Today on the flagship podcast of enterprise certificate hacks: The Verge’s David Pierce chats with Riley Testut, founder of AltStore and developer of the game emulator app Delta, about how his app finally made it into Apple’s App Store. Then, David walks us through his experimentation with the many software and hardware solutions for “AI voice notes.” Later, we answer a question from the Vergecast Hotline about the Rabbit R1.
Category
🤖
TechTranscript
00:00:00Welcome to the Vergecast, the flagship podcast of Enterprise Certificate Hacks.
00:00:04I'm your friend David Pierce, and I am having a project day.
00:00:07I feel like being an adult is just like a long, unending list of maintenance tasks.
00:00:13There's always things to hang up, and things to take down, and filters to change, and sinks to
00:00:17clean, and did you know you're supposed to clean your garbage disposal and your dishwasher?
00:00:22Apparently, there's all kinds of stuff like that, and rather than let it hang over my head,
00:00:26I've discovered that what works for me is to just keep a long-running list,
00:00:31and then once a month or so, whenever I'm feeling particularly energetic and have some time,
00:00:36I try to just knock out as many of those things as I can, right? It just feels better,
00:00:41and most of those tasks take like one to five minutes anyway, so I can get up, do a whole bunch
00:00:46when I'm feeling energetic or have a show I want to watch or whatever, and then it's done.
00:00:51And then I don't have to think about it for a month. I'm sure there are a million things that
00:00:54are supposed to be on that list that I don't even know about, but I've made it this far. I don't
00:00:58want to know. I'm sure it'll cause me pain, and I'll deal with it as it comes. That's what being
00:01:02an adult is, right? Anyway, we have an awesome show coming up for you today. We're gonna do two
00:01:07things. We're gonna talk a bunch about the Delta Emulator, because that little retro game emulator
00:01:13has kind of taken over the App Store the last couple of weeks, and it turns out its story
00:01:18is both longer and weirder and more interesting and maybe more reflective of kind of the whole
00:01:24evolution of the tech industry than you think. So we're gonna talk to the person who made Delta
00:01:29about the whole journey to get here and what it says about where we might be going next.
00:01:34After that, we're gonna talk about AI, because I've been testing AI gadgets, AI apps, AI platforms,
00:01:40AI everything for like a year and a half now, and I've found one use case for me that is perfect,
00:01:47and it's awesome, and I want to tell you about it. And this probably won't surprise you, but
00:01:51it does involve talking into a microphone. We're also going to answer a Vergecast Hotline question
00:01:55about the Rabbit R1. We got lots to do. Super fun show. All that is coming up in just a second,
00:02:01but first, I just noticed this like pile of laundry and socks behind me, and it's starting
00:02:06to stress me out. So I'm gonna go deal with that. Project number one off the list. This
00:02:11is the Vergecast. We'll be right back. Welcome back. For all of the last two weeks,
00:02:18the most popular app in the Apple App Store has been a game emulator called Delta. We've
00:02:22been talking a lot about Delta on this show over the last few weeks, both because it's a huge hit
00:02:28and because it seems like maybe the start of a new era for the App Store. Like in the history books
00:02:34of smartphone apps, there might be a before Delta and an after Delta for lots of reasons,
00:02:40but there's actually nothing even remotely new about Delta. The story of this app is actually
00:02:46a full decade long, and it starts with this guy. I'm Riley Testut, and I'm the founder of AltStore,
00:02:52developer of Delta. Riley has had, as you might expect, a pretty wild few weeks. In the middle of
00:02:58it all, I asked him to come on the show and just tell me the whole Delta story. Why he wanted to
00:03:03build an emulator in the first place, how the regulatory and technological world changed around
00:03:07him, and in general, how Delta went from a big hack to a big hit. It starts with Riley in high
00:03:14school, just kind of looking for something to do. I was just back in high school. I think it was my
00:03:19junior year, and I just one day came across just like the open source repo for a jailbreak Game
00:03:27Boy Advance emulator, GPS phone. I was like, oh, this is cool. This is all the code. I can
00:03:33probably just download this and tweak it and then put it on my own phone without having to jailbreak.
00:03:37Because yeah, I've never been a fan of jailbreaking. It's just never really appealed to me.
00:03:41So, I found this code base. I hacked out of it for a couple months, put it on my phone. I put
00:03:48it on a few of my friends' phones. They liked it. Then I uploaded it to GitHub and called it
00:03:52GBA for iOS, and then kind of left it there for a bit. Then people found it on GitHub and started
00:04:00talking about it. So, I was like, oh, there's interest in this kind of thing.
00:04:05So, you just made it as sort of a... It sounds like a mix of a thing that you thought would be
00:04:09useful and just kind of a thing to do. Yeah. I just wanted Pokemon on my phone. That was
00:04:13literally just... I found this. I was like, oh, cool. I can put Pokemon on my phone now,
00:04:17and that's awesome. That's really all I was thinking at the time. Then I put it on my
00:04:20friends' phones, and then we were playing just Pokemon ourselves. Then just completely
00:04:24coincidentally, there was this service that came out around the same time that lets you connect
00:04:29open source GitHub projects and sign it with Apple's enterprise certificate, basically.
00:04:34There was a company that had an enterprise certificate, and you could just connect your
00:04:37GitHub account and then re-sign your apps like that. So, I was like, oh, cool. I'll connect this
00:04:42to GBA for iOS, and then people can install it outside the app store. Maybe a few people will
00:04:46download it. That's basically when I learned, oh, wait, there's a huge demand for this thing.
00:04:52Oh, interesting.
00:04:52I did that, and then within a week, people were posting videos about how to install it. There
00:04:58were lots of Reddit posts. I was honestly caught off guard because at this point,
00:05:02it was still just a project I made for me and my friends in high school. Then I was like, okay,
00:05:07there's demand for this thing. Let's take this seriously. Let's build a real version of this app
00:05:13that's not throwing together, not hacky. That's when I got my friend Paul Thorson on board,
00:05:18also still in high school. We were like, we're going to do a brand new version of GBA for iOS,
00:05:23GBA for iOS 2.0. We're going to redesign it from the ground up, just build it and make a really
00:05:27polished experience. We basically spent all of senior year working on that. Halfway through
00:05:34senior year, we released GBA for iOS 2.0, and that's the version that most people know. That
00:05:39was a version that you installed from my website. You set your date back like 24 hours, and you can
00:05:44install it because of a weird bug that allowed enterprise certificates to work that way, even
00:05:47if they're expired. Yeah, wait. So real quick, just pause for a second and explain the enterprise
00:05:52certificate thing to me. Because I feel like part of the story here is like Riley's adventures in
00:05:58figuring out how to install apps on phones. Yes. It's slowly getting more and more complicated.
00:06:04So this was the easiest hack. It's just, yeah, as long as I found an enterprise certificate,
00:06:09I could just sign GBA for iOS with it, and literally anyone could just download it from
00:06:13a website on the phone. Super simple. And the idea is basically it's like that's the certificate you
00:06:17would use if I'm like a company wanting my employees and no one else to have that app,
00:06:22right? Yeah, the exact same flow. And so it's just, you're not supposed to do it for this reason.
00:06:26And so what Apple normally would do is they would just like revoke the certificate.
00:06:31This is no longer valid, and all apps that are signed with it can no longer be installed.
00:06:35And so the bug that I took advantage of was apparently at this time, like in iOS 7,
00:06:41if you just set your date back on your phone by more than an hour, the checks for whether a
00:06:45certificate is valid or not just don't work. They just don't do it. Sure. You did try to install GBA
00:06:51and Apple would have revoked the certificate. But because you set your date back, iOS was fine
00:06:56installing it. And so it was just like a really weird bug that I could take advantage of because
00:07:02Apple couldn't do anything easily about it. They had to actually fix iOS to address it.
00:07:06They couldn't just be like, okay, nope, turn off certificate. It was about nine months later that
00:07:10they finally addressed it. And then that was when I moved on basically from GBA for iOS is when
00:07:15they, it was like iOS 8.1. They killed the date trick. It was literally on my birthday when it
00:07:20actually happened. And I was like, oh, that's funny. But honestly, at that point, I had just
00:07:24gone to college. And so I was like, you know what? I'm going to move on from this whole thing.
00:07:28GBA for iOS is really cool. I wanted to go on to the next thing. Do you know how many people were
00:07:32using GBA for iOS? Like at that sort of peak moment before it got shut down? A lot of people,
00:07:38like more than 10 million. Wow. Whoa. That's like a zero or two bigger than I thought you were going
00:07:46to say. Yeah. There was a stupid amount of people using GBA for iOS, which is why I was so motivated
00:07:51to do this for this whole time was I just knew that there was a market for this kind of thing
00:07:56that everyone just kind of assumed there wasn't. I'm like, no, GBA got 10 million downloads outside
00:08:00the app store. Like it's just, people just really want to do this, like play old games.
00:08:05Yeah. Okay. So you kind of, you're sort of forced to shut down that project a little bit.
00:08:10Was there a moment where you're like, okay, I'm ready to go to war with Apple. Let's
00:08:14fight this to the death and figure out how to get GBA for iOS back. Or when it, when the sort of
00:08:19exploit gets shut down, do you just kind of say, you know what? Like you said, I'm going to college
00:08:23anyway. This is just a moment. Let me focus on other things. It was basically that. Yeah. I was
00:08:28like, you know what, this is just a sign that I should move on to something else. This was really
00:08:32cool. I'm glad I got to do this. Got a lot of experience, but maybe I'll try to make an app
00:08:36in the app store next time. That was what I was thinking. I'm like, this is a lot of work to have
00:08:39an app outside the app store. So then that whole thought process lasted like a few months and then
00:08:43I got bored and I needed a new app to work on. And so yeah, Swift had also just been announced
00:08:48like the previous year. And so I was like, oh, you know what? Let me just make another toy app
00:08:52for myself to learn Swift just for me. And you know what? I'll make another emulator. It'll have
00:08:58more systems, but it's just for me. Like I'm, again, like it's just like a fun thing I wanted
00:09:01to work on to learn Swift. That was like my entire motivation. So you're just run the exact same
00:09:05playbook again without even realizing it. Exactly. I was like, you know, I'm just going to make more
00:09:10fun. Yeah, I know. You'd think I'd learn. That's okay. So, and that's where Delta starts. Yes.
00:09:16And so that's the beginning of Delta was like, okay, I'll call it Delta and just build it for
00:09:20fun. And that is what I thought I was going to be working on. And then I went to WWDC that year.
00:09:25It was my first time ever at WWDC. What year is this at this point? 2015. Okay. So I go to WWDC
00:09:31and I talked to the app review team because they have like a lab there. And I basically just say,
00:09:36hey, I'm working on this emulator app. Is there any world that it could be approved in the app
00:09:41store? And then they were like, actually, yeah, we'll allow that. You just got to like comply with
00:09:46some weird things. Basically they said I had to like submit a list of approved games to them
00:09:51whenever I submitted a version of Delta so that they could just make sure all the games that
00:09:54Delta supports were safe. Like they just didn't want to have like an open pipeline of like games
00:09:59and everything. And so like, okay, fine. That works for me. I'll submit a version of Delta to
00:10:04you and I'll have like a list of games that you can play. And then every app update, I'll expand
00:10:08it to include more games. And it seemed like that would be something I could do. And so I was really
00:10:13motivated. I'm like, cool, I'm going to get Delta in the app store. And so then I spent the next
00:10:17year, like actually building it for real. I was like, okay, this is now a real app. No longer just
00:10:22for me. I like really want it to be polished. I want it to be good. I took a lot of my time. I got
00:10:27on my college schedule to be working on it. And then I went to the WWDC next year and I talked to
00:10:33the same person and I said, Hey, I'm ready to submit. Can I submit it now? And he was just like,
00:10:40so I talked to some people, we can't allow emulators. And that was just all I got. I didn't
00:10:47get a why. It was just like, we can't allow that. I was just pissed off. Do you think this person
00:10:55was wrong the first time they told you it would happen? Or do you think something happened in that
00:10:59year and they changed their mind? Honestly, I don't know. Either one's possible. My guess is
00:11:04though that they thought it was okay. And then they talked to someone in the year and then
00:11:08maybe they had discussions about it. And then they came away basically, it's not worth the risk.
00:11:13Because I think that's what it's been this whole time. Apple's like, it's not worth the risk to
00:11:16allow these emulators. There's so much stuff that could open up. So I think they're just like, eh,
00:11:20it's one kid, whatever. We don't need to allow his app in the app store and it'll be fine.
00:11:24And at that point, I mean, I confess I was late to the iOS emulation universe, but
00:11:31it seems it's a weird thing. Like GBA for iOS was huge. And then it kind of went away.
00:11:36And so I would assume we're back to like lots of pent up demand. People are very excited about it.
00:11:40So it's not like you're the first person in history to ask Apple this question, I'm sure.
00:11:45But at that point, emulation was just kind of, it was sort of nowhere on the iPhone. So I guess
00:11:49for them to say, no, we're going to continue to not allow it is not totally out of character.
00:11:55Like, yeah, I wasn't surprised. I was just pissed off. Like, I was just like,
00:11:59I wasted a year of my life working on this. And because you told me I could release it,
00:12:03that was really just all I was feeling. Yeah, that's fair. That's fair. So what do you,
00:12:07at that moment, you're faced again with this same decision of like, okay, do I just
00:12:12call this a successful project? I learned how to use Swift. Let me just move on with my life.
00:12:16Or do I go the other way? I know the answer, historically speaking, but what is that?
00:12:20What do you what is that moment? Like? Yeah. And this time, I was just so pissed off. I was
00:12:25just like, you know, this is not okay. I am gonna get this out there somehow. It's just like,
00:12:30really what I came to. And I did at that time, I really didn't know what it would be.
00:12:33But I was like, I just wanted to show Apple that they can't treat developers kind of like this
00:12:37was really just like, I was like, you can't treat us like this. You can't tell us one thing and then
00:12:41change your mind the next year, because real people are spending real time doing this kind
00:12:44of thing. So I just motivated to just kind of prove them wrong. This is the best I could say.
00:12:50And so the I was like, I'll figure out a way. So then next project is figure out a way like,
00:12:55how do you how do you even start? I just got lucky. Honestly, Apple had also just around that time,
00:13:01made it so that you can install apps from Xcode for free with a free Apple ID that like they just
00:13:07changed that rule of right around now, so that students could start developing. And so I saw
00:13:12that I was like, wait, you can install apps for free. If you just have an Apple ID, there's
00:13:17something there. Like I was like, that's all the technical thing I needed to I could make something
00:13:21from that. And so that was just like, the beginning of all stores like, okay, I got to do something
00:13:25with this free Apple ID sideloading from Xcode thing that Apple has just announced.
00:13:32And then I was basically just like, took a few years to research that build it up,
00:13:37find out what would be the most convenient flow because yeah, there are a lot of obstacles with
00:13:41that it was apps installed this way only could last seven days, you could only have three apps
00:13:46installed at a time, this method. So it was just a kind of trying to come up with creative ways to
00:13:51make a lot of these restrictions, not as annoying, like the fact that alt servers, the thing was the
00:13:56number one that alt server just refresh your apps every week in the background. And when I thought
00:14:00of that, that's when I was like, okay, I think there's something here we can at least work with
00:14:05that's good enough. So when did the first version of alt store like work? What was sort of the first
00:14:12version of it that you had that you're like, okay, this does the thing that I need it to do?
00:14:16I think that was 2019 is when I finally like, had put together all the pieces into one flow,
00:14:23whereas in like, I had a program on my computer that could install alt store onto my phone,
00:14:27and then alt store could send apps to that device or to my computer and then reinstall it.
00:14:31And then it was early 2019. I was like, okay, this can work like just seeing it actually work
00:14:37in practice. It was totally janky, like the UI was horrible, but just like the fact I could press
00:14:42a button on my phone, and it would do everything, send it to alt server and install it was just
00:14:46honestly kind of magical. So explain to me why that worked in the sort of infrastructure of
00:14:51Apple's universe. Like why did that system work? Because so the way Apple allows students to
00:14:58test out their apps is they have to have a Mac and they have to be using Xcode and then to plug
00:15:03their phone into Xcode. And then they have to install an app from Xcode onto the phone. Like
00:15:08you can't just download a file and do it. It's all through Xcode and all through a computer.
00:15:13And so alt server is basically replicating everything Xcode does. You can have an app
00:15:19file, but iOS can't install an app file on itself. You have to send it to a computer and then alt
00:15:24server does the same stuff Xcode does behind the scenes to install it back onto the phone,
00:15:28as if you're a developer testing out yourself. Got it. Okay. And you were able to basically
00:15:32build a thing that lets anyone kind of mimic that system for themselves.
00:15:37Yes.
00:15:37When I set it up, it was, I wouldn't say it was a lot of work, but it wasn't no work.
00:15:42Yeah. It's a lot. It's a bit. Yeah. Especially on the Mac, there is the mail plugin for a long
00:15:47time. That was just really good. Yeah. It was confusing to everyone. And I was like,
00:15:52oh, this is too nerdy to try to explain even to people why it matters. But just, yeah.
00:15:57People got through it.
00:15:58Well, so, and this is, this is, I feel like the sort of ongoing story here is you're,
00:16:02you're finding these increasingly complicated ways to do this and continuing to discover that
00:16:07there is so much demand for something like Delta that people will jump through basically any hoops
00:16:12that you asked them to jump through in order to do this. Yeah. The biggest thing I was surprised
00:16:16about with alt stores, we need to get your Apple ID and password to like log in with your account.
00:16:22And I was like, wow, we really are just asking people to trust us and give us your Apple ID
00:16:26and password. And I mean, we're, we do everything we can with it. We literally don't touch it
00:16:29ourselves. We send it straight to Apple. But I was like, Ooh, that's a big, scary thing that
00:16:34you really just aren't supposed to give your password to random people. And then nope,
00:16:37people are still giving us a password. They don't care. I'm like, this is cool,
00:16:41but also it makes me nervous for security as a whole that people are just doing this
00:16:45without thinking about it. Oh, for sure. And it puts you in a really interesting position.
00:16:48I would think as the, as the developer of this to say, okay, what is this thing? Like, am I just
00:16:54building this as sort of a fun side project for people who want to play Pokemon on their phones,
00:16:59which comes with, I think one set of like security and privacy implications. Like to some extent,
00:17:04honestly, if you send your Apple ID and password to just a dude, I don't think you have any
00:17:10reasonable expectation of privacy. Like, I think it's a good thing that you were taking care of
00:17:14people's privacy as best you could. But I think for me as the user, I deserve what I'm getting
00:17:19at that point. Right. But then if you're, if you're like a company that is, that has a terms
00:17:23of service and is set up to make money. And you're like, we're in this for the long haul
00:17:26and we're serious about it. I think that changes both the way you have to treat stuff and the way
00:17:31that I can expect you to treat stuff. And it seems like as you were going through this,
00:17:34you were also trying to decide like, how professionalized do I want this thing to be?
00:17:40So that, is that fair? That's pretty fair. Because when the first version of Altstore
00:17:44that launched was literally just like a Delta installer, the only app in it was Delta,
00:17:49it could install Delta. And I just knew that was important. But even at that point,
00:17:53I knew I wanted Altstore to be a bigger platform. That was like kind of what motivated. It was like,
00:17:58I didn't want to build a whole nother workaround just to install Delta. The fact that I could do
00:18:02it for Altstore and for other apps too, made it more appealing to me. I was like, fine,
00:18:06I can do the work and then lots of other apps can take advantage of it. And so it was always the
00:18:10plan to expand beyond Delta. But I just didn't know really what that would look like until
00:18:16probably like a year or so after we launched. And I started, we started actually having some
00:18:19third party apps wanting to be on the store. And then I got a sense of, okay, this is the type of
00:18:24apps people wanted to install. This is what I got to be dealing with. And yeah, that's helped a lot.
00:18:29See, so what, what was that? Like, we're going to get into the kind of history moved
00:18:34towards you a little bit aspect of this in just a sec. But I am curious in that moment,
00:18:39the idea of like, I'm going to get my third party app store to be sort of, you know,
00:18:43legitimate in Apple's eyes was completely off the table. Like 2019, 2020, there was just no
00:18:48world in which that was going to happen. But you're like, you're talking about, you know,
00:18:51other apps that want to do this kind of stuff. What were you starting to see? What was kind of
00:18:55the, the like unifying thesis behind apps that wanted to be part of this thing you were doing?
00:19:01I mean, at first it was just any app that wasn't allowed in the app store. And honestly,
00:19:04it's still a big appeal day. It's just like, for whatever reason that was rejected for whatever
00:19:08type of rule, they just couldn't be the app store. And so then they'd see Altstore and then
00:19:12they'd be like, oh, an alternative. There are a lot of emulators, like obviously. So there's
00:19:16plenty of those. There's stuff you can expect, like iTorrent had like an app. And you can be
00:19:21like, yeah, I can see why Apple wouldn't approve it, but stuff like that. UTM, the virtual machine
00:19:25app, just running Windows on your iPad. Really cool. Not allowed in the app store. So it was
00:19:30just like a bunch of different random project or another one I liked that I think is really cool.
00:19:36Old OS, if you've heard of that one. It's like a high schooler just recreated iOS 4 in Swift UI.
00:19:43And so it's just like the whole experience, you open the app and it's a home screen from iOS 4.
00:19:47You can open all the apps and it looks exactly like it did back in the day.
00:19:51Oh, that's awesome.
00:19:51And it's really cool. And so it's just like, oh, a fun little idea that not allowed in the app
00:19:56store. So yeah, a bunch of stuff like that. So I launched Altstore and the first year was really
00:20:00great, but the first year was also a lot and it was just me working all the time. And then I
00:20:05honestly was like, what am I doing here? Am I going to keep with Altstore? Is this ever going
00:20:10to be a real thing? Should I just move on to something else now? I was really going, trying
00:20:15to figure out what I want to do. And then it was really at that point I realized what I wanted was
00:20:19I saw basically that the US was investigating Apple and that the EU was investigating Apple
00:20:24around this time. And I was like, you know what? I feel like there is going to be something that
00:20:28happens in the next few years. I want to be ready for that, but I need help. And that's when I
00:20:33basically brought on my partner Shane Gill, who is my best friend and I've lived with for like 10
00:20:38years. And I was like, hey, I need you with this with me. Just do it together. I need you to do
00:20:43all the business stuff. And so I can just focus on the programming and then we can make that happen.
00:20:47And at that point, we started making the plans for what Altstore is today, or at least an
00:20:53Altstore in the EU, Altstore Pal. Okay. So you actually really were betting on the theory that
00:20:58eventually this was going to be a real thing that you could do and not just sort of a series
00:21:03of ever more elaborate hacks. Yes. By that time, I was convinced it would happen just,
00:21:09I didn't know when, but just at some point in the future. That's fair. Okay. So, and this was like
00:21:14peak COVID, if I'm doing my timing math here correctly. Everything's weird at this time.
00:21:19And you're like, well, let's just go for it. Everything's weird. Everyone's re-evaluating
00:21:23their lives. My original plan was I was like, hey, Shane, let's move to New Zealand together.
00:21:27Let's just escape everything. Let's build this in New Zealand because I'm a New Zealand citizen.
00:21:31So it'd be cool to go home for me, but that didn't work out. It was too hard to get New
00:21:35Zealand during COVID, understandably. So we're in Dallas now, which is the same thing.
00:21:40It's this close. Okay. So you start building this thing and was the idea, we want to build
00:21:49sort of a full App Store. Like was that, is it as simple as that? That was the pitch?
00:21:53Yeah. When Shane came on board, that was the full pitch. Like we're going to build a real App Store
00:21:57and show people what life could be like if there are other apps, just like, just be the example.
00:22:02And then whenever it was official, be the official, like be an official store.
00:22:05And what was, what were developers saying? Were you like reaching out to folks being like,
00:22:08do you want to be part of this? And what were they telling you?
00:22:10Honestly, we didn't need to be reaching out to anyone. People were just reaching out to us.
00:22:14Like there are a lot of people who just have app ideas that just aren't getting in the App Store
00:22:18and just, and even I'm like, wow, there are just so many apps that like, we're getting
00:22:22like reaching out to us. I'm like, oh wow, that was rejected. That was rejected. So we just,
00:22:27people just reaching out for us.
00:22:28Yeah. And were you having any contact with Apple at this point? Other than, you know,
00:22:32those two conversations at WWDC, was there any inclination that Apple was even sort of aware of
00:22:36your existence throughout this process?
00:22:38We did at one point when the EU started investigating stuff, me and Shane sent an email
00:22:45to the executives of Apple basically saying, hey, y'all, I think the EU may be going too much.
00:22:50And do you like, we just reached out, hey, like we're here and we don't want like a crazy world
00:22:55of sideloading. Do you want to like be on the same page here? And they didn't respond at all.
00:23:00But we were at the time just thinking, we don't want a crazy free for all. And that's what it
00:23:04seemed like maybe the EU was going to be. We wanted a more restricted sideloading world where the App
00:23:09Store was still the main App Store. Why? I really think that's what makes iOS so good for like the
00:23:14vast majority of people. Like it is just so simple that you can get any app that you trust through
00:23:19one store. I think that's really valuable. It's why like so many people I know personally use an
00:23:24iPhone that aren't techies. And so I just always knew that there needed to be a way for sideloading
00:23:29to exist without taking over the App Store or taking like so many apps away from it.
00:23:33Okay. So you then are watching the regulation stuff happening. Are you like
00:23:38barreling through all the, you know, DMA white papers and trying to get the deep sense of what's
00:23:43going on? Like how into this were you as that regulation was happening?
00:23:46Once the DMA was like an actual thing and they were talking about, oh yeah, we were very much
00:23:50deep in it. And then we were just like, we got to be ready for anything. Because literally until
00:23:54Apple announced what they were doing, we just were like, are we going to be a sideloading tool? Are
00:23:59we going to be a bit of a full App Store? Are we going to exist at all? Like we just had no idea
00:24:03what was going to come next. We just knew that there would be some way of installing apps.
00:24:07Okay. And so I guess with that idea, you can kind of keep building Delta and AltStore the way you
00:24:14had been without making too many kind of unchangeable decisions at that point.
00:24:20Yes. We basically were working on stuff that we knew would be true no matter what. We were
00:24:26building up the UI for browsing apps and we were expanding or building the Patreon flow so that
00:24:31people could connect their Patreons and install that. Because we just knew no matter what, that
00:24:36would be good for this world. We just didn't go into the really technical stuff about the actual
00:24:41installing process. It was just basically making the app look as much like a store as it could be.
00:24:46Yeah. That makes sense. And at that point, how are you thinking about the business of Delta?
00:24:50You've been working on this thing for an awfully long time without making any money on it.
00:24:53I'm assuming there were conversations about how do we all get rich from this thing that
00:24:58clearly lots of people want. People do keep asking me. But the thing that I've believed in,
00:25:04and Shane, before he even joined AltStore, he was the one that convinced me to make a Patreon for
00:25:08all this in the first place. He was like, hey, when you're working on Delta and AltStore, just
00:25:12have a Patreon on the side to support yourself. Because honestly, I thought my original plan was
00:25:16to release AltStore and Delta and then move on to something else and just let it exist and let
00:25:20people install it. But Shane was the one telling me, no, people want to support creators right now.
00:25:26You should just have a Patreon that people want to subscribe to and just support future development.
00:25:30And so, I launched AltStore with that, and it did really well. It has paid for everything ever
00:25:37since I started AltStore. We really believe that's a really important thing that should exist.
00:25:42Because developers right now, you can't sell apps through Patreon in the App Store,
00:25:46and it's just such a really good way, I think, for creators to build a relationship with their
00:25:51audience. People also don't mind spending a few bucks to a person, but they do mind spending a
00:25:58few bucks on an app. So, it's just a very different relationship. And so, we're basically just trying
00:26:03to really promote the Patreon idea. Our own income basically is just through the Patreon.
00:26:08We're still selling or having access to our betas for our patrons,
00:26:11and that's worked out pretty well as a way of monetizing Delta, honestly.
00:26:15Okay. Yeah, I mean, even just the idea of thinking of an app developer as a creator in the same way
00:26:20that you think about a content creator as a creator is really interesting and is totally
00:26:26not how most people perceive it, right? Exactly.
00:26:29And I think it's coming around a little bit. Personally, at least, I've seen a lot more
00:26:34of the sort of bootstrapped one-person app become a thing that people like in a way that I didn't,
00:26:40and they associate the app with the person who makes it. But I think for most people, it's all
00:26:44so faceless, right? It's just an app that exists. I don't know if it's made by one person or 10,000,
00:26:50and I don't care. I'm just mad that they're charging me money. And it's interesting to
00:26:54think that maybe the business model is part of the problem there, that because it all just sits
00:26:58in the app store, and it is so divorced from personality and people, that maybe that's the
00:27:03problem. So, actually, connecting Patreon to that is a really interesting way of just changing the
00:27:07way people think about what an app is and how it gets made. Yeah, exactly. And so, yeah, I'm really
00:27:12excited for that. Another nice benefit is we can just have our apps be free, and then people just
00:27:19get the betas. So, it means there's no features that are locked behind paywalls forever.
00:27:24It's just that you can wait forever and get it for free, or if you just want early access,
00:27:28you can donate. And I think that's more appealing to people when they're trying to support an app
00:27:34versus, oh, you've locked away a feature, and now I can never use it unless I pay.
00:27:38So, it's like, yeah, that difference, I think, makes a huge deal.
00:27:41Yeah, no, I totally agree. So, okay, so right before the DMA drops, and I guess, what was it,
00:27:48March of this year when Apple put out its third-party App Store plans?
00:27:53It was end of January, yeah.
00:27:54Okay, so this year, Apple, right before that, what was the status of Delta and
00:28:01AltStore? It sounds like it's growing, people are using it. How were things?
00:28:07Things were good. We had, at that point, 4 million users on AltStore, just existing. I was like,
00:28:13oh, wow, this is going great. Things are going great with Delta. We had been prioritizing,
00:28:17basically, a new version of Delta, 1.5, to release alongside whatever Apple announced.
00:28:24And so, we're just working on that. But then, basically, by January, we're just like, okay,
00:28:30everything's in a good spot. We're ready to go. We just need to know what the hell is happening.
00:28:35And so, we're just basically just waiting around to see like, okay,
00:28:39we can do this. The store had been finished by then. All the UI was done,
00:28:44Delta was ready, Clip was ready.
00:28:46So, you have this thing, and you're like, this exists, we made it, it's good,
00:28:51we have absolutely no idea what we're about to do with it.
00:28:53Exactly. That is exactly where we were.
00:28:55What a strange place to be.
00:28:57It was very strange. It was, yeah, because every day, we're like, what do we do?
00:29:00We can't move forward on a lot of things. So, it was just a lot of planning and strategizing.
00:29:06Yeah, it makes sense. So then, end of January, Apple drops like a million press releases and
00:29:11technical documents about how all this is going to work. What happens to you on that day? What do
00:29:15you do?
00:29:15Well, that was an overwhelming day, for sure. And the first thing that we try to do is figure out,
00:29:22okay, can we exist at all in this new world? Because yeah, there's all these restrictions,
00:29:27and we're trying to figure out what the hell an alternative marketplace is. And so, yes,
00:29:32we spend that whole day basically trying just to parse what Apple's announced.
00:29:36And then, we basically are like, okay, we think we can do this. We think we can meet the criteria
00:29:42if we get very creative. Because yeah, the criteria was like a lot of things. Like,
00:29:46you needed to have like the standby letter. We need to have a subsidiary in the EU.
00:29:51Right.
00:29:52And just like a bunch of things that we just sat down like, okay,
00:29:56we got to make all this happen in a month, basically. Because we're like,
00:29:59okay, we have until March 7th was like the DMA compliance day. We're like, okay,
00:30:04let's set up a subsidiary. Let's get the standby figured out. Let's build everything. Because
00:30:08also, they announced all the stuff we had to build. And it's like, okay, let's build everything
00:30:11we need to do. And then, for the next month was like extreme crunch time, just like getting
00:30:17everything done. Shane, honestly, was incredible. He was calling so many countries to figure out
00:30:23where we could incorporate. We finally incorporated in Ireland. So, we have a nice subsidiary in
00:30:27Ireland, which is great. He was on talking to banks to get someone to believe in two people
00:30:34who are trying to do something crazy. And they need this.
00:30:37How hard was that process? You're like, so we build a game emulator and also
00:30:41other app stores. I'm imagining a bunch of banks are just like, this is a prank call and hang up.
00:30:45A lot did. So many people just wouldn't listen to us. It was pure luck that we found one person.
00:30:52His name's Logan. He is just like Morgan Stanley. And he just understood what we were doing and
00:30:58really wanted to go out of his way to help us. And I honestly think if we had not met Logan,
00:31:03I don't know if we would have gotten this all figured out. But he was just like, okay,
00:31:07let's do this. I believe in y'all. I can work with my bosses. We can make it happen.
00:31:12That was a big win, getting that one for sure.
00:31:14Yeah. I mean, I remember when the news came out thinking that there were, like you described,
00:31:18a bunch of technical challenges, but there were always going to be a bunch of technical
00:31:21challenges, right? I feel like that you sort of know going in, there's going to be a bunch of
00:31:25weird work you have to do to get this thing to comply with what Apple wants. But the business
00:31:30challenges of just the hoops you had to jump through as an organization seemed like they
00:31:35were designed to keep almost everybody out, except for the truest of true believers and
00:31:42maybe just the most stubborn people building this thing.
00:31:46Yes, exactly.
00:31:47And part of me wonders if going back a few months, Apple just deeply underestimated
00:31:51your willingness to pick this fight. After so many years of picking this fight, you're like,
00:31:56I've been through your nonsense before, Apple.
00:31:58Oh, yeah. I know they were caught off guard. They told us that. When we ended up going to
00:32:03the lab and everything, they just said straight up, oh, we did not expect anyone to be this
00:32:08prepared, especially as a two-person team. They just did not expect us to be as prepared as we
00:32:13were, which I think...
00:32:13That had to feel good.
00:32:15Oh, it did feel good. I was like, cool. But at the same time, I was like,
00:32:18well, then let us launch. It was like, cool. Okay, that's amazing. Let us launch. We're ready.
00:32:22Was there any question for a team of two whether to invest all this time on something that,
00:32:28at least for now and potentially forever, is only going to exist in the EU?
00:32:32Like, there's a lot of people in the EU, but there's a lot of people not in the EU also.
00:32:36Yeah. I mean, so two parts. One, I do think it's going to come outside the EU at some point. I
00:32:41think it's just a question of time. So yeah, I expect to see within a few years, it'll expand
00:32:46out to like, I don't know, Japan, United States eventually. So we viewed this as like the beginning
00:32:51of a much longer process for us to expand. But then the other thing was like, we had nothing to
00:32:57lose. This is literally all we're doing. All our income is working on Altstora Delta. We're like,
00:33:02well, we might as well just go all the way with it. Like, what else? Like, why not?
00:33:06Yeah, I mean, I think the long bet, I think, is probably a safe one. Again,
00:33:10it's a question of how long it takes, but I do think you're probably right.
00:33:14It also seems like you were in a position where even if it didn't sort of immediately reach a
00:33:20ton of people, you were going to get so much notice just because this is such a big thing
00:33:23and such a big change that like, all of a sudden, Altstora just appeared in so many headlines
00:33:28in like February and March of this year, which is like...
00:33:31Yes, that was very, very surreal seeing that all of a sudden. Yeah.
00:33:35Yeah. I mean, it's so funny because like explaining Altstora to people, it's like,
00:33:38okay, it is this bizarre sort of convoluted experience where you have to set up a server.
00:33:43And I'm seeing this in like mainstream news articles.
00:33:46I know. I'm like, oh my God.
00:33:47It's like, oh my God, if my mom asked me what Altstora is, I'm not going to know how to answer
00:33:50that question for her.
00:33:52Same. Yeah, I explained to my friends like, eh, don't worry. It's just a tool to install apps.
00:33:58That's it.
00:33:59So, okay. So then we're almost to the present time, but I feel like we still have not gotten
00:34:03to the single wildest part of this whole timeline.
00:34:05Yeah, that's inaccurate.
00:34:07So you're building the store, things are going well. You launch, I think on launch day, right?
00:34:12March 7th, was it ready to go?
00:34:14No, we were ready. We were all ready to go. And that was the beginning of like this frustration.
00:34:21So we were, yeah, we were invited to a lab in Cork for the Apple to like help us implement
00:34:27our backend and everything. So like a week long lab for us just to work with Apple engineers to
00:34:31get everything working. So then we flew out to Cork. It was like the week before March 5th.
00:34:36So the end of it, the end of February, we flew out day one. I showed up. I said, okay,
00:34:40here's everything. We're ready to go. Can we launch? And then they were like, oh, hold on.
00:34:47They expected you to show up and be like, I'd like to build an app store. And instead you show
00:34:50up and you're like, I have built an app store.
00:34:52They literally, I really think they expected everyone to show up with like no code base,
00:34:55like to just start from scratch and how to do this. And yeah, so we show up and we're like,
00:34:58here's the entire product. It has store pages, it has everything you could possibly need,
00:35:02downloading works. And then they were just like, oh, wow. Yeah. Just very clearly. Yeah. They were
00:35:07not expecting that. And so it basically was then, okay. So Monday I was like trying to pressure
00:35:11them. Can we launch tomorrow? Can we launch tomorrow? iOS 17.4 is coming out tomorrow.
00:35:15Can we be there? And then it just became clear that this whole process was going to be much
00:35:21longer than we expected for a lot of foreign reasons. Like basically just legal, like
00:35:26giving them documents, having them review documents, stuff like that. So I was 70 before
00:35:32came out and me and Shane were very bummed because we're like, we were ready to launch
00:35:37on launch day. We were all ready to go. And then, uh, so that was frustrating.
00:35:42So how long did it take before you actually got the thing up and running?
00:35:46Then I think it was two months later. Like it was, yeah, it was April 17th. Like, yeah. So
00:35:52from March 5th to April 17th was just waiting. So a month and a half. And so we did that. And then,
00:35:59yeah, the reception was immediately just phenomenal. People just so excited. And,
00:36:04but then we also immediately saw, which we expected that Delta was the main story.
00:36:08Sure.
00:36:09That Delta in the app store was the big deal, but we had to figure out what to do. And we
00:36:12knew that would be the outcome. If we released Delta in the app store that Delta would get like
00:36:16it's big moments to be everywhere. And we just were like, you know, that's fine. It's free
00:36:20marketing for AltStore. If everyone just knows about Delta and then they talk about it and then
00:36:23they get it through AltStore in the EU.
00:36:25So you just, you just glossed past part of it though. Like the, the whole
00:36:29Apple suddenly allows emulators in the app store.
00:36:32Yeah.
00:36:33You say that as if you expected that to happen. I don't think anybody expected that to happen.
00:36:37Not at all. It, the most, yeah. Unexpected thing that could have happened. And me and Shane had
00:36:42different reactions at first. Shane was more like, this changes everything. Now we got to think
00:36:48through it. And my just reaction was, wow, what validation that Apple had to change the rules to
00:36:54allow emulators because they were threatened by us about to launch with an emulator. Like that to me
00:36:59was like invigorating. I was like, this is, yes. It means we're doing something right. That Apple
00:37:04is literally like doing something I never thought they'd do because it was the only way they could
00:37:08try to make the story. Not the EU now has the best apps outside the EU has, doesn't have the
00:37:13coolest apps. And I just knew Apple, I guess, yeah, they just couldn't have that story, which
00:37:17is also why emulators are allowed worldwide is my guess is the story just could not have been in
00:37:21the EU has better apps than the rest of the world. Yeah, no, I think that's exactly right. And it is
00:37:26very hard to argue that this wasn't specifically about Delta. Like I just can't the timing it's
00:37:32too, it is all about Delta. We haven't had a direct confirmation from Apple, but our app Delta
00:37:39was being notarized for like five weeks. They changed this rule. It's approved the next day.
00:37:44Oh, wow. It's almost like they were holding it to figure out what they could do about it.
00:37:48Did you get any heads up that they were going to make this change?
00:37:51Nope. Nothing at all.
00:37:53Well, and I would guess after that happened, you would have expected it to be big. It sounds like
00:37:58you did expect it to be big, but I feel like it's been, it's certainly been bigger than I expected.
00:38:03Like it's, it's bigger than I expected. It's been nuts, man. It's so validating. Like I really can't
00:38:09like, cause yeah, we're doing this because we just really believe that people just want to play
00:38:14old Nintendo games or old, all games. We just, I think it's just a thing people want to do.
00:38:19And for so long, everyone just says, no, that's only for the nerds. It's only for like people
00:38:24who know what they're doing. And I've just been like, you just have to make it accessible. That's
00:38:28all it means. Like all you got to do, if it's accessible, people will know how to do it and
00:38:33they will love it. But we're seeing that that is exactly the case. And yes. So even more people
00:38:36than I thought are enjoying it. It is the most surreal thing seeing it not only in the app store,
00:38:41but being the top app in the app store for like a long time. Are you still the number one app in the
00:38:47app store? I mean, I think last time I checked, but yeah, since we've launched, we've been number
00:38:50one and yeah, so surreal. Like the most, like anytime in my life, I've never felt more like
00:38:58I actually am dreaming right now. Like I need like pinching myself, like, cause it was just
00:39:02too perfect. Like work on this app for 10 years and then day one in the app store, it's the top.
00:39:08It was just incredible. Has it changed what you think about how to do all of this? I mean,
00:39:13if you had charged $2.99 for Delta, you would have made a crap ton of money by now. Like,
00:39:18have you, are you thinking about the business of all of this differently after it's blown up the
00:39:22way that it has? I think we want this to be bigger than just like Delta. And we want this to cause
00:39:31actual change in like the emulation scene or in the gaming sphere. Just like, we want to make this
00:39:37more accessible for everyone. And we want people to start talking about it and to have real
00:39:40conversations about it without just being like piracy, piracy. And so a lot of what we're doing
00:39:44is we thought Delta to make that change happen, we had to like, it had to be free, had to be,
00:39:50everyone had to have it. And that's, I think just what we're focused on really is like,
00:39:53as long as we're making, we're making enough to support ourselves. And so we're not in need of
00:39:57more money, but I just think for Delta to have the biggest impact is just make it as fully
00:40:02free and accessible to as many people as possible and let them just start reliving games. Are you
00:40:07worried at all about legal ramifications? I mean, there was obviously the Yuzu thing that happened
00:40:12kind of right before this. It's been a weird moment in the emulation community. Have you
00:40:18heard from any Nintendo lawyers? Yeah, that came out as we're also like in Europe trying to launch
00:40:23interesting, very, very relevant to us, but I'm not going to say I'm not nervous about things,
00:40:31but I am confident in what we're doing. I've learned a lot of what not to do over the past
00:40:3610 years. And I think that we are really trying to show how you can do something like this legal,
00:40:41like we're not Yuzu, we're not emulating a current generation console. We're not doing a lot of
00:40:46things that other emulators get in trouble. Like there's no DRM we have to deal with to like for
00:40:50the game. So just the game files ripped from cartridges. So we don't have to do anything
00:40:55like that. We're doing everything we can to do it right. And there was no world that I wouldn't
00:40:59have done this. I guess it's a better way. Like I wouldn't have got this far and not released
00:41:03Delta. I just had to. No, I think that will. And I think that's part of why this moment of
00:41:07your story is so interesting because we're at the beginning of this weird new era of apps in the app
00:41:15store. And I think a lot of stuff is changing and a lot is going to continue to change. But we're
00:41:19also in a funny way, kind of at the end of your 10 year journey, like the, this you, you finally
00:41:25finished the thing that is what it feels like. It's really weird, but I'm like, wow, it actually
00:41:31happened. Like it got in the app store. And yes, it does feel like the culmination of that whole
00:41:37journey, which is very weird. Does that make it feel like I finally got to do it and it worked
00:41:43and it was huge and it was awesome. And now I can go do something else. I honestly, the fact that
00:41:48Delta's doing so well, it's just motivated me to just want to work on it so much for right now,
00:41:52because I've just been working on it in isolation, assuming one day it could like be a big deal that
00:41:58I don't know, actually seeing everyone play it right now has made me really motivated to like
00:42:03really focus on it again. And also because so much of the past two years has been like a priority
00:42:07working on AltStore. And now that we launched AltStore, we can just like have some time with
00:42:11Delta again. Like there's just so many fun features I want to add, but I haven't had time to, because
00:42:15we had to get AltStore up and running for this deadline. But so I'm looking forward to that. But
00:42:19I don't know if this is a 10 year thing. I don't know if this is the beginning of like another 10
00:42:23year adventure. To me, I think it's I'll work on Delta for a bit. And I think I got to finish up
00:42:28what AltStore is like. We still have some to do to prove it out. Like we got to get the third party
00:42:33apps on there. We got to build up the platform. And I think that's, to me, the next step is just
00:42:39making sure that AltStore is a real app store that people are using and happy with. And then maybe
00:42:43I'll be done. And then then you will have resoundingly defeated Apple, which very few
00:42:48people can say. That's what I want. Yeah. All right. We got to take a break. And then I have
00:42:53to tell you about this weird obsession I've developed in recent months. We'll be right back.
00:43:04Welcome back. I've spent most of the last month and really most of the last year talking about
00:43:09and testing AI gadgets. It's been a long time, maybe since the early VR days, but maybe even
00:43:15as far back as the early smartphone era, since we've seen this many new things with this many
00:43:20new ideas. Most of those ideas right now are very bad. Well, not bad necessarily. It's just that
00:43:26they don't work. So they're bad. You know what I mean? The humane AI pin has some interesting
00:43:31theories about how you might use a device without a screen and without getting sucked in and
00:43:35distracted. It's just that none of those things work very well. The Rabbit R1 just kind of puts
00:43:40an AI chatbot on a device that looks like a rearranged phone. And again, none of it works
00:43:45very well. So far for me, at least, there is only one thing about AI that really feels like it works
00:43:52and actually makes my life better. And that is voice notes. A good microphone, plus good
00:43:56transcription software, plus some useful models for summarizing and grabbing the most important
00:44:01bits of what I say. It works. And it's awesome. I kind of feel like a doctor now, running around
00:44:06talking nonsense into my tape recorder, like JD from Season 4 of Scrubs.
00:44:11Mrs. McCalla, 40s, moderately attractive, condition improving.
00:44:15Did you just say I was moderately attractive?
00:44:17Excuse me, Mrs. McCalla?
00:44:19Patient's complaint of hearing loss was clearly just to get attention.
00:44:23Though I also, to be fair, have this experience a lot, doing that.
00:44:27Thought, I like toast. They aren't all winners.
00:44:31In actual reality, by the way, doctors have used voice recorders forever.
00:44:35Talking is a lot faster than typing, and a transcript of a conversation can often give
00:44:39you a lot more information quickly and with context than you'd get from just a written thing.
00:44:45And a lot of doctors are now using AI to help summarize those transcripts and help them find
00:44:49key information later. Because the problem with recording everything on a voice recorder
00:44:53is that it's a lot of information, and it takes a lot of time to parse through it all.
00:44:58Listening to your own voice for hours at a time just to find that one thing you said?
00:45:02Not fun.
00:45:02The big thing that happened to make all of this stuff work recently
00:45:06is that OpenAI built a speech-to-text tool called Whisper.
00:45:09When OpenAI launched it in 2022, it said that the model, quote,
00:45:13"...approaches human-level robustness and accuracy on English speech recognition."
00:45:17In its blog post announcing it, it used this audio as an example.
00:45:21This is the Micro Machine Man presenting the most midget,
00:45:23miniature motorcade of micromachines. Each one has dramatic details,
00:45:26That's a guy named John Moschita Jr. doing a commercial, by the way.
00:45:29He's actually in the Guinness Book of World Records as the world's fastest talker.
00:45:33I've now listened to that ad approximately a million times,
00:45:36and I do think Whisper gets one word wrong. It says
00:45:40R near the end, and I think Moschita says and.
00:45:44But I can't even tell for sure. And that's all Whisper missed. That's how good this is.
00:45:48OpenAI reportedly developed Whisper as a way to transcribe millions of hours of YouTube videos
00:45:54in order to get training data for its GPT models,
00:45:56which is a strange and somewhat nefarious way to have gotten to this point.
00:46:00But let's just leave that aside.
00:46:02Now we have this open-source Whisper model that works
00:46:04incredibly well and is being used all over the voice notes world.
00:46:08The first gadget I got to try all this out is called the Plaud Note.
00:46:12P-L-A-U-D.
00:46:13It's a $160 recorder about the size of a credit card, just a little thicker and a little heavier.
00:46:18It's just a microphone, basically,
00:46:20that connects to an app on your phone to store and process all of your recordings.
00:46:24You just press the one button on the Plaud to start recording,
00:46:27and a little red light turns on to let you know it's working.
00:46:29It has one mode for just recording your voice,
00:46:32and another that's supposedly optimized for recording phone calls.
00:46:35It actually comes with a case you can attach to the back of your phone for recording those calls.
00:46:39And it's surprisingly good at it.
00:46:42Note that this call may be recorded for quality assurance.
00:46:46If you have an emergency, please press one now.
00:46:48If you would like to fill a prescription, press two.
00:46:51As a reporter who has spent most of my career
00:46:54constantly trying to find ways to record phone calls, this is awesome.
00:46:58For years, I've used this weird rig where I have a headphone splitter
00:47:01that goes into a pair of earpods and also into a voice recorder.
00:47:04This is just way better.
00:47:06But you can also use it to record sales calls and meetings and that sort of thing.
00:47:10Plaud definitely imagines the Note as a business tool, first and foremost.
00:47:14I have been using it mostly for things like to-do lists,
00:47:17or when I wake up in the middle of the night with a fun idea and want to get it down before I forget.
00:47:21And I love, love using it while I wander around my kitchen
00:47:25to see what needs to go on the grocery list.
00:47:27Okay, this is my grocery list for the week of May 1st.
00:47:32I need bagels, potatoes, almost out of bread flour, so I should get bread flour.
00:47:38I need more paprika, but the smoked paprika, not the other stuff.
00:47:44So I record that, and then a few seconds later, I open up the Plaud app,
00:47:47which is very bare bones, but works basically fine.
00:47:50And that recording shows up.
00:47:51I open it up, and I get three options.
00:47:53Transcription, Summary, and Mind Map.
00:47:55Mind Map is useless, let's not worry about Mind Map.
00:47:57I tap on Transcription, and it asks what language the recording is in
00:48:00and what kind of note it is.
00:48:02Call, meeting, interview, task assignment, that kind of thing.
00:48:05I pick task assignment in English, and a few minutes later,
00:48:08I get a full and very accurate transcription of my grocery list.
00:48:12That's kind of useful, but really what I want is for this to just make me a grocery list.
00:48:17So I tap on Summary.
00:48:18It tells me at the top,
00:48:20the meeting outlined the necessary grocery items to be purchased for the upcoming week,
00:48:24ensuring a sufficient supply of food and essentials.
00:48:26Which is true, I suppose, but totally unnecessary.
00:48:29And then it gives me the list.
00:48:31All is a single action item.
00:48:33This is not what I was hoping for.
00:48:35But all the stuff is here.
00:48:36And the coolest part is, when I said I was good on peanut butter and good on coffee,
00:48:40it actually kept those things off of the list.
00:48:44Which means it actually in some way understood the difference between me saying something
00:48:48and me deliberately adding it to the list.
00:48:51That's the stuff, right?
00:48:53The Plog Note is very cool, but it's expensive for just a voice recorder.
00:48:57And you do also have to pay 80 bucks a year for the AI features.
00:49:01That is a lot for a grocery list tool.
00:49:04So instead, I turned to the Rabbit R1,
00:49:06a more full-featured AI device that in general is pretty bad.
00:49:10It can't do a lot of what it promises,
00:49:12but it can record, transcribe, and summarize audio.
00:49:15So I started carrying this thing around just to see how it would do with to-do lists
00:49:19and journal entries and all that stuff as well.
00:49:21Yeah.
00:49:21This is me testing out the voice recorder.
00:49:24Let's see what I have to do today.
00:49:27I need to finish my Apple story.
00:49:31I sent edits back to Jake.
00:49:32He's going to get edits back to me.
00:49:33I'd like to get that at least in the top edit today.
00:49:36We have the Verticast at 2.
00:49:38Once you record something, Rabbit stores it all in an app called the Rabbit Hole,
00:49:42which is not a good app in general, but for these purposes, it's fine.
00:49:45It pulled out a very basic plain text list of stuff from my note
00:49:50in addition to just giving me access to the recording itself.
00:49:52And it got most of the stuff on that to-do list that I said,
00:49:56but it did just completely ignore a couple of things.
00:49:58And since Rabbit doesn't just make a transcript and let you see it,
00:50:01I don't actually like this one quite as much.
00:50:04At least it is easy to download the audio and upload it somewhere else,
00:50:07which is something.
00:50:08And the mic is pretty good.
00:50:09I need to write the container post for tomorrow,
00:50:11but I can do that tonight while Anna's at class.
00:50:13That's my day.
00:50:14But in general, buying a whole device just to record audio
00:50:18is probably only for the most dedicated users of this stuff.
00:50:23The main advantage of a dedicated voice recorder
00:50:25is that the battery tends to last a lot longer.
00:50:27They tend to have better mics.
00:50:28And also because they're not an app,
00:50:30you don't run the risk of weirdness with background apps shutting down
00:50:34or getting a phone call in the middle that screws everything up
00:50:37or just loud pinging notifications constantly popping up in your recording.
00:50:41Not everyone needs a voice recorder all the time,
00:50:44but if you do, you'll probably appreciate it being a dedicated device.
00:50:48But if you want to get most of the functionality of those devices,
00:50:51but you'd rather just use your phone,
00:50:53oh boy, do I have good news for you.
00:50:55There are suddenly a million apps out there
00:50:57that are making use of Whisper and other models,
00:50:59especially from OpenAI,
00:51:01to transcribe, summarize, and make use of all of your audio.
00:51:05You can basically Google and find a thousand options,
00:51:07but let me just tell you about two.
00:51:08The first is called Cleft Notes,
00:51:10and it works on Mac and iOS.
00:51:12This one I think is the best voice notes app I've tried, period.
00:51:16You open it up and hit record,
00:51:17and it shows this amorphous meshy blob that moves as you record,
00:51:21which is kind of delightful to look at.
00:51:23Like most other apps,
00:51:24it gives you both a full transcription of your audio
00:51:27and tries to make some sense of it after you finished.
00:51:29But Cleft goes further, actually.
00:51:31It tries to format and structure everything you said
00:51:33so that if you come out of a call
00:51:35and you want to write down some takeaways...
00:51:37Okay, I just got off the call
00:51:38where we were planning Steve's bachelor party.
00:51:41It sounds like the plan is to go to Myrtle Beach for three...
00:51:45It then turns that into a document
00:51:47with bullet points and numbers and headings.
00:51:49It's not always right,
00:51:50but it's usually actually a pretty good start.
00:51:53Like that voice note I just played
00:51:54turned into a Cleft Note with an itinerary,
00:51:57golf plans, a transportation section,
00:51:59a food and drink section,
00:52:00and a note about renting golf clubs.
00:52:02It's written strangely formally,
00:52:05but it is really helpful.
00:52:06There are, like I said, tons of apps that do this,
00:52:09and since they basically all use OpenAI's models,
00:52:12they're all roughly equivalent
00:52:13in terms of the actual quality of the transcript and information.
00:52:17But I just want to quickly mention one other app I really like.
00:52:20It's called AudioPen,
00:52:21and it's a web app, which means it works everywhere,
00:52:23and it's super easy and fast.
00:52:25You just record some audio...
00:52:27My grocery list for today...
00:52:29I need eggs, I need nut butter,
00:52:31I'm good on coffee and good on tea,
00:52:34I need red wine for this weekend,
00:52:36I need coffee filters,
00:52:37I need three balls of mozzarella for pizza...
00:52:39Once you do, it gives you the same summaries and information as the rest.
00:52:43But what I really like about AudioPen in particular
00:52:45is that if you pay the $99 a year for AudioPen Prime,
00:52:48which gets you more transcriptions and storage and whatnot,
00:52:51you can also tell it how to format what it creates.
00:52:54So if you want everything in bullets or paragraphs
00:52:57or like iambic pentameter, you can do that.
00:53:00Here, for instance, is my grocery list in iambic pentameter,
00:53:03which I had a different AI from Eleven Labs read out loud.
00:53:07Eggs and nut butter I require.
00:53:09On coffee, tea, I'm satisfied.
00:53:12Red wine for weekend.
00:53:13Filters too.
00:53:15Three mozzarella balls will do.
00:53:17Pepperoni for pizza's taste.
00:53:19More bread flour, not to waste.
00:53:21Broccoli, mandarins combine,
00:53:23and peaches sweet.
00:53:25These fruits divine.
00:53:27You probably don't want your groceries in iambic pentameter.
00:53:30More likely, you'll want to tell it to be formal for business stuff
00:53:33or write like an author you like or something.
00:53:36But it works pretty well, whatever you pick.
00:53:38And it even goes another step toward turning my rambling thoughts
00:53:42into something that feels finished and useful
00:53:44and not just like rambling thoughts.
00:53:47Another thing you can do in AudioPen is upload existing audio.
00:53:50So like if you have a meeting you already recorded
00:53:53and now need to get something from
00:53:54or you want notes on a YouTube video you liked,
00:53:57you can just dump the audio in and get info out.
00:54:00I've heard a lot of people say recently that summarizing YouTube videos
00:54:04is one of their core uses for ChatGPT and other AI models.
00:54:08And this is another way to get the same kind of thing.
00:54:10You could put this podcast in
00:54:11and it would probably give you some information.
00:54:14Basically, all of these tools do cost money
00:54:16if you want to do stuff like that,
00:54:18at least if you want to use them heavily.
00:54:20Whisper is an open source model that anyone can use.
00:54:22You can actually download it to your computer
00:54:24and run it locally if you want to.
00:54:26But it is really computationally intensive.
00:54:28And for all the summarization stuff,
00:54:30these apps still are mostly using OpenAI APIs anyway.
00:54:35Oh, one exception I should mention, actually.
00:54:37If you have a Google Pixel phone,
00:54:38the built-in Pixel recorder is a terrific voice recorder.
00:54:42And there's a new Summarize button in there that also works really well.
00:54:45Apple's Voice Memos app doesn't do any of the transcription
00:54:48or summarization stuff,
00:54:50but I wouldn't be shocked if that's coming soon.
00:54:52In the last few months, I've been using AI audio for lots of things.
00:54:55Meeting notes, to-do lists,
00:54:57searchable transcripts of YouTube videos that I want to refer back to.
00:55:00But I actually like it best as a journal.
00:55:02I've never been good at keeping a journal over the years
00:55:04because it always involves taking the time
00:55:06and sitting down to do it.
00:55:07And it's an activity and it's a whole thing.
00:55:10But now I can just sit down in the morning
00:55:12and quickly brain dump everything I'm thinking about
00:55:15while I brush my teeth.
00:55:16Today is Wednesday, May 1st.
00:55:18The first thing I have to do is finish up the Apple story
00:55:23and my Rabbit review.
00:55:26I also have to book train tickets to get to New York for tomorrow.
00:55:30Should be a relatively chill day.
00:55:32And then at the end of the day,
00:55:33while I'm doing dishes or getting ready for bed
00:55:35or just walking around,
00:55:36I can just ramble about what happened that day for a few minutes.
00:55:40I was in Chicago today for the Chicago Humanities Festival.
00:55:44It was a really fun day.
00:55:45We did a good job of, I think,
00:55:47not having the whole thing be a bummer,
00:55:49but also having it not be too sort of Pollyanna-ish optimistic.
00:55:54When I do that, I now get, instead of just an audio file,
00:55:58a really good transcript that I can find again later
00:56:00and a nice summary of the key points of everything I talk about.
00:56:03It's both incredibly low effort and surprisingly useful.
00:56:08Honestly, I hope day one, which is the journaling app
00:56:10I've been trying to force myself to use for like a decade now,
00:56:13adopts some of this stuff.
00:56:14This feels like a good way to get things into a journal.
00:56:18Honestly, I think these apps and devices
00:56:20are kind of perfectly indicative of how AI works right now in general.
00:56:24For all the hype over image generation
00:56:26and AI's abilities to write emails for you
00:56:29or, you know, replace your job entirely,
00:56:31the thing these models do best
00:56:33is take a bunch of information and try to simplify and summarize it.
00:56:37That's handy if you're trying to get the gist of a long legal document
00:56:40or you want to know what a podcast was about
00:56:42before you devote 90 minutes of your life to it.
00:56:44That's why I'm so intrigued by Notion AI and all the other tools
00:56:47that aren't trying to understand all of the universe's information.
00:56:50They're just focused on the stuff that you care about
00:56:52and making sense of that.
00:56:54That is a really good use case of AI, if you ask me.
00:56:57And it's working pretty well for me
00:56:59because, as you guessed, I do love rambling into microphones.
00:57:03Actually, you know what? I should stop doing that for a minute.
00:57:05We're going to take a quick break
00:57:06and then we're going to come back
00:57:07and take a question from the Birchcast hotline.
00:57:09We'll be right back.
00:57:12All right, we're back.
00:57:19Let's get to the hotline.
00:57:20As always, the number is 866-VIRG11
00:57:23and the email is virgcast at thevirg.com.
00:57:25We love your questions.
00:57:26Please keep sending them, keep calling,
00:57:28and we try to answer at least one on the show every week.
00:57:31This week, we have a very timely question from Dax.
00:57:35Hey, guys.
00:57:35It's Dax from Virginia and I can't go to sleep
00:57:39because I've just been thinking,
00:57:41why is the Rabbit not just an app?
00:57:44At $200, it's like a very buggy, bad-looking plastic thing
00:57:49and it doesn't do anything that I feel like my phone can't do.
00:57:53So why isn't it just a paid subscription service app
00:57:56that I can install on my phone?
00:57:58We'd love to know your thoughts. Bye.
00:58:00I've been thinking about this a lot.
00:58:01I have the Rabbit R1 in front of me.
00:58:03I have used it a ton.
00:58:04I just reviewed it.
00:58:06It's on the site.
00:58:07We have a video.
00:58:08You can watch me wandering Washington, D.C.,
00:58:10yelling questions at the Rabbit R1.
00:58:12And I've been thinking about this question a lot,
00:58:14ever since really using the humane AI pen.
00:58:17What are these devices actually for?
00:58:19Why are they devices?
00:58:20And I think I have a cynical answer
00:58:22and I have a non-cynical answer.
00:58:23The non-cynical answer is that I think there is room
00:58:27for something that isn't a smartphone.
00:58:29In a funny way, we're actually going through the beginnings
00:58:32of an unbundling process in a lot of ways.
00:58:36I think there is a set of people who are realizing
00:58:39that they want multiple devices for multiple things.
00:58:42The idea that I should just have a smartphone
00:58:44that does everything was never really true.
00:58:46I mean, think about that brief moment
00:58:49where we tried to use our smartphones as TV remotes.
00:58:52It turns out that having buttons you can press
00:58:54without having to look down at is a good idea.
00:58:57Something like the Ray-Ban MetaSmart glasses,
00:59:00I think is really interesting,
00:59:01or the Snap Spectacles that they've been working on
00:59:03for years.
00:59:04The idea that actually a camera that's not in my pocket
00:59:08but is on my face is actually a good idea.
00:59:11So I think there are some things
00:59:12that should have been shoved into the phone, right?
00:59:15Like my phone as a music player
00:59:17and not necessarily needing a dedicated music player
00:59:20makes a lot of sense.
00:59:21But I think there is room for some kind of AI device
00:59:25that isn't your phone.
00:59:27The idea that it could be a little more accessible,
00:59:29it could be closer to your face and mouth and eyes
00:59:32and the sort of biometric stuff
00:59:34that actually can make AI really useful.
00:59:36My guess is that's probably something
00:59:38that looks more like a smartwatch or a pair of glasses
00:59:41than it is a sort of dedicated handheld device.
00:59:44But the reason I was so intrigued by Humane
00:59:45is I think the idea of a wearable
00:59:47that's like on your lapel,
00:59:49kind of always within arm's reach is really interesting.
00:59:52That's also how I felt years ago
00:59:54about things like Alexa and Google Assistant
00:59:57and the idea that I can talk to technology
00:59:59without having to pull my phone out,
01:00:01look at a screen, unlock a thing.
01:00:03This idea of ambient computing
01:00:04that's been floating around
01:00:05actually requires lots of different kinds of devices.
01:00:09I think the smartphone is going to keep being the main one
01:00:11for a long time.
01:00:12And I think any product that tries to subvert that
01:00:15and say, ditch your phone and use this other device
01:00:18is just gonna fail
01:00:19because A, people don't wanna ditch their phone
01:00:21and B, you shouldn't.
01:00:21Phones are great and they're useful
01:00:23and they do lots of things.
01:00:24But we can attach stuff to phones
01:00:26that does more things in more ways.
01:00:28And I think that's gonna be very cool.
01:00:30The AI gadget remains to be seen,
01:00:33but there are ways it could work
01:00:34and there are ways it could happen
01:00:35and there are ways that it might be exciting.
01:00:38That's the non-cynical answer
01:00:39is I think there is something here.
01:00:40The cynical answer is that
01:00:42they're just desperately trying
01:00:45to find a way to sell you something.
01:00:47We have gotten into a place in the smartphone market
01:00:50where there are two companies, Apple and Google
01:00:52that are total gatekeepers for the entire industry.
01:00:56If you wanna do something on the iPhone,
01:00:57you have to play nice in the app store.
01:00:59You have to play by Apple's rules.
01:01:00You have to give Apple a cut.
01:01:02That is just how it works.
01:01:03Same in the Play Store.
01:01:04There's like a little more openness in Android,
01:01:06but for all intents and purposes,
01:01:08if you wanna be on Android,
01:01:09you have to play by Google's rules.
01:01:11That means there are certain kinds of things
01:01:12you just can't do.
01:01:14That means there are certain kinds of apps
01:01:17that you can't make a real business out of
01:01:19because of the way that Apple's rules work.
01:01:21And so what you're left with is saying,
01:01:22okay, there are two companies
01:01:24that are just going to tax us
01:01:26every time we wanna do anything.
01:01:29And I think we've kind of spent a decade
01:01:32watching companies that didn't win in mobile
01:01:35try to win in the next thing.
01:01:37And I think in a huge way,
01:01:39that was why then Facebook, now Meta
01:01:42bought Oculus all those years ago,
01:01:44because it was saying,
01:01:45okay, we think that there is going to be
01:01:47a huge new VR platform
01:01:50and we cannot lose it again to Apple or Google
01:01:53because there's so much money in being Apple and Google.
01:01:56You get to do all kinds of things.
01:01:57You get to insert yourself in the middle.
01:01:59You get to take the tax.
01:02:00You get to become a huge ad business.
01:02:01You set the defaults.
01:02:02It's just a massively powerful thing to be.
01:02:04And so when there was the idea that,
01:02:06okay, VR is the next thing,
01:02:08you saw a bunch of companies chase VR.
01:02:10When everybody thought AR
01:02:12was gonna become the thing,
01:02:12that was where Snap showed up and was like,
01:02:14we're a camera company
01:02:15and we're gonna build spectacles
01:02:16and we're gonna take over.
01:02:18And the HoloLens did it and magically did it.
01:02:20And there was this crazy rush
01:02:22to not just be the first great device,
01:02:24but be the first great platform.
01:02:26The same thing is happening right now in AI.
01:02:28As I talked to both Humane and Rabbit
01:02:30and frankly a lot of other companies in this space,
01:02:33the idea is to start with hardware, right?
01:02:35You build a product,
01:02:36you make people buy the product,
01:02:37you make people like the product
01:02:38and then you build a platform out of it.
01:02:40Humane's big idea is this thing called Cosmos.
01:02:42That's their operating system.
01:02:44They're hoping basically
01:02:45to have an app store for Cosmos.
01:02:46You'll be able to use the APIs to plug in.
01:02:49It'll work on lots of different devices.
01:02:50Like they are very much trying to build
01:02:53a software platform with the size
01:02:56and power of iOS or Android.
01:02:58Rabbit has been a little less ambitious
01:03:01about how it talks about that.
01:03:02But when it talks about the large action model,
01:03:04this thing that can actually go
01:03:06and use the apps you use for you,
01:03:08that's the same thing, right?
01:03:10Like there's a version of this that is kind of hacky
01:03:12that is you teach the system how to do it,
01:03:13but there's a much easier version
01:03:15that is just partnering with companies that exist
01:03:18and have existing products.
01:03:20And for Rabbit, if it can be that center point,
01:03:23the sort of place that you go to use the apps,
01:03:26that's very powerful.
01:03:27OpenAI is trying to do the same thing
01:03:28with the GPT store and custom GPTs.
01:03:31Google, I suspect is going to start
01:03:33to push into that pretty hard.
01:03:34I think you're going to start to see Apple
01:03:36have AI APIs that you can plug into.
01:03:39The idea is just if you can be the sort of hub
01:03:43for all the different features of AI,
01:03:46that is going to be hugely powerful.
01:03:48We've seen in this first phase,
01:03:49a bunch of companies kind of try to do everything, right?
01:03:51You build the one huge, perfect model,
01:03:54that solves all your problems.
01:03:55Everybody uses it for everything, the end.
01:03:57I don't think that's the future.
01:03:59I think the future is actually
01:04:00sort of an interconnected galaxy of models
01:04:04and systems and tools that you use.
01:04:06And so the question then is who sits in the middle?
01:04:08What is the platform that you use
01:04:10that translates all of that stuff for you
01:04:12and picks the right agent depending on what you need?
01:04:15And I suspect all of these companies think
01:04:17that if they can be that platform,
01:04:19they can make an awful lot of money
01:04:21off of being that platform.
01:04:22Hardware is part of the business.
01:04:24I think actually a lot of these companies
01:04:25genuinely would like to make money from their hardware.
01:04:29Rabbit is hard to tell.
01:04:30And frankly, so is Humane
01:04:32if there's actually any money in it right now.
01:04:34But they do seem to both want to be hardware companies,
01:04:36but more than that, they want to be platforms.
01:04:39And you cannot build a platform on top of iOS or Android.
01:04:42They just won.
01:04:43And as they push into AI,
01:04:45they're going to entrench that even further, right?
01:04:48I think we're going to see a lot of AI stuff at WWDC.
01:04:51We're going to see a lot of AI stuff at Google IO.
01:04:54Gemini is already all over Android.
01:04:56Those companies are not interested
01:04:57in letting other AI platforms in.
01:04:59And I think there's really interesting regulatory questions
01:05:02about whether they're going to be forced to.
01:05:03Like what if you picked your AI model
01:05:06the same way you picked your default browser?
01:05:09Really interesting idea.
01:05:10These companies will not do that
01:05:12unless they are forced to do that.
01:05:14So we'll see.
01:05:15But the way that these companies are thinking
01:05:17is that the only way in is to build a new thing
01:05:21that doesn't require a smartphone
01:05:23and start to slowly siphon you away
01:05:26from these two companies, Google and Apple,
01:05:28that have essentially owned our internet experience
01:05:31and have been between any company
01:05:33that wants to build something and you for 15 years.
01:05:37That's a hard, hard, hard thing to upset.
01:05:39It's why a lot of the antitrust stuff going on right now,
01:05:43especially in the US, is happening.
01:05:45And if I'm a new company,
01:05:47that's the only reason I would build hardware
01:05:49because building an app is too hard.
01:05:51There's too much stuff in the way.
01:05:53And you're probably just going to get eaten alive
01:05:55by Apple or Google.
01:05:56So that's the cynical answer.
01:05:58I hope for the non-cynical answer.
01:06:00And in reality, I think it's a little bit of both.
01:06:02Like we are in a phase
01:06:04where there are interesting new ideas about hardware
01:06:06because there are interesting new ideas
01:06:07about how we interact with technology.
01:06:09We're going to spend, I think,
01:06:10less time tapping on our phones over time
01:06:13and less time looking at screens just to do basic things.
01:06:18So if that's the case, we are going to need new devices.
01:06:21But also anyone trying to compete in those spaces
01:06:25knows that the only real way to build a company
01:06:28the size of Google or Apple
01:06:29is to do it outside of Google or Apple.
01:06:32And that's what everybody is desperately
01:06:33trying to figure out.
01:06:34That said, all the AI stuff we've seen so far
01:06:37should pretty much just be an app.
01:06:38I have yet to see the one that I am like,
01:06:40this is a new category of device.
01:06:42But I'm holding out hope.
01:06:43I haven't tried all the glasses yet.
01:06:45I'm excited to see.
01:06:47All right.
01:06:48That is it for the Verge cast today.
01:06:49Thank you to everybody who came on the show.
01:06:51And thank you as always for listening.
01:06:52There's lots more on everything we talked about.
01:06:54My rabbit review, lots of our Delta coverage,
01:06:56all kinds of stuff on AI voice notes,
01:06:58lots more on theverge.com.
01:07:00I'll put some links in the show notes.
01:07:02But as always, read the website.
01:07:03It's a good website.
01:07:04We try hard to make it great.
01:07:05If you've thoughts, questions, feelings,
01:07:07or other AI gadgets you want me to try out,
01:07:10you can always email us at vergecasts at theverge.com
01:07:12or just call the hotline 866-VERGE11.
01:07:15We love hearing from you.
01:07:16Send us all your thoughts and questions and ideas
01:07:18for what we should do on the show.
01:07:20Feelings about everything.
01:07:21It's my favorite thing.
01:07:22We have a Slack room, by the way,
01:07:24where every single one of the hotline questions
01:07:26comes in as they are received.
01:07:27And I get to listen to them.
01:07:29And it's just so fun.
01:07:30I listen to every single one.
01:07:31I promise.
01:07:32Whether or not it makes it on the show,
01:07:33please know our whole team hears all of your calls.
01:07:36We read all of your emails.
01:07:38We love hearing from you.
01:07:39This show is produced by Andrew Marino,
01:07:40Liam James, and Will Poore.
01:07:42The Verge Cast is a Verge production
01:07:43and part of the Vox Media Podcast Network.
01:07:45Nilay, Alex, and I will be back on Friday
01:07:47to talk about earnings, iPads, and a whole bunch more.
01:07:49We'll see you then.
01:07:50Rock and roll.