll do all right. Yeah, I think so.

Dave Jones: All right. Night? No, it's daytime here. It's midday here in Sydney, Australia. I'm sorry. Yeah. Anyway, thanks for joining us, Chris. This is going to be awesome. So you are editor of Wired Magazine. You are co-founder of Do-It-Yourself Drones.

Chris Anderson: Founder of Do-It-Yourself Drones, co-founder of 3D Robotics, which is the electronics manufacturing arm, the commercial arm of DIY Drones. We make autopilots and aerial robotics hardware. I've got a couple books. Longtail and Free are the two, but there's one coming out in October called Makers, which is all about the stuff we'll probably be talking about tonight. Absolutely. I founded Geek Dad. Oh, yeah. That's right. Yeah, which is all about raising children who are excited about science and technology. So I've got five kids. I need to say I've completely failed at raising any children who are excited about science and technology.

Dave Jones: Oh, no. Yeah. So is this like a blog? What is this? I haven't heard of this.

Chris Anderson: Geek Dad? Yeah. Yeah, Geek Dad started as a group blog. It's now one of the official Wired sites. There's a geek mom as well. I hasten to add. It's now a series of best-selling books. There's some products in the offing. It's a little mini parenting empire.

Dave Jones: All right. Oh, I'm going to have to get into this because I've got a 14-month-old now.

Chris Anderson: Yeah. Well, the whole point was that there's lots of things that are fun for kids and boring for grown-ups or fun for grown-ups and boring for kids. But to find projects that are really kind of cool and geeky that are fun for both was a challenge. So the site is designed to share best ideas on that.

Dave Jones: What seemed to be the most popular things that actually hit it off with parents and kids in terms of projects?

Chris Anderson: Well, you can't go wrong with Comic-Con kind of stuff. You know, Star Wars and what order should you watch the films and how do you bring the kids out? Oh, no. You know, Lego is always a good start. Yeah. You know, sort of basic electronics, all the science experiments, especially backyard science, rocketry, kites, things like that.

Chris Gammell: Yeah. It seems to dovetail nicely with a lot of the make stuff as well. I mean, I don't think it's not associated, right? But it seems like it's very well aligned with a lot of the same kind of projects and stuff.

Chris Anderson: We're all good friends and we all kind of mix in the same circles. That's awesome.

Dave Jones: We've had quite a few episodes where we've talked about women in engineering and girls in engineering and getting more involved. How have you found the – like you said, you've got a geek mum, right? Is that an important thing to get girls involved? It is. Is it harder? Well, no.

Chris Anderson: I mean, I want to be super clear, but one thing, geek dad doesn't mean geek father son. It just means geek child. Of course. So, you know, we're all about our daughters. As a matter of fact, my daughters are actually more into science and technology than my sons, largely because we have a 3D printer. And they love using a 3D printer to make dollhouse furniture. Awesome.

Dave Jones: Oh, there you go.

Chris Anderson: Yeah. Interesting. Yeah. So, the geek mom thing is just for geeky mothers who also want to infuse their children with science and technology. Right. And sometimes a mother's approach is different from a father's.

Dave Jones: Got it. Speaking of 3D printers, I've just got the new Make-A-Bot Replicator and it just broke on me, not through any fault of my own. I'm just curious. Where do you see the 3D printing industry? I know it's in its infancy. Where do you see it heading? Do you have any input on that?

Chris Anderson: Yeah. It's going to be absolutely huge. It's huge. So, I've had like every 3D printer ever. I mean, I'm still using a thing-o-matic, which is the version ahead of yours. And I've broken my share of 3D printers as well. They're still a little fiddly. I think we're kind of in the dot matrix phase of the whole thing. Right.

Dave Jones: That's a good – that's an apt description, I think.

Chris Anderson: And we need to get to the inkjet phase. And that means plug and play and reliable, but also multicolor and fast and simple and better desktop software. I think there's two things that are going to happen. And the first is by the end of the year, you're going to see a crop of really kind of commercial quality ones that aren't sort of laser-cut plywood and fiddly Python scripts in the setup process and all that. So, you're really going to see it starting to feel like something you might actually just buy in a box and use. And that, I think, is going to make the kind of user experience easier. And then from a technology perspective, right now we're dealing with extrusion, ABS plastic that's melted and extruded. The next level of sophistication is going to be lasers tracing in a resin, a liquid resin. The thing about ABS extrusion is that it limits the speed at which you can produce. I mean, the first thing people realize when they get their 3D printer tuned in is that it's slow. It's crazy slow. And there's really no good way to speed it up because it's based on the thermal properties of the plastic. You sort of need the plastic to harden just enough in each layer so the next layer comes on cleanly without smushing. So, there's a physical limit there. There's a physical limit there. Whereas with the laser curing of the resin, you don't have the same physical – there are some physical limits, but they're nowhere nearly as constraining.

Chris Gammell: Did you see the Kickstarter project for that?

Chris Anderson: Yeah, I did see the Kickstarter project for that. I have not tried that out myself. And to be honest, after having been kind of alpha and beta testers on all the 3D printers to date, I think I'm going to kind of let this phase – let me let the first phase pass and be a late adopter on that one. We don't have – Where's your faith? Come on. I know. No, I'm alpha and beta testing too many things as it is. So, this is one where I'm happy to let others take the heat. But also, I'm really happy with my MakerBot. And we don't have serious production requirements. So, if it takes two hours to do a big print, so be it.

Dave Jones: Of course. Where do you see that laser technology? Is it actually mature enough at the moment to get a usable product or sort of a dot matrix level usable product out the door?

Chris Anderson: Oh, yeah. Absolutely.

Dave Jones: We've got some ways to go in terms of material science and things like that.

Chris Anderson: The professional quality ones are fantastic. Oh, of course. And you can buy one for $50,000 or $60,000. That's amazing. So, the technology has actually been around since the 70s. Right. But bringing it down to the sub-$1,000 level is still a challenge. Among other things, the resin, the consumables are quite expensive. So, there's a few things that need to be done. But I think we're talking a couple years at most. No more than that.

Chris Gammell: And that's definitely the path, you think, is laser next instead of the powder plastic or anything else like that. I mean, I see a lot of those.

Chris Anderson: Well, powder plastic typically uses laser as well. It's just a different substrate. Oh, that's true. Like sintering. Exactly. I'm not expert enough to know which of it it's going to be. My sense is that the consumables are a little less messy with the liquid resin than they are with the powder.

Chris Gammell: Well, the powder – one of the powder ones is actually – it actually is like an inkjet where it actually deposits the organic – it's like an evaporative basically. So, you kind of like put liquid on top of the powder and it creates this like sludge and then you kind of bake it and that's the –

Chris Anderson: There's tons of competing technologies. There's also things that actually cut out really thin films of plastic. Oh, yeah. And so, there's a bunch of them. I'm – right now, I think ABS extrusion is going to mature to the next level and that's usually about usability and sort of reliability and just ease of setup. And I expect that to sort of carry us through for the next year or two. And at that point, I'd expect the laser ones to be at a consumer price point.

Chris Gammell: I have a question, just a general question though. Where do you find time for all this stuff?

Chris Anderson: I get this question all the time, especially from my wife. Do you sleep? Is it a sleep thing or – I sleep a normal amount. I'm very good at starting stuff and then I tend to kind of delegate. I mean I run actually more than the magazine. I run the overall brand which includes the website and the iPad app and the events and all that. So, that's a pretty big operation and that's obviously my day job and that takes me all day. Yeah. Yeah. Then the biggest thing that takes my time – Geek Dad has been handed off to another team. So, I'm an emeritus to use the phrase there. My robotic stuff, the drones and the UAVs is a huge time suck and I mean that's been just sort of consuming every other spare minute. But I'm getting – I've now sort of delegated most of the stuff including the dev teams. The software dev teams are now in the hands of professionals and I by and large reduce this to a couple hours a day. The books – when I'm working on a book, which I have been for the last couple of years, those suck up my weekend since I spend way too much time in Starbucks with headphones on. And then there's the small matter of the five kids and the speaking gigs and all that and I think I'm just sort of – I'm happy when I've got too many things going on or maybe I'm just happy when I'm stressed and overwhelmed. And it just seems like we live in exciting times and I don't want to let anything – any opportunity drop.

Chris Gammell: Yeah. You're in the right crowd right now. Me and Dave were talking about that before the show actually. Both of us are loaded down.

Dave Jones: Too many irons in the fire, you know. And one of my questions is – we talked about this the other week. The pit of despair is what Nathan Seidel, who's the Spark Fund, founder of Spark Fund. He calls it the pit – okay. He calls it the pit of despair. I'm not sure if you've heard him say that. Yeah, I suppose. Oh, okay. Yeah. Where something like Do It Yourself Jones, for example, you got caught where it – well, this wasn't supposed to be successful. Now it's sucking all of my time. And there's that intermediate zone where it's just successful enough where it's sucking all your time and you can't hand it off. And when it was just taking a little amount of your time or it goes to the other end like, say, the Wired stuff, Geek Dad, where you can hand it off to somebody else and just sort of oversee it from a top.

Chris Anderson: So I think I have three – How do you deal with that? I have three pits of despair. The first one – I mean there's probably more than three. I've only hit three so far.

Chris Gammell: Levels of hell at this point, right? Yeah.

Chris Anderson: The first one is definitely the moment where a hobby becomes not fun anymore. Oh, yeah. So you got into it because it was fun and then you kind of started to industrialize it and take it up. I mean let me give you an example. Let's say that you write some code and you're pretty happy with it and you post it and you open source it. And then people start using it. And that's awesome. You feel great about that. And then they like want technical support. And there's like bugs. They want better documentation. And suddenly this little sort of gift to the world becomes this massive responsibility, this kind of albatross around your neck. And what started as a kind of like, woo-hoo, I'm coding, turns into, oh, my God, I'm doing tech support. Yeah. So that's the first one that's sort of not fun anymore.

Chris Gammell: That reminds me of – we talked to Bunny Huang. He said he had people calling him at 2 in the morning for like sales support. Yeah. And he was like, this is my personal line but I'll be happy to figure this out in the morning. Like that kind of thing, right?

Chris Anderson: Yeah. So we've got all that. But I mean once you really scale up, it's just crazy. I mean like you wake up in the morning and there's like literally 200 emails. Oh, my God. And they're all unhappy, right? Yeah. These are people who are getting stuff for free and they're all incredibly like entitled about it. It's like, you know, I can't believe your documentation sucks so bad. And I was like, I just gave – it's just got it for free. I mean anyway, so there's that. The second pit of despair is the – when things are working really well and then you scale it up and you realize that the edge cases just kill you, right? So it worked on like one configuration. Now people are bringing in other configurations you've never seen before and your code didn't anticipate. Or maybe it turns out your 1% failure rate on your boards multiplied times 10,000 is like a really big number. And well, it's not a really big number. It's a very predictable number. But nevertheless, it's a daunting number. Yeah. And then suddenly you realize that all you hear is complaints. And that's the kind of the – oh, my God. Maybe I don't have a robust platform here. Maybe there's something fundamentally non-scalable about the project or the code or my algorithms or something like that.

Chris Gammell: It just sounds disheartening too. It just sounds –

Chris Anderson: Well, that sucks. I mean that's really when you – when something that looked easy turns out to be harder than you thought or the problems are more subtle or the algorithm that worked in a kind of a narrow case doesn't work in a broader case. And that's the sort of like, oh, my God. Is this actually going to work? I mean is this actually – did I choose the right approach? Is this really going to get me where I need to be? And that's – and there's a crisis there where you really have to kind of ask – you and your team ask some tough questions like have we made the right choices? Are we on the right path? Is this – are we going to – can we fix this? And there's these moments where everything you do makes it worse and your fixes create two bugs and it's really dispiriting. And that's a moment where it's all about team building and leadership and communications. And that's where a robust open source system kind of pulls together and just powers through it because that's horrible. And if you've got something big and gnarly, if you take on a big gnarly problem, you're inevitably going to hit that moment where it's gnarlier than you thought. And then the third pit of despair is once you've got it working and it's successful and you're building a business, then you deal with popularity. It's like you can't meet demand. You've got supply chain problems. You've got inventory problems. And it's like – and now everyone thinks you're incompetent because you can't run a proper business.

Chris Gammell: Right, because a 12-week lead time because of a Japanese tsunami or something.

Chris Anderson: Exactly, exactly. Or the Chinese went on holiday and who knew? Oh, my God.

Chris Gammell: In February. Oh, God. Chinese New Year. Run for your life.

Chris Anderson: Exactly. So the sort of pitfalls of success and volume are – and then you spend all your life dealing with supply chain issues and QA and yield and multiply and 30 – trying to negotiate 60-day net and cash flow. Anyway, we've at this point discouraged everybody from that. No, no.

Chris Gammell: We talk about this stuff all the time on here, honestly. Supply chain is a shockingly huge part of electronics and I think a lot of our listening audience understands that. So we feel your pain. We really do. Yeah. So those are the three I've run into so far.

Dave Jones: Yeah. Oh, there could be more. And what's all this talk about team? Team. Geez, I'm a one-man band. You know? David and Wolf. The luxury of having a team. No. I can't believe it.

Chris Anderson: I'm a huge believer of teams. I mean, basically, I'm a terrible completer. I'm a good starter. Yeah. And fundamentally, I think open source is all about the social factors. You know, pulling together a team, creating what Tim O'Reilly calls the architecture of participation. You know, just finding – I mean, you know, it's really interesting. We have about 100 developers now in DIY drones and working on – you know, RG Pilot is our hardware and RG Plane and RG Copter and RG Rover are the software sets. And it's interesting. You know, we started off real loosey-goosey saying, hey, everyone – you know, everyone play. It's emergent properties and all this kind of stuff. And now we have, you know, relatively, you know, traditional-looking, you know, command and control architecture. We have, you know, roles and responsibilities and team leads. Wow. And, you know, people get, you know, get kind of interviewed before they're added to the team. And, of course, we have, you know, proper version control and, you know, code reviews and all this kind of stuff. And it kind of looks like GE. Yeah. It's like a – it's organized around normal kind of, you know, almost corporate structures, except for people are motivated by basically non-monetary, you know, reasons. They're volunteers, which means you have to – the funny thing about volunteers is that you have to be even more careful and attentive to managing them because, you know, if they're not happy, they'll leave.

Chris Gammell: It's like having that – what Dave affectionately calls FU money but without the money aspect. Exactly.

Chris Anderson: In a funny way, you can mismanage employees but you can't mismanage volunteers. Yeah. Wow. Wow. That's impressive.

Dave Jones: And you found that was essential for – to run it like that to get the, you know, RG pilot where it is at the moment? Would you say it was essential or was it just a –

Chris Anderson: It's a really complex project. This is the most – it's based on the Arduino platform and it's the most complex software that's ever been written for Arduino. You know, we found, you know, bugs in the compiler just due to the file size that no one has encountered before. So, you know, at this level of complexity, I think we really do need, you know, pretty sophisticated management structure. I'm not to say we're perfect by any means. We've got a long way to go. But people like – people you may have heard of like Andrew Trigel from the Samba Project have been really instructive in kind of coaching us on best practice and open source software, which we – and our software – the truth is our hardware team is really much tighter and rather smaller. And many of them are employees. So, that's easier to manage. And the software just kind of builds on traditional open source software projects. So, we're structured around things like Ubuntu or – so, we had Jono Bacon from the Ubuntu Project who wrote a fantastic book called The Art of Community. He coached us on some of our community management practices. And again, Trig, Andrew Trigel coached us on some of the experience from the Samba Project. So, we feel we've got kind of, you know, great open source, you know, experts who are kind of coaching us on being better managers and coordinators.

Chris Gammell: Could you tell us a little bit more about the hardware versus software? I mean, how much of it is hardware these days and how much of it is software?

Chris Anderson: Yeah. You know, it's really interesting. So, we're part of the open source hardware movement, which is to say that we release the design files, the CatSoft Eagle files for all of our boards. And, you know, it's – hardware is, as, you know, a number of people pointed out before, has about seven layers. Unlike software, which is really a single layer, it's just ASCII text, which we, you know, can share on a version control system. Hardware starts with, you know, schematics and board designs and bill of materials and instructions and documentation and, you know, sometimes CAD, you know, images as well. And each one of these layers requires a different form of communications and collaboration. So, like version control systems, it's very hard for them to show you a diff on a pretty circuit board file. It's very easy to see that in code, very hard to see on a board design. And, you know, there's questions like what are the right licenses for each one of these? What's the sort of preferred file format? So, again, with software, it's super easy. It's ASCII. Everybody can read it. You know, what's the preferred file format for pretty circuit boards? We use something called Katzoff's Eagle, which is okay. We use it because it's largely free for non-commercial use. But, you know, some people don't like giving away their Eagle files because it's so easy to clone. And, you know, when we release a product, it is now – it now takes less than seven days for the Chinese to clone us and basically put a clone of our product into the marketplace.

Dave Jones: And they have been actively doing that, haven't they? I know they started like a year or two back and they –

Chris Anderson: Yeah, yeah. So, places like Good Luck Buys basically have our stuff within seven days of us releasing it. Wow. So, part of me says, that sucks. You know, I just put – we just put like $10,000 and six months worth of work into designing that. And they just took our product and they just took our design, which we gave to the world and they're selling against us, you know, at a lower price, you know, within – you know, at the same time. Yep. And so, that feels bad. What feels good is what's supposed to happen, which is when people take our designs and they approve upon them or they do derivative designs that add features and functions that we hadn't thought of. So, our partners – and we do have lots of partners out there – you know, do that and they'll do two things. They'll either say, look, actually, here's a version of your board that we think actually works better. It's a more graceful design or whatever. And we're like, oh, that's a great idea. Well, thank you because it's an open source design. They put their design back in the public domain and then we can use theirs. So, for someone, for example, approved on one of our GPS modules by adding a battery. And we're like, that's a great idea. We should have done that. Thanks. We'll do that. And then sometimes they can do – and sometimes they'll do derivative designs for specialty applications that we hadn't thought of as well. So, basically, we just sort of accept the fact that the cloners are going to clone because we believe that at the end of the day, putting things out there as open source are going to benefit us more than it hurts us. And we have a community as well. You can buy from – good luck buy and good luck to you. Maybe it will work. Maybe it won't. But don't come to our community complaining if it doesn't work because you didn't support them.

Chris Gammell: Right.

Chris Anderson: And so, we believe that at the end of the day, community wins and we want to do the right thing for our community.

Chris Gammell: What kind of price differences are you seeing? I mean is it shockingly different?

Chris Anderson: It's not that much. I mean again, we price as low as we can. So, we have a really transparent pricing model. It's 2.6 times the bill of materials. And that number is based on two 40% margins, one for us, the wholesaler, and one for another 40% margin for the retailers. And what's nice about that is that we don't charge for intellectual property. This is just the bill of materials. It's just the parts and manufacturing costs. So, when you charge 2.6 times the BOM, it's really – you can't really underprice us by much and still make money. So, the Chinese will underprice us by maybe like one of those 40% margins. But – so, it will be sort of 40% less than us. But there won't be much more than that. And at that point, we feel like if a 40% price difference makes that much difference to you – it's not actually – if a 40% discount means that much to you and it's often less than 40%, by the way. Then – and you're willing to take the chance. Then so be it. You're probably not our perfect customer anyway. Right.

Chris Gammell: And they're probably taking on the chin too seeing the less margin. I mean maybe not as much but –

Chris Anderson: Also, they don't have transparency into the development process, so we release new boards all the time. Right. And if they just made 10,000 of the board that we've just obsoleted with our new version, whoops. Had you been part of the dev team and played nice, you would have known that.

Chris Gammell: Right. So, is that a built-in software kind of feature – or not a software feature but almost a security feature of software being up to date with hardware and just kind of being current with the community? I mean –

Chris Anderson: Again, the software is open source. So, we don't – we support actually old hardware as long as we can, so at least a year and often two years. So, it's not like our software will like stop working on our old hardware. But the community will always want the latest, greatest stuff. By the way, I want to quickly add something. When I say the Chinese, I'm not referring to all the Chinese. Some of our best contributors are Chinese as well. I want to tell you a little story about – there was one of the guys who had started a company who was selling derivatives of our board. It was actually being – he wasn't just selling via the Chinese eBay versions. He was selling on his own site. There was a sense they were doing some proper support and they were part of our – they were – I noticed they were translating our manual into Chinese. And I thought that was really cool. Yeah. And I got in touch with them and I say, hey, I'm delighted you're translating our manual into Chinese. If you're going to do it, why don't you do it on our wiki so that there's – we use the Google Code wiki and there's a way that you can actually switch between languages. And I said, it would be awesome to do it on our wiki so that you can share our images. And so, when we update them, yours will be automatically updated. And he's the guy's name. He went by the screen name Hazy, H-A-Z-Y. And he said, cool. Okay. So, I gave him a login to the wiki. In those days, we were using a version control system on Google Code and I'd set it up wrong so that if I gave him a login to the wiki, you also had login to the code repository. And so, first, he ported the manual to Chinese. And by the way, he still – he updates like almost every day. He's amazing. So, he translated the manual to Chinese and then he started translating some of the software interface into Chinese. And I was like, okay, cool. Then he started fixing bugs. And today, he's like one of our best contributors. And because we didn't treat him like a pirate, we treated him like an open source community participant and we sort of invited him in. He's become a great contributor. And there's lots of teams in China who are like that. So, I know people kind of tend to use a broad brush and refer to China. But there's – just like in every other country, there are people who are doing great innovative work and there are people who are just ripping people off. And I just want people to know that there's a lot of the former as well in China.

Chris Gammell: Right. And there's no reason there wouldn't be a cloner in – straight in California right next to you guys as well, right? Yeah, exactly. Absolutely. That's a good distinction to make.

Dave Jones: If they could get away with it, if they could make a buck from it, they would.

Chris Gammell: Yeah, exactly.

Dave Jones: Yeah. I like to think, though, that there's some unwritten rules in the open source hardware community. And one of them is that you do not rip somebody off, just blanket, without adding any value in some way back to the project, back to the community, even though you're legally allowed to do it. Because it's open source hardware, they're legally allowed to take your Gerbers and print them, as long as they don't use your trademark name, of course.

Chris Anderson: Oh, they use a trademark name with no problem.

Dave Jones: Well, that's legally wrong. But if they physically remove that, then they're legally allowed to do it. But that doesn't mean you should do it. Exactly. I think that's overstepped the mark of what the community expects. And you have to give back in somebody. You have to make a derivative, make it useful, not just plain rip off, even though you're legally allowed.

Chris Anderson: Completely agree. It is, however, an unwritten rule. Of course. It is the culture and the philosophy and the spirit of the thing, but it's not legally enforceable.

Chris Gammell: That's right. Chris, could you tell us a little bit more about your actual manufacturing operations, since we kind of started to touch on that? I mean, I've seen a couple of things about your PCB assembly hardware and everything there. Could you just tell us about an overview of what you guys do there?

Chris Anderson: Sure. I actually don't run it day in, day out, and I'm not actually there. We have two factories now, one in San Diego and one in Tijuana. I live in San Francisco, or in Berkeley, actually, outside of San Francisco. But we started in San Diego, and that is basically a standard pick-and-place line with the ManCorp pick-and-place, reflow oven, solder paste printer, and all that. So it's a relatively state-of-the-art assembly line. The one in Tijuana is a clone, actually, of that line with the addition of specialty devices for through-hole soldering. Interestingly enough, we would have thought we'd – because it's in Mexico where labor is lower, we would have used people for that. And we did start that way. But we find that the reliability of mechanized through-hole soldering is such that automation really pays off. So there's basically two pick-and-place lines, and we also have two CNC lines that do quadcopter and hexacopter frames. And those run 24-7 as well. That's pretty much it. We have the R&D and the prototyping side is all in San Diego. So we have the usual range of 3D printers and such to prototype hardware for our pipeline of new products.

Dave Jones: At what point did you go, right, we're going to actually run our own pick-and-place machine as opposed to – and our own CNC machines for mechanical, as opposed to just subcontracting out that stuff to people who already know how to do it and do it properly?

Chris Anderson: Yeah. Well, a lot of this – the credit goes to my co-founder, Jordi Munoz. But basically, it started like this. We started a community, and it was like – we used like, hey, we're going to do autopilots, and we put up some designs out there. And it's like, hey, here's some links to batch PCBs so you can have these PCBs made for you. And then here's the link to the DigiKey products so you can just order the bits, and here's some instructions on how to solder them together. And that's like almost nobody who can handle that. And they were like, could you possibly make them for us? That sounds really, really hard. And we're like, okay. And so we bought – we batched up a lot of boards, and I bought stacks of components. And Jordi and I said, well, we'll just make them ourselves and laid them out on the kitchen table and then basically went blind. That was horrible, hand-soldering that stuff. So that took us a couple hundred boards, and we're like, we're not doing that anymore. Then we sent them out to a component manufacturer, to a production house.

Dave Jones: A board assembly house.

Chris Anderson: And so we used advanced circuits, and they were great. But the problem is that to get any kind of economies of scale, we'd have to order big volumes. So we'd order like 1,000, which was a big volume for us back in those days. We'd order 1,000. And then you sit there with your heart in your throat hoping that nothing is wrong with them. Or you haven't made a mistake. Or they're broken, yeah. Yeah. And then it comes in. Good news. They work. Fantastic. And then you'd realize that you can't change the design until you've sold all 1,000 of them. So you have incredible inflexibility. And at that point, we realized that the only way we were going to get both – we were going to be able to innovate and run kind of an efficient economic operation was to have our own pick-and-place line. Because that allows us to make small batches. So we basically do a kind of on-demand production. So we'll make like 100 in a day. But we don't end up with a huge inventory of products of anything because that would discourage us from innovating and improving on the product. And as a result – so we're constantly changing the product by bringing in new components or things we were able to source more cheaply or fixing small bugs, et cetera. So we'll have – in the space of a year, we'll have maybe 30 different versions of a board. No one will notice this, but each one is designed to be easier to make and more reliable or cheaper to make. And all that's enabled by the fact that we have the flexibility of our own production operation.

Dave Jones: But how do you trade off – like you said, you might have 30 different versions of the board. How do you trade off the keeping track of that documentation and having people go, oh, my one's slightly different. What the hell is going on here in all the support?

Chris Anderson: We're not perfect at that.

Dave Jones: Where it's better to go just stick with the one version for a long time and try to milk as much out of it as possible.

Chris Anderson: You know, most of these versions no one would notice. I mean it's a component here or there and it doesn't require documentation. When we do change something that people would notice, we do update the documentation. We try not to have too many versions that do that. And sometimes they're just more – just the yield is higher, which is to say it's only a difference that we would notice in the production, which is that more than past QA. Oh, okay.

Chris Gammell: So like solder paste differences and drop-ins and stuff like that?

Chris Anderson: Exactly. So we're constantly kind of doing production and process innovation.

Chris Gammell: Yeah.

Chris Anderson: But we're still at the stage of the kind of the high growth or fast innovation period of our industry. So we largely expect to rev every board every six months and just to keep up with the industry. And some people – it's definitely too fast for some people. They're like, would you please stop releasing new versions because A, it's making me feel bad about the product I just bought and B, everyone requires a new learning process or new configuration, et cetera. And we do hear people on that. By the same token, if we did stop innovating, then people would pass us and that's not what we want either.

Dave Jones: That happened to me. I bought an Arducopter. I'm developing a quadcopter that actually goes through a canyon and actually, you know, it'll autonomous, do the whole thing, you know, and completely autonomously. And I just bought the Ardu Pilot and the Ardu Mega or whatever it was. And then like the next month, here's this brand spanking new 2.0 version or something. And I went, oh.

Chris Anderson: Well, the good news – I'm sorry about that. The good news, however, is that the 1.01 will fly just as well and is supported entirely by the code. And, you know, so I know it feels, you know, kind of off. At first, yeah. But you should know that it won't fly any differently.

Dave Jones: Once I was over that bummer period of, oh, man, I just bought the old model, then I realized, well, it doesn't actually make a difference.

Chris Anderson: Yeah.

Dave Jones: I was bummed for a couple of days. Yeah.

Chris Anderson: And that is actually intrinsic in our business, which is to say that we're kind of at the cutting edge or, you know, bleeding edge, if you will, of robotics. And, you know, if what you want is stability and, you know, and such, then, you know, then you should, you know, buy a pair of AR – not you. One should buy a pair of AR drone and have fun for now and wait until things stabilize. But there are people who really do see this as a kind of a, you know, a frontier and an experimental platform. And so they're very demanding about wanting, you know, the latest sensors and computing platforms, et cetera. So we're going to be releasing – when is this podcast coming out?

Dave Jones: Oh, it should come out like today.

Chris Anderson: Oh, okay. Well, then I can't talk about what we're going to be releasing because it will be coming out mid next week. But we're going to be releasing a series of new boards.

Dave Jones: Hold off on all your sales there, folks, because they're about to release something new. It's the – what is it? They call that the –

Chris Gammell: The release party?

Dave Jones: The Osborne effect. The Osborne effect, yeah. Yes, because Osborne went under because they announced a new product and didn't deliver and nobody went and nobody bought their old one.

Chris Anderson: Well, I can't say that we're going to be releasing the APM 2.5 board next week. But that's actually, again, functionally identical. It's just designed to be easier and more reliable – easier to make and more reliable. So that shouldn't stop people. But there will be a brand spanking new board, which is actually designed for a different market segment for really kind of advanced, you know, academic use. I don't know whether you saw that University of PEM swarm lab, the swarming quads. So if you ever want to do that yourself, now you can. Oh, sweet. So you should see an announcement this week on that.

Chris Gammell: I'm hearing Zigbee this week. I think that's my prediction. Zigbee talking amongst them and stuff, right?

Chris Anderson: Yeah. So you probably know that we have released a competitor to Zigbee with the 3DR radios, which are – I didn't know that. Competitor to the XB radio. Yeah. Yeah. So the 3DR radios use the Hope RF chip. So we love XBs, but they're closed source and that cannot stand. Nothing will get crushed in the way. And they're also a bit of a hassle to set up and they brick. And we just said, you know, enough of that. So we've released our own radios called 3DR radios in 90433 megahertz that do all the same thing, but they're half the price and longer range. And they never brick. And they're 100% open source.

Chris Gammell: What is the chip set on there? Is it the same base chip set?

Chris Anderson: It's the Hope RF. It's the Hope RF chip set.

Chris Gammell: Hope RF. I've never heard of that. I'll have to look that up. Okay. That's cool. So that's actually the manufacturer, though, is Hope RF?

Chris Anderson: Yeah. It's basically the Hope RF chip set. And then we just built the kind of logic circuitry and the firmware. And more importantly, all the configurations and utilities and such around it. So they're optimized for long range use right now. So if you want like a mile, that's the way to go. Very nice.

Dave Jones: Is that legal? How are the legal requirements handled? Or have Hope already done all the legwork there?

Chris Anderson: You know, I don't know, frankly. And it depends on the country. You know, we sell to developers by and large. So we don't deal with FCC clearance ourselves. Oh, yeah.

Dave Jones: So you just leave it up to the user. If they want to use it commercially, that's their business.

Chris Gammell: Chris, I had a question about the manufacturing side. Well, when did this all went down? Mid-2000s, right? That's when this all started? Was it later than that?

Chris Anderson: The manufacturing side?

Chris Gammell: Yeah.

Chris Anderson: Oh, no. 3D Robotics is only three years old. Oh, three years old. Okay. Yeah. I mean, DIY Drones started in 2007. Okay. DIY Drones, just to be clear, is the community. It's just a free community. And that's where, you know, people post blogs and talk about this stuff. And that's where everyone hangs out. But 3D Robotics is the company.

Chris Gammell: Okay. So I was confounding the two. So I don't think Kickstarter was in the mix at that point. But if you had to do it all over again today, would you have done a Kickstarter to actually get this up and running? Or what's your take on that whole thing?

Chris Anderson: You know, that is a great question. I love Kickstarter. And Kickstarter does three things really well. Two of which we probably would have used anyway. But the three things it does well is it basically moves money forward in time. It moves money from like the, you know, from like, you know, the when you sell period to when you need it period, which is when you're designing the product and buying parts, etc. So, you know, that's really the problem with making stuff is that you don't get paid until it's all done. You need the money at the beginning of the process. So by taking pre-orders, it just moves money forward in time to when it's needed. The second thing it does is it does market research as to whether people really want the product. And that's super, super important. And the third thing it does is it creates a community that's basically invested in the product and will market and sort of evangelize for it and be your boosters and supporters and reality check and all that. And I think all products basically should have communities going forward. So of those three, we only needed – we already had two of them. We weren't money constrained. We had plenty of money. So that wasn't an issue for us. And we already had a community in DIY drones. So that wasn't an issue either. The market research, you know, probably would have been good. I think we – you know, there's other ways to skin that cat. Just, you know, in our – just kind of sharing designs with the community and getting feedback is a good way. So I think that wasn't super crucial for us. So I think in our case, because we already had a community and, you know, and money was not a constraint, Kickstarter would not have been the best process for us because there's such a level of commitment. You're basically promising to create a product. Oh, yeah. You know, and people are – sort of feel like they're buying that product. And if you change gears a lot in the process, as inevitably happens in development, you get – there's a slight worry that you've kind of bait and switched them. And, you know, I kind of reserve the right to change gears. That's right. Because I usually don't get it right the first time. So that's the one reason why we probably would not have used Kickstarter. But others who didn't have their own community or didn't have, you know, enough money up front, I think is a great option.

Dave Jones: That's always the problem when you've taken somebody's money, you know, and you've promised something. That is another pit of despair in itself or it can be anyway.

Chris Anderson: You know, you take a dollar of money and you basically – people feel like you owe them their life, your life. It's just incredibly demanding. So no app development in your future then? I mean – Well, you know, we do take people's money for hardware. That's true, yeah. One of the real complexities in our world is that, you know, we sell hardware. But this is just commodity hardware, the basically Arduino boards with commodity sensors. But the software is created by a community of volunteers and given away for free. And, you know, the hardware usually works fine. But people, you know, have – you know, find bugs or performance issues or the documentation or whatever with the software. And they're like, why don't you do a better job of supporting this? And we're like, well, you know, we actually didn't sort of create the software. We didn't – it's free. It's kind of created by, you know, the internet. And, you know, you didn't pay them, right? You got that for free. And people just don't understand the distinction.

Chris Gammell: You don't think it's getting any better? I mean –

Chris Anderson: I think that, you know, at the end of the day, you know, the hardware side does have – I think, you know, maybe not a responsibility, but it's – they're well-advised to be able to do some kind of software support. But, I mean, look at the examples with Linux. You know, Red Hat – so Linux is basically given without warranty and, you know, maybe without enough documentation. That was a market opportunity for companies like Red Hat and IBM to provide third-party services for those who want sort of full end-to-end support. And I kind of feel like the same thing is going to happen with open source – you know, the open source hardware movement is that, you know, let's say you get your – I mean, MakerBot actually does quite a good job of handling both of it. But let's imagine that MakerBot sold you the gear and then, like, it was like, hey, and you're on your own with the software and you're like, well, either I'm, like, really competent with Python and I can handle that or I'm willing to pay extra for someone who's going to give me kind of, you know, full-on service. Yeah.

Chris Gammell: Do you want the extended warranty, right? Exactly.

Chris Anderson: So that would be an opportunity for people to sort of sell kind of value-added MakerBots. Yeah. And I think, you know, you're already seeing people with the UAVs, you know, packaging them up and selling them, you know, system-integrated in planes with service attached. And, you know, we think that's great. The more the merrier.

Chris Gammell: That's awesome. So how does it play off against – I mean, I wrote on the discuss forum where people were asking questions not to ask about the book, but I have to ask because you kind of said it there. How do you play that off where your book kind of says, well, you should give stuff away for free and then you say, well, giving stuff away for free kind of sucks?

Chris Anderson: Well, the book, in fairness, says that it's all about freemium, which is basically you should give things away to drive sales. Right. So our model is right out of the book. As a matter of fact, it's even covered in the book. We give away the software. We give away the bits and we sell the atoms. I've heard that before many times. That's Neil Gershaw as well? Someone said that. He – Gershenfeld has the atoms and bit construction. I'm not sure he does the economic side of it as well. Okay. But yeah, so the software is free. Hardware is paid for obvious reasons, that the marginal cost of distributing software is zero and the marginal cost of distributing hardware is not zero. Correct. And so that divide is – I mean a lot of people sort of read the title of the book, which is free, but didn't actually read the book itself, which is all about getting paid. And so they're like, I can't believe you charged for this book. And you're like, maybe you should look up freemium. You know, it's about finding a way to build – you know, to use free as a marketing tool to give people something of value while getting a fraction of them to upgrade to a paid version, which is more value.

Chris Gammell: Full disclosure, I have read the book. I just – it was a couple months ago, so that's why. Okay. I should have been more memorable. That's my fault. So speaking of the forum, we should get to some of their questions real quick. There was a lot of questions about the actual – you know, the UAV side of things and a lot of news items about, you know, robotics in the sky kind of coming under scrutiny now and the FAA and all that other stuff. Kind of what's the deal with that and where do you see that going?

Chris Anderson: Sure. I mean let me give you a quick overview of where it's at. So I'll give you the U.S. rules. There's somewhat similar to the European rules. Australian rules are actually a little bit more liberal. But I'll just give you the U.S. rules. Right now, there's – the FAA controls the national airspace. And they're – right now, UAVs can fly under basically model airplane restrictions, which is under 400 feet. So manned aircraft are like – their floor is 1,000 feet. There's a 600-foot buffer and then there's a 400-foot where like kites and balloons and model airplanes can go. So under 400 feet, you have to stay within visual line of sight and that's to achieve what's called sense and avoid. So that means that – so the number one rule about the airspace is that you've got to be able to see, you know, other aircraft to avoid collisions. Now, if you don't have a person in the cockpit, then you don't have a good way to see. Even if you have a video camera in the cockpit, you don't have the same peripheral vision. So you don't have that same sense and avoidability. So you need someone on the ground to be able to see the aircraft in the sky and see the sky around the aircraft to avoid any other aircraft. So that's the visual line of sight rule. Then you need to stay away from buildup areas, you know, over populations, away from airports and things like that. Oh, yeah. So that's the way we fly. Now, that's only for non-commercial use. For commercial use, you have to get this crazy – you have to go through this paperwork process to get what's called a certificate of authorization, a COA. And that's really hard, practically impossible for most people. So what's really interesting is that my nine-year-old can fly drones on the weekend in the park and often does. But like, you know, Lockheed Martin can't or the University of California – or the University of California, Berkeley around – you know, next door to us is not allowed to fly, but my children are. And that's all because we're amateurs and they're commercial. Now, in 2015, the FAA is supposed to be releasing a set of rules that sort of create some categories by which small and, you know, relatively, you know, not dangerous UAVs could be allowed to fly commercially with a kind of a minimum regulatory burden. And, you know, we'll see what they define as small. We'll see what they define as minimum regulatory burden. We'll see what the rules are. But that should allow commercial use of UAVs. Right now, people are doing it but kind of under the radar and, you know, kind of out of harm's way. But going – you know, around 2015, it will probably be delayed. But around that time period, there should be a way for them to do it legally and safely. And we're really hardcore about adherence to this because it's just common sense. This is just safe. And, you know, one of the nice things about UAVs as opposed to radio-controlled airplanes is that we can actually enforce this. You know, you can create – you can tell the thing that if it gets beyond a certain distance away from you, it flies home. And, you know, so flyaways are a problem with radio-controlled airplanes. But with UAVs, you just click a button and it returns to launch.

Dave Jones: But how does the – how does that autonomous aspect come into it? Because you're no longer in control or you may not have line of sight. Should it be allowed to fly home or should it just immediately land where it is regardless? I mean, that's all –

Chris Anderson: So we don't hard-code that. But basically, all of our drones have the ability of manual override, which is that, you know, if anything goes wrong, you can always – you always can and should take over manual control. We can't stop you from flying out of visual line of sight, but we can make it very easy to do it accidentally – just to not do it accidentally, rather. So, for example, we have a geofencing feature by which you just describe a 3D box in the air, and it won't leave that box. By the way, it's a great training feature for kids or people learning how to fly. Basically, you can't crash. You can dive it straight to the ground, and when it hits the bottom of the blocks, it will take control away from you, fly back to the middle of the box, and then return control over to you. That's awesome.

Chris Gammell: Yeah. I actually have very limited experience with UAVs. Most of mine comes secondhand from Dave talking about his. So I haven't taken that leap yet, but that sounds like a good way.

Chris Anderson: You know, everyone is really into, you know, multicopters, quadcopters, hexicopters, et cetera. And, you know, we've got a lot of them, and that's – most of our community is into that as well. But we actually recommend people start with planes. You know, they're real foam planes that cost like 60 bucks. The nice thing about the planes is that, you know, when things go catastrophically wrong, as they sometimes do, they don't fall out of the sky. They kind of glide. Rock mode. And, you know, when they do fall out of the sky, they kind of bounce, you know. And they hit you in the head. It doesn't hurt. And they're really kind of – you know, they're kind of fun and lovely and just a relaxing experience. Whereas multicopters are just inherently unstable and inherently un-aerodynamic. And, you know, catastrophic failures are expensive with copters and pretty cheap with planes.

Dave Jones: Well, you know what they say about helicopters. They actually beat the air into submissions. That's how they fly. You know, they – Right. And a hexacopter does it six times. They shouldn't have any right to fly. I know. I know.

Chris Anderson: And it's just – it drove me crazy watching the Avengers. You know, when they had that, like – that had that aircraft carrier quadcopter. It turns into the quadcopter. And, like, one motor goes out, and it just stays horizontal. Yeah, I know. It's like, that's not what happens.

Dave Jones: It's going to be flipping crazily to the ground. Yeah. Oh, man. That's funny. What other questions have we got on here? I think they're a lot related to the electronics manufacturing side of things as well. But – Yeah.

Chris Gammell: Tony Rob wrote in about – he asked, what do you think is missing from homebrew UAVs? And what would you like to see in RG Pilot 3? And he said, what silicon hardware do you think is missing from homebrew? Oh, yeah.

Chris Anderson: So, I mean, we're all upgrading to the ARM core right now. I mean, we've stuck with the Arduino platform. We're big fans of the Arduino platform, not because it's about feeds and speeds, not because it's the fastest and most powerful processing board. It's not. But because it's got the lowest barrier to entry. It's got the best community, the easiest to use development tools. We believe really strongly in sort of inclusiveness and accessibility. And we thought that was more important than clock speed and all that. And the truth is we have a really good performing autopilot out there on basically an 8-bit processor. So, everyone thought – everyone who said it was about feeds and speeds, it turns out it's more about algorithms. And the best way to get good algorithms is with a big community. And the best way to get a big community is to have a very accessible, easy-to-use development tool chain. So, we're big believers in Arduino. Arduino is moving to the ARM processor later this year. They've already released – Oh, wow. I didn't know about that. Yeah, they've released a development board called the Due, which is built on the SAM3 – Atmel SAM3X. Oh, that's right. I think it's a Cortex-M3 core. So, we're going to be upgrading with them. So, we'll be on ARM by Q4. In the meantime, there are other platforms, including one that we may be talking about later this week, that are built on the STM32 – Oh, yeah. Anyway, the 168 – yeah, the 168 megahertz ARM core, which is I think the M4, if I'm not mistaken. Anyway, I'm probably – I get really confused with ARM nomenclature, but – Oh, yeah. Yeah. It's endless. It's endless. At any rate, we all have – we'll be on ARM across the board by the end of the year. Now, the question is what are we going to do with that? What do we need the power for? And the answer is – it's not immediately obvious. Right now, we're not even – we haven't even maxed out the existing Arduino boards. We're running at about 60 – about 60%, 70% of capacity. We have the chipset, the sensor chipset we use, the InventSense, InventSense MPU6000, has a built-in processor right in the chip, which does sensor fusion. It's called digital motion processor, and we're going to be turning that on next – in the next couple weeks. That will free up a lot of processing power from the main board as well if you want to use that. So we have – we still have a lot of kind of overhead in the existing chips. But as you go forward, you know, when you think about what you might do with like infinite processing or close to it on these new chipsets, the first is you run all your algorithms faster. You know, so you're doing it – you know, rather than doing our sensor fusion at 100 hertz or 200 hertz, maybe do it at 400 hertz. You know, that's good. The second is you run a real-time operating system. So, you know, a lot of that kind of preemptive – you know, all of that time scheduling stuff and the preemptive stuff gets easier and you don't worry about collisions and, you know, adding a feature that ends up slowing down or messing up other features. And that's good. It should ease development. A third thing, and this is probably the most important, is learning. Right now, one of the biggest problems with UAVs is that you have to configure them and train them. All aircraft are different, you know, and especially in open source ones where everyone can put together their own frame and different motors and props and, you know, and dimensions and all that. All that has, you know, really kind of profound implications on the aerodynamics. And you can have – there's like, you know, dozens if not hundreds of gains and PID settings that you optimally would tune for this.

Dave Jones: Tell me about it.

Chris Anderson: It's really hard. And if you don't buy the stock platform, you know, some tuning is involved. That should be a learning thing. That should be machine learning. Basically, what should happen is that you should take off and fly it manually a little bit. And in the course of flying it manually, it should just figure out for itself what those gain settings are. Of course it should. Yeah.

Dave Jones: That's an obvious step forward.

Chris Anderson: So that's going to require, you know, that's going to require a lot of AI, basically. It's a lot of pretty hardcore statistics and no small amount of memory as well to, you know, to record all that data and process in real time. It could be done on the ground, actually. And, you know, we have a relatively fast telemetry link that can – that does do a lot of, you know, some kind of AI, including scripting on the ground. So you could do it on the ground right now. But ideally, it would be, you know, a black boxed thing in the air. So that – I think learning is kind of the frontier of robotics in general. And that's probably true for aerial robotics as well.

Chris Gammell: That sounds really impressive and really hard, actually, like you said. And any thought towards, you know, moving towards parallel processing? I mean, you said RTOS, which is, you know, quasi-parallel because you do scheduling either on Robin or whatever else, you know, the RTOS is. But ever – any thoughts of FPGAs or is that too far up?

Chris Anderson: You know, it's interesting. I mean, we only hear about FPGAs from, like, newly minted engineers. Oh, come on. Well, it turns out that, you know, what's going on because of smartphones, what's going on with just the standard ARM cores is – you know, the Moore's Law is so fast. I mean, it's accelerating so quickly. There's just so much power coming out from these kind of standard platforms that we don't really see the need for custom hardware. It's just the kind of commodity hardware is just getting so good so fast. And I think that's taken a lot of the, you know, motivate – a lot of kind of the appeal, FPGAs away, that you can just kind of throw silicon at it. You know, commodity silicon and things get done. And so we do parallel processing already. You know, on the existing Archipilot board, we have the APM2. We have three processors, you know, running simultaneously. One handles the PWM, you know, the RC input and processes that. Another one is the main processor. Another one is the DMP processor on board the sensors, the gyronic accelerometer sensors, which can be doing some sensor fusion. So there's that kind of, you know, physical separation of processing. And then as the chips get more powerful, then you can just, you know, preemptively multitask with software and an RTOS as well. But there are people who use Parallax propeller chips and have tried to do that. I mean, there are some autopilots out there. I don't know how many of them have kind of really effectively multitasked or parallel processed with that. And I just don't know enough about the architecture. It hasn't really taken off. I think there's only two autopilots out there that use it. So for whatever reason, it hasn't been a kind of a gating factor in our business.

Chris Gammell: I mean, I guess I didn't think it – I didn't know it was like 100 hertz kind of stuff too. So that is lower than I would have guessed. I don't know why I thought it would have been higher. But in terms of scan rates and everything.

Chris Anderson: Yeah. I mean, 100 – so basically when you just do the math, I mean, when an airplane, you can't move a control surface faster than 50 hertz. The surface only goes at 50 hertz. So there's no point in doing your computations any faster than that. For copters, you're basically changing the speed of a motor. And past about like 200 hertz, you're talking about less than – you're changing the speed within a revolution. So you're multiple times within a revolution. And it's really questionable whether there's any kind of physical reality of that. Yeah. That makes sense, yeah.

Dave Jones: Where do you see the autonomous aspect of these things going? In terms of – you talked about the increased processing power afforded by the high-end ARM cores. Does that mean the community is going to get more – the stock platforms are going to get more into 3D mapping, real-time mapping, and all that sort of stuff so that they can actually map where they're going. They can actually map objects and avoid them as they fly back autonomously and all that sort of stuff.

Chris Anderson: Yeah, exactly. Where is that heading? So right now they're full – in robotics, you talk about sort of the inner loop, which is kind of the basic stability. So if you're thinking about like the human equivalent, you're walking down the road. What keeps you upright? It's your inner ear, right? Your inner ear. Your balance. Your balance, yeah. So just kind of knowing where down is, we call that the inner loop. Then there's the outer loop, which is kind of where am I going? You know, left or right or just figuring out what direction to go. And that's kind of basic autonomy. That's following waypoints, automatic takeoff and landing, et cetera. Kind of, you know, you can dynamically change it, but it's basically just sort of following commands. Then there's kind of the third loop, which you think of as sort of as the kind of, you know, what's my mission here? What am I really trying to do? And this is the loop of like, you know, I'm trying to go to the store and I don't want to run into people. And, you know, I need to wait for the stoplight and things like that, the kind of AI layer. And that's the layer. That's, I think, where autonomy is still going, is starting to get interesting. By the way, the military is really not there either. You know, the amateurs are pretty much up there with the military in terms of autonomy.

Dave Jones: I think they're way ahead.

Chris Anderson: When you think of like a Predator out there or one of these military UAVs, there's like 60 people on that crew. I mean, you know, with an airplane, there's – yeah, it's between the operators and the ground crew and the, you know, the kind of the system analysts or the mission analysts and the CIA, you know, analysts, et cetera. I mean, there's just tons of people working on those crews. So they're really – there's almost like no fewer people than a manned aircraft. And they're often actually, you know, not flying autonomous missions. They're doing the inner loop themselves, which is they're staying upright. But the outer loop as to where you're going is being actually point and clicked on a map. So the amateur stuff, you know, the Archercopter and Archerplane and others like it are basically fully autonomous in that respect, which is you give them a mission, automatic takeoff, you know, follow waypoints. They will avoid – they will kind of, you know, follow the topology, which is to say that, you know, they know – with Google Maps, they know the elevations of the various waypoints and you can maintain an elevation for that. Then it will finish its mission a little – you can program it to look at a certain point of interest, keep the camera focused on it, loiter at certain areas, you know, do various camera functions and then land. So that's a fully autonomous mission. But what it isn't doing is, you know, noticing that, oops, there's a telephone line, you know, in the way or, oops, the cloud came in and, you know, and now my camera can't see anything because you're over the clouds or all that kind of stuff. And that kind of, you know, adaptive mission, you know, control, you know, is the next big step. And it's going to start with – it's going to start with things like image processing, you know, optical or LiDAR. You know, as you probably heard, one of the big, you know, crises right now in the UAV world is GPS spoofing or just loss of GPS blocking. And you take GPS away from ours and they're dead. I mean they don't have any idea where they are. Now, there's a couple ways to solve that. I mean one is just dead reckoning, sort of saying, oh, I lost GPS. But, you know, still I kind of know which direction I'm going and where I came from and I can kind of backtrack through just kind of, you know, inertial, you know, data and make it more or less back into it like GPS again. The second is actually looking for other wireless sources and just sort of saying, well, you know, here's some kind of known frequencies, known radio signals out there. I can tell through Doppler effects and others and signal strength what's – you know, whether they're getting weaker or stronger, what direction I'm going. So let me just sort of fly along the gradients of the existing RF space. So that's really kind of interesting and very doable because there's RF everywhere. And the third is imaging of the ground beneath you with LiDAR or optical and sort of saying – and having a database on board that, you know, that knows what the ground looks like and sort of saying, whoa, okay. Well, I mean what would you do if you were a pilot and you lost – and you lost, you know, your radar, your compass or your GPS? You'd look down to the ground and sort of say, do I recognize anything? And, you know, the great thing about drones is that they can have, you know, an infinite database of terrain characteristics. So, you know, between LiDAR and optical and other kind of multispectral imaging, they should be able to kind of look down and kind of find their way home.

Dave Jones: All right.

Chris Anderson: So that's the next frontier.

Dave Jones: That sounds exactly what I need for my canyon copter, my autonomous canyon navigating. Actually, my friend Roger who's working on this with me, he'll probably kill me if I don't ask you. We're developing this canyon navigating quadcopter, okay. It's a standard quadcopter based on the RG pilot. But we want it to navigate autonomously through a camera and get one continuous – canyon, sorry – one continuous video shot of the trip through. And we don't need it to do anything fancy. We need it to move slow, steady, avoid obstacles up and down, rocks and water and stuff like that. It's in a canyon. The walls are only 10, 20 meters wide at most. There's no GPS reception. Yeah, that's hard. What are you – are we – well, yeah, we feel like no one else in the world seems to have done this. We're probably biting off more than we can chew.

Chris Anderson: Well, we actually – so here's a couple of things. We do support sonar already. Right. Yes, we've been thinking about that. Yeah, so we do support sonar right now.

Dave Jones: How accurate is that? Because we're only like a couple of meters off the ground here. We're not going to be 300 feet up. We're going to be – Sonar is five. Oh, yes.

Chris Anderson: Sonar is actually ideally 15 meters with – 15 and less. That would be our tops. So sonar is good within a 15-meter range. So we absolutely support sonar. We actually use sonar already on APM1, which had a kind of – which it shouldn't have a great barometric pressure sensor. We actually use sonar for altitude hold. On APM2, we've got to improve barrow so we don't need that. So we support sonar. There's no reason why you can't have sonar in all four directions. If you go onto our site –

Dave Jones: I was going to say, that's what I was going to ask. What about the different directions? Because we basically want to center the thing within the walls of the canyon. So we need left and right. Not a problem. And then forward and down.

Chris Anderson: Not a problem. As a matter of fact, if you go onto our site, a guy – a member named Jose Julio has a great shot of him playing tennis with a quad copter. Oh, nice. I've seen that. Yeah, it's great. So basically they just have the – it's a very slow motion tennis, but they basically swing at the quad and the quad sort of moves away and moves in the direction of the other guy and the other guy swings at the quad, et cetera. So if sonar is good enough for you, sonar is totally doable. Right. Optical gets harder. Yeah. We do have optical – we also support optical flow and we actually sell an optical flow sensor. And the nice thing about optical flow is that you get these kind of vector fields that will give you a sense – they'll give you a sense of kind of objects. Obviously, as objects get closer, their vector fields diverge. Oh, yes.

Dave Jones: This is sounding promising.

Chris Anderson: Yeah, but it's probably more – I mean to do it in four – in what? I guess in six directions.

Dave Jones: Yep.

Chris Anderson: It's probably more processing than you want to do to which I would say wait until later this week.

Dave Jones: Oh, okay. There we go. Dave's going to like this week. And or are we better off doing like a waypoint base actually going through and then programming the – actually, ideally, right, we wouldn't even want to fly it manually. We want to just move it along because we can walk through the canyon. So hold it here and then – like we can just manually hold it there and then program that as a waypoint. Yeah. So it would need –

Chris Anderson: So I think – so basically our waypoint – so GPS is good. You know, should – I think with the 2.7 code that we're going to be releasing this week as well, you should be able to stay within about a sort of a three-meter box.

Dave Jones: That's not good enough for us. Yeah. It needs to be – okay.

Chris Anderson: Because, you know, winds can – you know, we try to stay in a one-meter box but, you know, winds, et cetera. Also, in a canyon, you might have GPS multipath problems. Oh, yeah. So I wouldn't, you know, bet my life on standard GPS on that. Yep. No. You could fly manually, you know, and just –

Dave Jones: Yeah, I know. Yeah, we could. But then the canyon snakes, twists and turns, and we wouldn't be able to keep up with it because we have to abseil down waterfalls and we've got to do all sorts of nasty things.

Chris Anderson: Well, so you say – so what's the width of the canyon?

Dave Jones: It can be anywhere from two meters wide to, say, 20 meters wide.

Chris Anderson: Yikes.

Dave Jones: Very, very, very narrow. 20 meters would be good.

Chris Anderson: Two meters, I just would not – I think that's just too hard for GPS.

Dave Jones: That's the thing. I mean, we – ideally, we want something – a quadcopter's overkill, they're designed – you know, they're very maneuverable and things like that. We don't – we almost wanted to do like a blimp, like a remote control, you know, helium blimp, nice and steady, stable, just goes forward and, you know, it just gently turns left and right and – but, you know, you can't lift the payload. The EUV log Zeppelin, huh?

Chris Anderson: Yeah, we actually – we started with the blimps. Blimp Duino was our first one. Blimps are fantastic indoors, outdoors, to be able to fight winds and thermals and things like that.

Dave Jones: We're effectively indoors. We're – with a canyon, we're effectively indoors. Oh, really? So – Oh. Well, pretty much. These are pretty enclosed canyons. They're almost cave-like at some point.

Chris Anderson: Our blimps were too small for a proper payload, but there's no reason why you couldn't do that with a proper blimp. Blimps are lovely. They're real fun to work with, with vectoring thrusters. You can get pretty good altitude control. You'd probably want a blimp of at least, I would say, you know, two or three meters. I know. That's the problem.

Dave Jones: Therein lies the problem. Sharp rocks. That's the real problem. You can haul all this gear down into the canyon. Right.

Chris Anderson: Yeah, and then there's sharp rocks and all that. Yeah, that's a tricky one. I mean, you know, I mean, is there a stream or anything underneath it?

Dave Jones: Yes, yes. There's water. It has to be fully waterproof, so it lands in water, sand, rocks. So here's a dumb question.

Chris Anderson: I mean, why not just have like a raft that goes down the stream and just a balloon attached to the raft?

Dave Jones: Well, you can't, unfortunately, because it's all rocky and there's waterfalls and there are all sorts of – all sorts of stuff.

Chris Anderson: Oh. Oh.

Dave Jones: So, yeah, unfortunately. Yeah, I know. It's difficult. That's, you know, why no one's ever done this.

Chris Anderson: Well, there's one last thing. FPV. Oh, yeah. FPV. Flying first-person view.

Dave Jones: Oh, yes. Yes, we were thinking about that. Yep.

Chris Gammell: Those are great videos, too. If people aren't already subscribed, the DIY drones feed is spectacular for that kind of stuff. Every day you basically get a new FPV kind of video from someone and it's just –

Dave Jones: Which is – if you don't know, FPV is first-person view. Yeah, flying with video. So there's a camera mound on the front which transmits video. Exactly.

Chris Anderson: You're flying with video. You're flying as if you were in the cockpit. You're wearing video goggles. It is on some. You have a real-time video feed. It's –

Dave Jones: Yeah. That's probably our only way to do it in the end, I think. It takes a little getting used to. But then when the canyon moves – you know, but when the canyon goes around a corner, our RF is just going to drop out, I'm sure. We haven't actually trialed it yet.

Chris Anderson: Well, you know, maybe. I mean, you know, here's the thing. FPV comes in three frequencies, 1.3, 2.4, and 5.2 gigahertz. And, you know, depending on – as you know, different frequencies bounce in different ways around corners. You can also – depending on whether you want to kind of bend the rules a little bit, you can go all the way up to a watt of out. Oh, wow.

Dave Jones: Yes. Well, this is in the middle of nowhere. This is like in the middle of the wilderness. So, yeah, we can –

Chris Anderson: Yeah, so I think you throw enough juice at it. Pick the right frequency. You might be good. Before you do it – Go through the rock. If you can borrow – if you know somebody with one of those Parrot AR drones, do try to fly that in FPV mode and see whether you can do it because for the life of me, I can't do it.

Dave Jones: I think that's what we have to do. I think we have to get one of those cheap Parrot drones and go down there with our iPhone and try and –

Chris Anderson: Yeah, the nice thing about going with a Parrot AR drone is that if you do hit the wall, it's not a disaster. They've got these nice foam rings on the outside. If you crash, it probably won't break. Well, we have to make it waterproof. That's the only thing.

Dave Jones: Is it able to lift the payload?

Chris Anderson: Well, it's got a pretty decent camera. The new AR drone too has a high-def camera.

Chris Gammell: Yeah, 720p built in. Although that's the forward-looking camera, not the downward one. Nah. Chris, there's one last question from – besides Dave's troubleshoot – How did you for troubleshooting? I wanted to hear a little bit about your book before we go because you've got a new book coming out in October and we wanted to talk about that real quick. And it kind of fits in.

Chris Anderson: So the name of the book is Makers, the New Industrial Revolution. And it's really about what I think is sort of the future of Western economies, which is taking the web's innovation model, which is all about digital stuff and open source and community, etc. And applying it to making real stuff. And it's basically that 30 years ago, we introduced the word desktop to computing with the Apple and the laser printer and all that kind of stuff. And it just kind of – it changed the world. And we now have the word desktop manufacturing. We talked about 3D printers at the start, but there's other digital fabricating tools, CNC machines, laser cutters, and so on. And these things are now right about at the point where regular people can use them. They've become democratized. And what's great about it is that because they're digital and because these ideas all start digitally, they can be shared. And suddenly everything we've learned about doing stuff together and about the advantage of innovation models in the last 20 years of doing the web, we can apply to manufacturing and making real stuff. And suddenly it's not about who's got the cheapest labor. It's about who's got the best ideas. Exactly. Yeah. And so that I think is – I think – I live in the United States and so obviously the future of the American economy is important to me. And real countries make stuff. Amen. We still make stuff, but there was a sense that it was a losing game. And I think that the day when the cheapest labor wins is gone and now the best ideas win thanks to these digital tools bringing manufacturing back to everybody.

Chris Gammell: As an Ohioan, I – yes, we have lots of – well, we used to have a lot more, but manufacturing is a very strong focus around here.

Chris Anderson: Exactly. And it will be again, I think. Those skills, I hope they're not lost because those skills are going to be very much in demand.

Dave Jones: They're never lost. They're always lurking. Yeah. They're always lurking.

Chris Anderson: And I can tell you that as somebody who originally did manufacturing in China and then brought it back to the United States myself, what you might lose in a little bit of margin, you gain in flexibility and quality control and all that. And once you start to see these kind of more high-value, high-innovation manufacturing, you realize that you add a little automation. You add a better innovation model. You start to see a kind of domestic supply chain building as we're starting now seeing electronics and suddenly making stuff in the United States make sense again. And it worked for us. And I think what we're seeing with the maker movement, which is what this is called, is really just the kind of the homebrew computing club start of what is ultimately going to be the computer industry. So I think we're on the verge of something that's – the web was big, but manufacturing is bigger. And to see – in the same way the web transformed media and content and information, I think the same forces are about to transform manufacturing, which is something like 70% of the world's economy.

Chris Gammell: How do you see the interplay between like the really big companies and the desktop manufacturing side of things? Because we always hear the things about desktop manufacturing. We've seen it ourselves in terms of 3D printing and like you said. But I mean what about the large manufacturers and stuff like that? How do you see that all fitting in?

Chris Anderson: You know, it actually takes me back to my first book, The Long Tail. And everyone said, you know, oh, this is the end of the blockbuster, you know, the end of the mass market, the kind of the Hollywood model of content. And I was like, no, it's the end of the monopoly of the blockbuster. I mean I think, you know, right now big manufacturers basically own it, you know, because of their economies of scale. And, you know, if you're not operating at Walmart scale, you can't compete. And I think what that means is that a lot of products designed for kind of more niche audiences just don't get made because big manufacturers don't want to operate at that scale and because retailers, which are the usual, you know, process of selling these things, don't want to operate at that scale either. And, you know, between the ability to do more kind of small batch manufacturing efficiently, which is what automation, you know, allows you to do, and the internet as a kind of a growing, you know, retail distribution channel for products means that suddenly you can have a long tail of things as well. That's awesome. So, you know, so again, the same way that, you know, mainstream media, you know, wasn't killed by the internet, but it was just, you know, the internet just created, you know, a huge amount of competition. And so you still have Hollywood, but you also have YouTube. I think you're going to start to, you know, you're still going to have mass manufacturers making, you know, making, you know, toys or, you know, or commodity electronics.

Chris Gammell: Or something like that.

Chris Anderson: Exactly. But you're going to have that, you know, that long tail of stuff that's designed for smaller audiences and maybe fits them better.

Dave Jones: That's great. Speaking of YouTube, I, of course, make my full-time living from making YouTube videos. And I wanted to ask about Wired magazine, go a bit off topic, UAVs all show. Wired, where do you see the print media going? Because Wired is a print magazine. It started out as a print magazine. Did it or was it always digital?

Chris Anderson: No, it started as a print magazine. Is it just print?

Dave Jones: And how has that moved? How is digital in comparison to print, you know, the print edition now? And where do you see the video side of things going?

Chris Anderson: So we started in 2003 as a print magazine. So we're hitting our 20th anniversary, unbelievably. I joined about 11 years ago. How did you join, by the way? I just, if I forget to ask. Yeah. How did you get involved? I was, I mean, my background's really kind of funky. I'm a kind of a high school failout, a college failout, played in like punk rock bands for most of my 20s. And then, you know, sort of late in life, decided to get serious about school and went to become a, went to study computational physics. And just in time. As you do. I thought after having failed for so long, I had to prove that I could do hard stuff. So I found the hardest thing I could do. And I was kind of okay at it. But then kind of physics went all pear-shaped because the accelerators got too expensive. So I ended up, I had to find something else to do. And I kind of decided the next kind of, you know, adjacent space was the science journals, if not science itself. So I went to work for Nature and Science. And then I got recruited by The Economist because they tended to hire people from the science world to run the science section. And this was in 93, no, this was, must have been 90, I guess 94 or so. So, and I said, you know, I, rather than do science, there's this new magazine out called Wired that says that like the internet's going to be really big. And so I think I, I think we should, I think I'd like to cover the internet. And they're like, okay, what do you need? And I said, I'll need a computer and an internet connection. So they got their first computer and they called British, we was in London, they called British Telecom and said, he, this guy wants an internet connection. I think they dropped a T1 right into the back of the, of the PC. My bandwidth has gone down every year since then. Yeah. Yeah. Doubled in price too. Yeah, exactly. There's a book in there, Chris. But so what I ended up, sorry, I was at The Economist and I was based in London and then Hong Kong. And then I was running a New York bureau when Condé Nast, a big, a big sort of, you know, magazine publisher that has titles like Vanity Fair, New Yorker and Vogue and things like that. And Reddit. A lot of fashion titles. Yeah. And Reddit. I had bought Wired a little bit earlier and wasn't, you know, it was kind of a hedge and kind of a, you know, laboratory for the future. And they, they asked me if I'd want to take over and I'd like, Wired changed my life. You know, it turned me onto the internet. So I was like, absolutely. So there's, there's a very long and convoluted path that took me from punk rock bands to, to Wired. And yet every step makes sense. That's correct.

Chris Gammell: And yet the whole path is completely nonsensical. And then further on to hardware and, you know, RG copters and everything else.

Chris Anderson: This is kind of, this is back to my, back to my roots. I mean, you know, sort of, you know, you know, open source hardware is really just, you know, the kind of same boards we were building for physics, you know, laboratories. It's just, you know, more mainstream adoption. Anyway, so the simple answer on the magazine is that. Yeah. The magazine is that started as a print magazine. It then very quickly started a website and was actually one of the first. It did the first, arguably the first blogs. I don't remember. Remember suck.com and all that. The first banner ad was on Wired. Some of the first content management systems. We developed Apache, you know, the server operating system. So that was, so that was really early days. Then, then, then, then that all kind of went along until about two and a half years ago when it became clear that tablets were going to be interesting. And so we were one of the first to have a tablet edition on the iPad. And I'm, I believe, I believe today we are tied with the New Yorker as the most popular iPad magazine. So tablet magazine, actually, we're on, we're on all platforms. It's a gorgeous version, too. It's very nice. Just, just came out on, on all the Android tablets via Google Play. So I, so the simple answer is that I think the, I think the future of print is, is, can be sort of foreshadowed by the future of books. Which is to say that today you have a choice between a print book and a, and a, and a digital book. Earlier you said you wanted the Kindle edition and, and, and me too. I prefer the digital books. And yet, and yet lots of people still prefer the physical books, even though they have that choice. And sometimes they prefer it because they just like the touch and feel of the thing. Sometimes they want to put it on their shelf. Sometimes they want to give it as a gift. You'll notice the production values of physical books are, are a lot better. I mean, the covers are embossed and die cuts and fancy. And, you know, on a Kindle, you'll be lucky to get a cover at all, actually. So, so for people who value those kinds of aesthetic qualities, you know, the, the print book still has additional value. And I think the same is going to be true for magazines that, you know, increasingly people have a choice between the digital magazine and the physical magazine. And, and, and, you know, the digital magazine obviously is instantaneous delivery and you can get it anywhere you are and you have your whole library with you. And it's, and it's, you know, interactive and multimedia and all that kind of stuff. And so the question is, okay, well, so as people, you know, have a choice between them, how, rather than have the print one be a one size fits all where we have to use relatively poor paper stock, you know, because we're, should we make it? We, we print a million of them every month, you know? So, so, so, you know, we have to, we have to use, you know, we have just for economics, we have to use relatively affordable paper stock and inks. What if we, you know, and I'm, I'm like thinking forward like three or five years from now, you know, what if we decide, okay, so here's the choice. You get the, the digital edition, which is relatively inexpensive, you know, reflecting digital economics. And this is print edition, which we're going to make absolutely beautiful, which we, the best paper stock, the best, you know, ink and, and sort of foils and, and die cuts, et cetera. They're no expense. And you can get a glimpse of it already. If you look at our Italian edition or our British editions, which print much smaller volume, like, you know, 50 or 70,000 copies, they're stunning. You know, they're, they, you know, they, they feel like this, like Fabergé egg, the covers are textured and, you know, they don't use silver ink. They use like foil and they're absolutely gorgeous. And they're a glimpse, I think of, you know, that maybe, maybe the simple answer is this, that, you know, as, as, as people choose between digital and print and, and, and, you know, and, and more people and, and, and people, and increasingly choose print, I'm sorry, choose the digital edition, then print becomes a smaller fraction of the total. Then maybe the print edition can get better. So as print gets smaller, print gets better. Yep. And maybe people will pay more for it because. I was going to say, it'll get higher margin. Exactly. And it becomes an artifact. It becomes a, it becomes a, you know, a really a kind of an object of desire and kind of retain some of its physical properties that are so unique to, to, to, to, you know, to, to, you know, to, and it retains physical value.

Dave Jones: Exactly.

Chris Anderson: The kind of thing you're really proud to have on your, on your, on your coffee table.

Chris Gammell: The way that LPs have made the resurgence as well in the same manner.

Chris Anderson: Exactly. So, you know, as go books, so, so may go magazines.

Chris Gammell: Hmm.

Dave Jones: That's interesting. And where do you see video fitting into all this? Dave's asking for a job, Chris. I don't, I think that's an undertone here. I've seen his videos. I love his videos.

Chris Anderson: I think I've even tweeted his videos. So, so, so video is a big part of what we do. We have a big video production team. I mean, increasingly we, we look, you know, we increasingly look like a, like a, like a Hollywood production house. I mean, we not only have a video team, we have a post-production team. We have an effects team. You know, we have people doing, you know, kind of. Robert Dean, Robert. Well, we're, we're, we're pretty big operation. Yeah. And the reason, the reason we do all this is because we integrate video into the digital magazines, the magazines, you know, on the iPad. And there's an, there's a sort of an obligation to kind of, to bring video up to magazine standards. Now, you know, the, the, you know, the good news is that, is that, you know, people, people, you know, want this and they now have this. The bad news is that that's not really our skill set. And, you know, we, we, we do our best, but, you know, increasingly we're starting to think about, you know, you know, what, what is a magazine really? I mean, you know, I mean, what's a, it's a real question. You know, it's, it's not, it can't be defined as the medium anymore. Is it just a, is it a package of content? Is it long form, short form? Is it, is it design? You know, you know, as it, as, as the, as the images are increasing and kind of Harry Potter-like and that they, you know, they move. You know, what are the, what are the skills you're looking for in the next generation of, of, of editors? I mean, are you going to hire people out of the video game industry? You know, these are all, these are all, you know, interactivity is a big part of what, of what we do. So these are, these are kind of big questions for the future of the business. And, you know, ultimately they'll be answered, you know, over a matter of years, not overnight. I mean, we're, we're, we're still really at the kind of the, the early days of digital magazines. And we've been successful. Not everybody has been successful in digital magazines in part because we're figuring out the medium in part because people are just figuring out the, the way they get, you know, how to get these magazines easily. And how to make it as, as, as, as seamless as music or, or books or, or, or, or, or film is on, on, on, on, on these devices. So I think that, you know, you know, I think it's a work in progress, but this is definitely the future. People, tablets have won. People love consuming their content on tablets. Magazines seem. I haven't got one. You haven't got one.

Dave Jones: I haven't got one. Yeah, you've got a Kindle though. I'm a bit of a Luddite.

Chris Anderson: You've got a Kindle though.

Dave Jones: I've got it. I love my Kindle. Yeah. Because it's the right tool for the job. I love the right tool for the job. I hate the universal tool.

Chris Anderson: That sort of, I think I, well, I'm a pretty big fan of, of tablets of all sorts. Try, try the Nexus seven. It's the.

Dave Jones: I will have to get one. Yeah. I'll have to try it.

Chris Anderson: Yeah.

Dave Jones: I think I'll be a fan of China, but I, like, I will not read books on it. Right. It's ridiculous. I just don't want to read a book on it. I want to read a book on the Kindle. What's the e-ink display has just killed it. Really? No, I see. I, I, I guess, I, I guess I. Not battery life and not, I just don't want to have to worry about the battery. I don't want to have to charge it every day.

Chris Gammell: Dave is a bit of a crank as well. He's a gray beard in training. As soon as he learns how to grow a beard on his baby face, he will be, the first one will be gray.

Chris Anderson: Nah. Right. Dude, let me ask a question. What, what, what, what screen is next to your bed, on your bedside table? What screen? A Kindle. You've got a Kindle. So how do you, so let's say you want to check your email or the web or something like that.

Dave Jones: I, I don't do it there. I sit in front of a computer bloody 16 hours a day. I can already do that all day, every day. When I go to bed, I just, you know, I occasionally just want to read a book and I don't want to have to worry about charging the damn thing. Fair enough. Fair enough.

Chris Anderson: And so you don't watch movies in bed or anything?

Dave Jones: Oh, no, no, no. All right.

Chris Anderson: Dave's got a pretty tight chain around his house.

Dave Jones: So that's called, that's called bed hygiene.

Chris Anderson: He's got good bed hygiene. Yeah, that's good. Yeah. Well, no, actually, seriously, for people who have trouble sleeping, it's very important not to have screens be the last thing you experience before you go to bed. The light sensitivity in your cones actually simulates part of your brain, right? To say nothing of that kind of slight, slight association with work and like things to do. Oh, yeah. Yeah.

Dave Jones: I love my Kindle and so does my wife. She loves the Kindle. We've both got Kindles and we love it.

Chris Anderson: Okay. Well, the Kindle app on the iPad is excellent. I'll just say that. There we go.

Chris Gammell: Right. Well, Chris, we've taken more than enough of your time. Yeah, we have an hour and a half. It has been an awesome, awesome conversation talking with you. I agree. It's very fun. I'm big fan of what you do. Thank you very much. Where can people find all your work? Maybe a quick rundown of sites.

Chris Anderson: Sure. So start with DIYdrones.com. That's the community. It's got something like 30,000 members and it's very active. There's something for everybody there. And my own personal stuff, I mean, you can follow me on Twitter. It's Chris A with the I being a one. I presume you have show notes and all that. That's pretty much it. You'll find 3D Robotics. It's just the DIY drone store. But, you know. Link in books and wired and everything else. Yeah. I'm kind of easy to Google. That's right. Just Google it. It's all out there.

Dave Jones: I couldn't find a wiki page for you. You can't find a Wikipedia page for me?

Chris Anderson: Oh, it's there. It's full of all my crimes and misdemeanors, I'm afraid. There we go. Oh, is it? I couldn't find it. It's a very common name.

Dave Jones: You weren't listed under the Chris A. Really? Well, maybe I got this one wrong.

Chris Anderson: They've got me under writer, yeah.

Dave Jones: Oh. Oh, right.

Chris Anderson: Okay. I know. I know. I'm not allowed to edit my own entry. So it's, you know. Oh, that sucks. I know I'm an editor and I can't edit my own. It's the ultimate in self-restraint.

Dave Jones: We don't have one.

Chris Anderson: Yeah.

Dave Jones: We're always, you know, we're always thinking, oh, we've made it if somebody's made a Wikipedia entry for us. Well, you guys could do me a huge favor.

Chris Anderson: You're allowed to edit my entry. You could go into that. You could go into the Chris Anderson disambiguation page and just write writer slash editor. There we go. Slash entrepreneur.

Dave Jones: We will add this episode in. Yeah, we will. To your wiki page. Thank you. Well, I think it's people want to know. They want to know the man behind the.

Chris Gammell: Good. The everything. The man behind the everything. You seem to do a lot of stuff.

Chris Gammell: All right. Well, thank you again, Chris. We really do appreciate it. You being on the show. We do. It's good fun. Anytime.

Dave Jones: It's been awesome.

Chris Gammell: Thanks. We'll talk to you soon. Talk to you. Bye-bye.

Speaker ?: Thank you.