Tech Brew Ride Home - (Bonus) John Gruber Talks Apple And AI

Episode Date: January 13, 2024

John Gruber joins Chris and I to talk about AI Hardware, Apple's AI strategy and Apple's Vision Pro launch. Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to another weekend bonus episode of the TechMeme Right Home. I'm Brian McCullough. This has been a long time coming for two reasons.

Number one, it's been about six months since Chris Messina's been on for one of these properly. We used to do them every other week for about a year and a half. But Chris is back finally. Hey, Chris. Hey, howdy. Welcome to 2024. Yeah. Let's get into what you've been talking about and arguing about.

eventually. But also, we have an excellent guest who, shockingly, has never been on the show. He needs no introduction, but John Gruber, welcome to the TechMeme Right Home podcast. I don't know about the needing no introduction, but I'm very happy to be here. There's not a single person listening to this that doesn't know who you are. So, before we do anything else, I thought what we should do is, bring up what Chris and I have been arguing about a little bit offline, and not arguing, but having a little disagreement about. But, you know, CES was this week, and as you heard on the show, I highlighted the Rabbit R1, that little AI thingy that was sort of like

a smartphone, but not. And then I did on the show on Thursday. that all of a sudden they sold out in 24 hours. I'm going to let Chris go first because I said on threads sort of facetiously that this was the best hardware launch of the last five years that I've seen. But Chris, give me your impressions of the Rabbit R1 and why you're a little skeptical of it. And then I want to hear John's take on any of these sort of smartphone replacements. Yeah, look, I mean, it's funny because I was seeing the rabbit ads on X, ironically enough, even though I'm not there anymore, days before the launch.

And so I was excited to actually post it to product, you know, as one does. And I believe that you sent me a link to the keynote once it was out. And you were like, you must watch this. And so instantly, I dropped everything and realized that now I had like 25 minutes that I had to blow on this long keynote. note. And so I got to admit that I sort of scrubbed through it. And the beginning I thought was pretty strong. I was like, okay, this is like an interesting entrant. You know, I watched the humane AI pin launch, had a certain thought for like, you know, which I would argue was a worse sort of intro video, but we can come to that. Well, it was at least a little bit shorter. However,

look, there's so many different ways to unpack this. What I would say is that there is seemingly an opening the marketplace to define what an AI type interface is going to look like that goes beyond sort of a black sheet of glass. And mostly its voice, but also it needs to have eyeballs so it can see things and, you know, read the refrigerator and tell you what to make. You know, it's got some way to like talk to you or whatever. So it's getting those capabilities. So I loved the visual industrial design of the rabbit. Teenage engineering did a great job on that, as they do with most of their products. These guys are kind of a generational talent when it comes to that.

But once I saw the founder, unfortunately, I don't know his name. Jesse Liu? Okay. L.YU. Liu. I don't know where he came from. He sort of knew on the scene, just sort of blew out of nowhere. And he started to go down the kind of demos.

Unfortunately, and look, I am a parable, like a terrible critic because now that I'm an investor in the AI space, I see a lot of stuff super early. And basically what he was showing was like the greatest hits of the last six months of stuff from the agentive AI space where there's things like LangChain and Langsmith and other types of essentially kind of piping one type of LLM into the next with the ability to then take actions through internet APIs. And it isn't that I disagree with the, I suppose, the orientation to the opportunity of problem space. It's just that this is exactly what we were describing six years ago when we were working

on bots and conversational AI. So does it need a physical form factor? Does it need to have its own lamb, this large action model? And furthermore, if it's not replacing the phone, which you said it's not, whereas the AI pin is supposed to, like, how do you relate to this as a product? Maybe I just don't understand like a multi-device person in the future. You know, you've got the watch, You've got the phone. You've got your rabbit. You've got a pin. You've got AirPods.

You've got, you know, goggles. Like, we're basically going to be strapped to the max, you know, with all these things that don't talk to each other. I said that on the show. I don't get it. And, John, go ahead and go in a second. But like I said, well, this is reintroducing a second device into your pocket because it's not like you're going to get rid of your smartphone. But what did you think of the rabbit, R1? It's very funny.

So as soon as I saw it, and I wasn't really paying attention to day one of CES or whenever it was announced, but it was like somehow one of the slacks I'm on a group chat, somebody posted, I saw it. And my first thought was, whoa, that thing is a rip-off of Panics Play Date, the little yellow game player. And within like 10 seconds found out that, oh, Rabbit had partnered with teenage engineering to do the industrial design. Well, that's who did the industrial design of Panics Play-Date. So of course they look like siblings. And I thought, oh, well, if this company is smart enough to partner with teenage engineering and I'm as effusive about their talents as Chris's, I was like, $199, screw it. I'm just buying it. And I literally pre-ordered before I watched the keynote. Because it's like, $200 is like just under my threshold or not just under. It's a beautiful device. So even if you just like have it lying around. Right. Also, we can get into that.

Like, that's also something that humane maybe needs to think about. But, you know what, John, what I've seen a lot of people say is, like, when they see the article or whatever, and then people send them the keynote, as opposed to Chris, a lot of people are like, oh, once I see the keynote, I'm convinced. Like, what did you think of the keynote? I love the keynote. I did too. I like that the CEO, Jesse Liu, I know I can't enunciate Chinese very well, but I'm glad he did it. And yes, he's not super polished. He's not a pitchman.

But there's a certain genuine... But there was an authenticity. Offensive. That's what made it better. And they did not overhype it. Like Humane had like two years of saying, this is going to change everything. And it's hard not to compare them directly against them,

rabbit against Humane, because they're obviously directly competing devices. And humane, every single way that you can contrast them, there's a difference. Humane had this years-long hype cycle beforehand. They have the Apple pedigree. The Apple pedigree that they've emphasized in that hype cycle. They've raised boatloads of money.

To the point where I know from my sources that they're the leaning on the Apple pedigree is, these are not popular people inside Apple Park. And there's a lot of, it's not just that people who leave Apple are expected not to talk about having been at Apple, but that the credit for Apple devices is Apple. And to the degree that they don't even put engineers names in the about boxes of the software anymore. And that's a thing that's not like that they've spilled secrets about Apple that they came out, but that the taking of credit for doing stuff is so contrary to the Apple.

Apple culture, that it, that more than anything. It's not like, oh, Apple is worried about this company or something like that. It's just, it's personal. But anyway, they had this hype cycle. They have a $700 price point. And again, I don't know what my threshold is for an impulse buy, but I, these two devices prove that that number is between $207. Yes.

It is. Because I did not, I have not preordered a humane AI pin. Right. And it's, you know, and I did, I did this without looking at the keynote. And then I did watch the keynote. And I had like this sort of sine wave up and down where I'm like, oh, yeah, I'm glad I pre-ordered. And then like at the three-minute mark, I'm like, no. He definitely needed editing.

But at the same time, they also at Humay needed editing. So they didn't seem like a weird cultish sort of, you know, guru sort of thing. But ultimately, I got through the end of the keynote. And I'm glad I pre-ordered. And then it so happens that I think within the last 24 hours of us talking, they've announced up there, they've cut off pre-sales and said that this is a sellout. They're not shipping. It's a very unusual to me way of putting a ship date.

They're shipping at, quote, Easter. And I don't even know. I grew up Catholic. I don't practice anymore. It's the most unusual holiday in the Western world because it changes every year. I don't know what that means. I mean, I get it, you know. But they did also.

They probably don't have a huge team to figure out the messaging of what they're going to do. Go ahead. But the other thing, the other big, I mean, we could keep going about the fundamental differences between the way that, even though they're same sort of device. But the big one that Rabbit did, that Humane didn't, is Humane went all in on no screen. We're not going to have a screen. We're not going to have a screen. And I can't help. Except for the one on your hand.

Right. I can't help but think that they got so caught up at some point in the development cycle. We don't want to have a screen on our thing that we're making. And then they got to these edge cases where it's like, but we have to show something. We have to show text in some way. And then they came up with this project a laser thing onto your hand and then have the camera see when you tap virtual buttons or whatever. And it's like, no, what Rabbit did is the more obvious answer. Just put a goddamn screen on the thing.

Well, okay. This will lead us into discussing what Apple. could be doing with AI going forward, which is the thing that was the most interesting to me was the sort of obviating the need for interacting with apps in the sense that, yes, you have a screen, but it's not a screen where you're doing Uber and tapping that and doing this and tapping that. It's literally talking to it. Now, we can, and, you know, Chris's product guru. You're a Mac and Apple sort of design. Yeah, aficionado.

So you guys can tell me how this could evolve, but the thing to me that kind of blew me away more than the device, which was sexy as hell, don't get me wrong, was the idea that, yes, that's it, that's it, that the agent goes between the apps and does the thing for you. So whoever wants to take this, we're assuming that WWDC this year will be about a lot of AI integration into iOS and whatever OS. What do either of you think about what that means for, I'm not tapping an app and going into an app and doing a thing.

I'm just picking up my phone and doing a thing, which is what sort of Rabbit is suggesting, Chris is squinting his eyes, is suggesting, the future might be. Whoever wants to go, go. I'll just jump on that. Just because, like I said, I'm leaning a bit on my sort of legacy experience here where the dream of conversational software was alive and well, basically in 2016, 2017. I worked at Uber where we had an API and you could book a ride via voice and that was all there. I think the real thing that we learned, to John's point about needing a screen is that unless you live in a world where there are infinite dollars and infinite undues and customer service is abundant and amazing.

Essentially for any time a robot screws up, you get your money back and there's no problem, people need to be able to confirm things along the way to make sure that the right thing happens. There was a whole demo that the CEO gave about booking a trip to some random place with a bunch of highly nebulous things. I mean, if I gave that to, you know, like my partner Joe and she was, you know, my partner Joe, she went off to like book the thing, I'd probably have some critique of like her choices, not that I don't think she has great choices or taste, but just because I want to make sure that this actually makes sense. So to get to a point where you can ask a computer to do

these things for you without human nuance or without a sense for kind of like the vibe that you want, even if you can kind of modify the request or the prompt to be like actually make it a little bit more like Lucy Goosey, the thought that you don't want to have agency in that process made it highly dubious that this product is going to work the way that humans actually prefer to work. And so, you know, the $200 price point is amazing. 10,000 people are going to get this thing. They're going to ask it to book a trip. It's going to have what, your credit card on file. It's going to execute all these things. Suddenly people are going to have tens of thousands of dollars of credit card transactions on their account that they didn't anticipate because they didn't

expect it to buy the $5,000 flight because it's available. And now you've got to undo all that shit. And so hopefully there's a robot that will do that for you and we'll, you know, kind of like talk down the prices. So that agentive web is something that's been available. It's just at the risk of moving that direction as quickly as this thing suggests. I think we're just not ready. I will just say that I, and I've had an Alexa in the kitchen for, I don't know how many years. How many flights has it booked for you? I don't even buy stuff from Amazon on it. And that's from their own store. So they're not like, the whole use case was like, buy me toilet paper. And you're like, it's going to

to buy the wrong stuff. And I don't know, it was just a couple of weeks ago, and Alexa churped up in the kitchen about something. It was like, it was like, I forget what color. It's like an orange light instead of blue. And I was like, what is that? And Alexa wanted to tell me that one of our recurring purchases, I don't know, it's like my wife has like trash bags. Subscribe and save or something. Yeah, subscribe and save and we get trash bags or paper towels or something like that on a regular basis. And Alexa was like, like, If you'd like to order more, just say, I don't might know. I do not want to order anymore. Stop. Okay. But there is, but that's just from Alexa. I don't trust it for that. And so I do think.

Years of experience. I mean, it's a huge platform. But that demo was way, was to me over the, over the line. But this is Chris's point. Chris's point is that the demo looked amazing, but he doesn't believe that the actual product will deliver what the demo is. So let me step back, right? I want to make my point clear because I don't want to just like, you know,

shittle over some tech demo that's like, you know, cool and interesting. I think there's a level of thought that needs to go into this next era of AI tech that is about service design and is about thinking about the moments where, you know, human agency needs to be brought into the mix, where computers can do a whole lot of work, you know, come up with like five different flight plans and, you know, present it to me as a pitch.com like deck, show me, like, you know, sell me on it as a travel agent might, but without any pressure and let me think about it. Like, that would be an amazing use of this kind of like conversational AI, such that I can make a better decision, more informed decision, and I have to do less work. And then I look like a hero to my family, friends, office, whatever.

But instead of delegating all access and authority and agency to the agents, that to me is a recipe for disaster. And that's where I feel like both the AI pin and this rabbit device haven't quite gone deep enough into how people really want to be able to relate to these things in a way that feels like they have some control and confidence. I do think though, and again, the proof will be in the pudding when it ships, but it seems like if there's, I don't know about a breakthrough, I don't know if it's unique, but it does seem like what they're talking about, what they're calling the LAM, the action. The large action model. Yeah, right. So that's, yeah. Right, which is like, so the language models are the ones that whether you're typing or whether it's voice, you know, it's human using words and the, AI giving words back. It doesn't really matter whether it's text or text to speech. It's the language.

The action model is what they're claiming is that their AI is able to parse website user interfaces, right? And so you don't have to go through an API to do these things. I keep using the verb drive. And let me, the way I think about this is, and I think this is undeniable, is where we're going eventually is to have like C3POs or R2D2s in our house. Like actual robots that are with us and that are you can talk to in a reasonable way and that can do things for you. Like, whether it would actually be, I mean, you know, an actual humanoid or some other shape, but an actual robot that can like go up and down your stairs, move around your house. And if you could say to C3PO, cook, you know, how about you make us a chicken dish for dinner? Or look in the fridge and tell me, what can you make us for dinner?

And C3PO could, what would C3PO do? It wouldn't, like the robot that can cook you dinner isn't going to pop the dinner out of its belly, you know, like a gag on the Jetsons or something like that. It's going to actually, it's going to go in your kitchen and look in your fridge and use your stove and your oven and your pots and pans. By that point, it probably is ordering the things on your behalf and making the, well, but I'm just saying, but I hear what you're saying. But you know what I mean? It's, it's not going to be a robot that is a robot that is. a refrigerator and a microwave and a thing. It's going to use this. And baby steps along the way, it seems like Rabbit's vision of this is sort of doing that with websites. Right. And it's,

it's just software, but it's driving the Uber website or the DoorDash website or whatever, the way that a physical robot, in the not too distant future, I'm convinced, will like use the things in your house, you know, and go actually open the door for you to take a pack. package from FedEx. The not too distant future next Sunday, AD, Realheads will know that reference. But let me bring this to the iPhone and the iOS and things like that. So, again, what you're describing is that the robot in your house doesn't, it just does the things that interacts with the things.

On an OS level, on a smartphone, the thing that struck me about this was the vision that what if eventually the way OSs go is that there's you're just interacting with a better Siri. And so when you're booking a car, you're booking the groceries, you're not going into the app. So like right now, I go into my smartphone and I, my experience, it's like, you know how the car companies are all obsessed with like, we don't want car play. We don't want Google's, whatever it's called, because we want to own the experience in the car, right? But in your phone right now, you still have the experience with seamless, with Spotify or whatever. But is the future of the OS is what made...

This is what I was thinking about whether or not Rabbit works. Is it... Does it aviate apps? Does it obviate icons? where it's just you're interacting with the bot, in which case, that makes sense to me, and it probably makes sense to Apple and everybody else. But what does that do to everybody that's not Apple or Google or whoever makes the hardware? So, look, again, the thesis that we're going to a post-app world is compelling. And I'm just going to repeat myself, it is what we were working on in the early

bought era because the thought was, and I literally had this in like a keynote that I gave, it's like once you've got two million apps in the app store, there becomes a question as to how a person is expected to navigate that space where you have a squircle chicklet that represents some functionality that has some obtuse name that you remember somehow. And then you launch that app, that little chicklet by tapping on it and by creating an account and then by doing the things that it wants you to do by going through a series of onboarding steps until finally you get into some loop or flow where you come back and use that app on a regular basis. And the reality is that years ago, most people download apps and then they never even

launch them for the very first time. So there is definitely like a gap and a challenge with the existing app ecosystem. I think what Rabbit did was perhaps one of the first examples where they're like talking about, here's what a post app world will do. And the way in which we get to post app is not actually post app. We will just have a computer that knows how to read a screen like, a human does and it'll push the buttons that a human would push because frankly, it's the easiest way to get to the API on the other side rather than writing a bunch of bespoke code that integrates

with everyone's bespoke API. So the whole thing that really happened with apps is that we moved from a world before where you had maybe backend systems that would integrate together through enterprise software deals to one where you needed an API. You as a business needed to open up an API to connect to an app that was in the app store that would then connect to your customers. That was the appification of the broad commercial space. Now that those APIs exist and now that most companies have opened up APIs to the outside world, now we will have AI that actually works to sew together a number of integrations between these things in an opportunistic sort of way.

In other words, Brian, to your point, the challenge is that maybe you use Tavola or HelloFresh And I use a blue apron or sunbasket. And we have a robot that lives at home and it needs to be able to order food on our behalf. Each of those platforms have a different way of taking in our preferences or our food allergies or the frequency in which we want meals delivered. And so they need to be able to go to each one of those different websites or each of those different apps and make sense of them. Because frankly, it's a bunch of different choices just based on a brand. And so I do think that we'll be outsourcing to AIs to do that. And we thought that we were going to do that with voice assistants like Alexa and the Google assistant.

And instead, we didn't have LOMs. LOM is the game changer that then allows us to have late bindings, which, in other words, is another way to say a computer can come up to an interface that it's never seen before. And because it has billions of examples, apply the, what are those things called where you put the coin in the top? And then it sort of goes, you know, Plico, Plinko, Plinko, yeah, Pachinko, whatever the thing is. Exactly. More or less, kind of what an LM is doing to get to the right action on the other end, and then it executes for you.

So we've been here before, and it's just now about to start happening, but apps will stick around for some period of time, just like newspapers have, as a means to get to the thing that we want to actually achieve on the other end. And so Rabbit is the first, I think, to say it out loud in a way that allows us for people to get excited about it. But it's because we have enabling technology that just arrived last year, that suddenly now – and so the last point I'll make on this. And John, I think you'll have some insights into this.

Apple has one of the best accessibility frameworks for building applications out there. And that, what I mean by that, is that whether you need voice assistance or you need touch assistance or you know, you have low vision, all of those things that exist in the Mac and iOS ecosystem allow computers to see the software as well as humans that lack those certain capabilities. those are the ways in which a system like a lamb will be able to interact better than anywhere else. A couple of great points there. I'm largely, largely, largely aligned. And I think part of what Chris was getting at there is that technology, we tend to think of revolutions as blowing away what came before it, but that's not really what happens. Technology evolves in a sedimentary way, and the old layers, stay there, right? And nothing exemplifies that better than the way that Mac OS still comes with a

terminal app and a complete command line that is more or less instantly familiar to a time-traveling Unix user from 19704. Can't get rid of the command line, yeah. Right. Like, you know, somebody from Bell Labs in 1974 would be blown away by everything else in MacOS. But if you sat them down in front of the terminal app, they'd be able to find their way around. It's still there. And I think apps are like that. And building AI to drive the app, and the accessibility thing, I actually just brought that up on my other podcast, Dithering with Ben Thompson tonight talking about the Rabbit R1. It has to be. It is a real, real advantage that's overlooked by casual users who don't need

the accessibility features because they're not. heart of hearing or they don't have vision problems or something like that. But they really do benefit everybody, even if you don't have the... Well, and especially machines. Right. And machine, what are machines except, you know, handicapped users, right? They have a handicap, which is that they're not human beings trying to use these things that are designed for human beings. But like, and, and, you know, to go back to my kitchen analogy and cooking, like a human being who knows how to cook can go into a kitchen for the first time and they never saw this brand of stovetop before. They'll figure it out, right?

Like, I've never used an induction oven. But you'll figure it out. You see there's a dial, there's temperature. AI excels at figuring things like that once it has the base idea of you've trained it on 500 existing ovens, now put it in front of a new one. It's going to figure it out. The difference, though, and this is where the pressure is all on Apple, is with the privacy model of iOS, where your app can't poke around another company's app, it's all on Apple's shoulders. So if Apple can do this and whether they keep the Siri brand name, which I think they'll probably do because it's so well established.

A future Siri that's truly a 2.0, a brand new underpinning based on LLM and modern technology, could drive your iPhone by using the accessibility APIs. And your apps are still there. And it's the Uber app where you've used the app and tap the little squircle and signed into your account and it has your credit card. And you could still use the Uber app the way you've always done it by using the app. But now the system can drive the app through the. these verbal commands and it would be using the same thing. And if you've ever seen like a completely lined user use an iPhone, it's astonishing.

It's absolutely. Yeah, it's so fast. And it'll bring it brings tears to your eyes to see that these people who, you know, otherwise, you know, literally can't see anything can navigate the iPhone and in ways that they, you know, because there are other senses like touch are involved in a way that they couldn't even do. It's one of the things that iOS devices can do better than classic Mac or Windows computers, because the touch gives them this extra sense that they're more attuned to. So something like that, you know, except with Rabbit, they don't have that ability.

So a Rabbit app on an iPhone or Android can't poke around the Uber. app or the DoorDash app because of the privacy restrictions. And it, you know, that makes sense. But that's not the way things were a generation ago with the PC. Like with the PC, a breakthrough company like this could make a software product that poked around the system and poked at the buttons in these other apps. And it was in sandboxing like there is now. Right. So without the sandboxing, like so they could do it on the Mac, but the Mac doesn't make any sense because nobody's going to carry around. So the other thing that I would add to that, there was like,

shortcuts seems to becoming more and more important. Now, I don't know if that's sort of like a hobbyist area of MacOS, but in terms of automation and the ability to string together, you know, what looks like kind of, you know, Langsmith, Langchain-style agentive models where you have a set of commands. That would be the way that maybe a third party could somehow do this. But you'd never get the permission to always be listening. You'd never get the access to the hardware to be able to take control of the camera

So I see it. So I'm super happy that they built their own hardware so that they can write software that can make its own rules for when it's listening and when it's using the camera. They have to. Okay. So let me ask something real quick, though, because my question that is about where this kind of goes, and Brian, this will create an opening for you for your next topic. Sure.

Yeah, sure. But what I want to sort of, I guess, parse out a little bit is to think about the future of these two companies and whether or not they've boxed. themselves into a corner or whether or not they've actually created an opportunity for successive generations of products. So when the humane pin came out, it was like, this can't be the only thing. Like this is just the opening salvo and this is the first product that they're launching and there'll be a suite of things that are part of the overall humane, you know, kind of stack. Rabbit feels a little bit more like all chips are on this one device to make it. And at the same time, you've got meta,

you know, with Quest 3 and those types of devices trying to again create. their own operating system and their own set of vernacular gestures that they can own. So I guess my question before we move on to Vision Pro is this question of those two new entrants. Do we see that Humane is going to have several successive hardware products or is the pin it? And where is Rabbit in a year? How much runway do they really have to establish a beachhead as you know, a player in the space versus, I mean, like the play date is great, but it's sort of like got a, you know, sort of niche community of people that love it. And it's not going to become like the next sort of, you know, Apple, you know, successor. Well, it's funny to compare it to the playday

because the play date is not Panic's main product, you know, panic still runs the company by selling professional Mac software, you know, the Toyota and stuff like that. Maybe Rabbit sells, you know, lamb, you know, and they provide access to that. Right. But panic can afford to have play date as a hobby, whereas Rabbit is a startup where this is their own thing. It's essential. I assume they have a, you know, if they, you know, and selling out if they decided, if they knew in advance, they were only going to try to sell 10,000 and doing it in 72 hours or 48 hours is a good sign that there's. So part of my question two is also about the price point, right? Like the rabbit is 200 bucks. That is, like you said, that is an impulse purchase. I am so concerned that they built in no margin. to cover customer service or support. Not to mention, so you've got $600 for the AIPIN plus it's another $40 or $50,

whatever it is a month, to get mobile service, right? So it's not just buying a play date and then playing it at home. It's like you're buying a relationship with a company where you want that product to be evolving constantly and it means a cloud connection. Well, I don't understand. The thing that worries me the most about Rabbit, and again, if I, you know, a year from now, my Rabbit R1 no longer works. I mean, so what?

I'm out. I bought this knowing that it might be disposable. Put it on the shelf. It'll be great. But I don't understand how a $200 one-time purchase can cover their ongoing cloud expenses. Just running their LLM. Right, because one thing we know is that these LLMs, the cloud-based

component of them, is extremely expensive. I mean, is not an investor in Rabbit as far as I know. Right, right. But by you know, so I don't know. I mean, I guess they've raised money and they have runway to go through, but presumably, I don't that they're selling the hardware loss, but the service is because the service has to have an

ongoing cost. I mean, there's no other way around it. The argument I made was this was the best launch of a hardware product I've seen in years, but Chris pointed out that as opposed to humane, they came out of nowhere, which is sort of like that's maybe the better way to do it. Oh, I think it is. I think it's also the more Apple way to do it. I think part of what was really weird about the humane hype cycle is that's the opposite of the way Apple does things. Apple does not mention things. I mean, and we all knew this Vision Pro headset was coming, but Apple never said a goddamn word about it until WWDC last year when they were ready to announce it. All right. Let's go to the Vision Pro, which is, you know, I, this is the most date, February

2nd, right? February, like, yeah. Pre-orders, start Friday the 19th of January, and shipping February 2nd. Now, I want to put on the record that I am more excited about this than any Apple product going back beyond the Apple Watch. And you're on the record for your experiences are incredible. I don't want to do that. But what you were just talking about, we were just talking about like, sort of like expectations and things like that.

So there's not an event again to do the launch. And, you know, and again, I am very excited for this product, but one of the things that I thought about this when they announced it six months ago or whatever it was or debuted it was that, okay, then they will come out with something that will be a great, amazing app that they can launch with. and or like they mentioned, other people have said, you know, Disney was on stage with them. You know, so the fact that there is no event that it's just going to come here and stand in line and you can demo it. I'm just curious what your thoughts are in terms of how Apple is thinking about launching this as a product. It's clear that they don't think they're going to blow the roof off. No one thinks that because it's a hugely expensive product. But what do you think about the slower role than even people like I thought they would slow roll?

I'm surprised that I'm not booking a flight for a Cupertino event, you know, in between now and February 2nd. I just assumed it. And it is unusual because Apple does not have new product lines very often. And I just thought, well, they tend to be a company of patterns, I often say. And I thought, well, look at what they did with Apple Watch. They announced the Apple Watch in September of 2014 at the iPhone event. And they had a whole big one more thing segment. And they showed it.

And they said, we've got the sport model and the steel model and the gold model and said it starts at $3.99. And everybody was left guessing what the stupid gold ones would cost. And then in, I don't know what month, but it was like March of 2015, they had a whole separate event to reintroduce it and say, and now this was the onstage era, but it was another onstage keynote. And they had new things. It wasn't that they redefined it since September, but they reiterated it and they had new demos. I thought we would see something like that, you know, like maybe just a half, you know, now that they do these pre-recorded sort of shows that are not really, you know, the keynotes are all pre-recorded. I thought maybe like a half hour one just about this. That would be the opportunity to be like, hey, here's how, pick your partner. Look at this amazing app that in the last six months they put together. Yeah, that's what I thought we'd see more like something like that.

And that that was part of the reason for pre-announcing it. And instead, even the first party demos that they're showing, example videos and this stuff, It's all the same stuff they showed in June. There's like, somebody's, you know, of course, somebody did it, but somebody. There was one new clip of video, yeah. Yeah, there was like one new clip of video. But otherwise, no new, no other features, no third party apps that they, I mean, they had like a list of some of the first third party apps that they know that are already coming out.

Fantastic Al and Jira and I forget who else is on the list. I think that it's why? Why is it sort of low-key? I mean, that's how I would describe it. It is very low-key. And I don't think that it's the price per se. I know $3,500 is a lot for a consumer electronics device. But inflation-adjusted, the original 1984 Mac was not inflation-adjusted.

It started at like $2,500. And inflation-adjusted, that's like $6,000 or $7,000. So like a 1984 Mac in today's dollars was like $7,000. thousand dollars i mean this is 3500 is not unprecedented and there's a lot more people who follow apple and are interested in computers like 1984 was a time when there were an awful lot of people who even were gadget hounds and bought lots of expensive hi-fi stereo equipment and TVs and projectors and stuff like that what i'm suggesting is that even apple knows that there will be people that will line up to do a demo but they're not going to be able to report to wall street that we've sold 100 000

Well, I think the problem is going, honestly, I think the problem is supply constraints. And I haven't heard anything to dispute it. I don't get to go too far into the rumors, but the supply chain rumors tend to be the most accurate because there's some of the companies that Apple works with, they blab. And Apple's relying on these, they're more than 4K each, but two of these displays, one for each eye made by Sony, and nobody else can make them. And by all reports, there's like a hard cap Sony's production of like, it's like $2 million a year at this point. So theoretically, the most of these things Apple could sell is a million. And it might be lower than that.

And that's if it doesn't matter with the prices. They can't sell more than a million of them. But Apple. And that's nothing for Apple. And Sony releases PlayStation's that they can't sell because they have supply chain constraints. What I'm saying is, does Apple know strategically that, this is, we have to low ball this and low sort of soto-voce this because no one's going to pay 35. No, they will. No, I'm telling you what's going. I think what's going to happen is this is going to come out on

the pre-orders start on January 19th. And people who are late to the pre-order on January 19th are already going to start getting shipping dates into March and April. And I think for the better half of 2024, I don't know if they'll be able to catch up by the holiday season, but I think through the first half of the year, I think this thing is going to be back-ordered 12 weeks or maybe even more. Nobody's going to be able to get their hands on one. And I think Apple knows that. There's no way they can make more of them.

They're not going to be able to meet demand. And so why get hundreds of millions or their billion user fan base, the sort of numbers who buy new AirPods or who, buy new iPhones, why get them all excited to buy this thing that they're not even going to be able to get their hands on? It seems so, so there's, I could sort of completely buy that aspect of it in terms of access. The other thing that I would point out or ask is whether or not this type of product and device requires a new type of messaging and marketing. And eventually, or essentially, it needs to be kind of invented. And because the, I don't want to say the response is kind of like

lukewarm, but there's not sufficient pull that says, this is the use case. This is the thing you want it for. Essentially, they have a number of kind of use cases that they developed into, you know, classy, attractive, you know, videos that are on Rails with Happy Paths. They've seen what meta did, and those things are just kind of a video game device. They don't want to be that. And so essentially, in order to lower the risk, I mean, yes, supply chain stuff, allwithstanding, there's almost a need to be like, come to the store, see it for yourself, because we can't show you what spatial computing is, you have to experience it. And that's a different level of commitment than just seeing AirPods, you know,

that are like 300 bucks. And it's like, you know, noise cancellation and, you know, Syrian, whatever else it is. And so that feels like one of the reasons why perhaps the approach is being taken. That's a lot slower, a lot more careful. It seems like, you know, you've got HoloLens in the previous era. You've got whatever Sony's doing that's at CSS this year, I think. I mean, the meta-meta-meta-stuff is pretty evolved at this point. I know, but like, it feels like...

But it's not evolved in the way that Apple's envisioning this. Apple's aspirations in this are so... They're truly profound. I mean, that's why they're calling it spatial computing. It's not because they wanted to invent a new term, and everybody else is calling it AR and VR and VR, and they want to invent their own term, and they're calling it spatial computing.

They see this as like the, you know, as big a deal, as desktop computing and then mobile computing as defined by the iPhone and the modern era of that, that this is like the third flag in the ground. There was, or maybe there was four. There was like the Apple II era where you could just have a computer at all in your home and it was all caps, you know, it didn't even have lowercase letters. And then seven years later came the Macintosh and the computer for the rest of us and truly establishing the paradigm for personal computing for decades to come. Still today, I'm using a computer today that would be very familiar to a 1984 Mac and Dash user.

That's what they are thinking. And I do think, I think Chris is right in the same way that Apple came out with the Mac in 1984, but when did most people buy a computer that had a mouse, whether it was an Apple computer or a Windows computer, but a computer that worked like that with Windows on screen and a mouse that you, that you moved around and you double click on apps and you cut copy and paste and stuff like that. It was the 90s before people bought it, right? And the iPhone was similar.

The iPhone was like a curiosity until, I don't know, four or five years in and then all of a sudden it seemed like everybody had one. I never imagined that I would tell my mother to buy an iPhone until she told me she bought an iPhone. I get all that. Let me ask you this. I've said this on the show.

And Chris and I, unfortunately, are investors these days. so we're thinking about go-to-market. Why is the go-to-market not the iPhone replacement, but the Mac replacement? Because that to me is more compelling. If I just... Wait, is it? I don't know that that's true. Wait, more compelling to me?

No, no, no. You're saying... Sorry, sorry, what I was hearing you say is that Apple is taking an approach where this is a Mac replacement. Well, okay, right. Apple's not saying... As opposed to this is a new category. Yes, yes.

But what I'm saying is the go-to-market... Okay, so forget what I said iPhone replacement because you're right. That's what all the press is saying. And the reason why this is important relative to our previous conversation is the AI pin is being pitched as a replacement for the phone. The rabbit is not being replaced as a replacement. It's being pitched as a watch adjacent product where you just carry lots of things around and you've got your bag of goodies and you pick out whichever one talks to you in the best way. The spatial computing as a concept is a destination. Okay, let me complete my thought because I'm curious to hear John's thought on this. To me, to sell me on a $3,500 face mask is it would replace my Mac because I have a screen here, I have a screen there, I have this whole setup here. And if I could do this and anywhere I went, I could have all of this, that actually makes sense to me. They're saying spatial computing, but why are they not selling this as a replacement for a laptop or a Mac or whatever? Because I think they think it might be, but I think it's years and years ahead.

And I think that one thing Apple's institutionally learned is they're a very patient company now. I think it's one of it's overlooked. I mean, they can afford to be more than ever. Right. But they look at things like Apple Pay. And remember Apple Pay first came out and then it was like, I don't know, four months later, somebody went around and it was like, oh my God, there's only like 7% of stores in San Francisco support Apple Pay. This thing's a bus, you know, blah, blah, blah.

Apple just, they knew it was going to take a while. And it required the retailers to swap out their hardware and stuff. And they knew it was going to take years. And now you go everywhere. And I'm shocked when Apple Pay isn't available somewhere, right? And that just happens. It didn't happen overnight. But we outside, we want things to happen overnight. We want that moment, like January 2007, when Steve Jobs held up the first iPhone. And things did change overnight. It's like, holy shit, that is from the future. That almost never happens. And the idea that something in the vision, you know,

and the fact that they're calling this first one, I keep saying, Vision Pro is that it's as close as Apple could possibly come out and say is, we promise we're going to come out with Vision non-proes that have lower prices eventually. The VisionS. Yeah, or the Visionaire or something like that. The Vision Air, that's pretty good. But it would be, yeah, and if it's lighter too, that would be a big help.

It would actually, again, like in the same way that a MacBook Air is often a better computer for a lot of people than a MacBook Pro because it's lighter

and you don't need the pro. It's like, I don't know. I think Vision Air might be a much better product, not just price, but it'll be years ahead. And maybe it is.

But I do think, I think this might be the first product Apple makes that could replace the Mac for more people than the iPod. Our iPad did because it is bigger,

right? And it's like even the biggest iPad, the 12.9-inch iPad, Pro is a tiny screen compared to what a lot of professional Mac users are like, whereas this device, you can have a giant cinema screen, literally like a movie theater screen in front of you. So let me build on this point, because I think there's something important and critical in what you're saying, and I don't even think we're thinking that far forward enough, right?

What you said is that this could replace the Macs for people. I think this is more interesting from a generational shift that the very first max some people might get in 10 years from now might be a descendant of the vision product. And this loops back to everything that we were talking about, whether it's in terms of conversational software or like the content that I'm seeing developed currently, whether it's generative AI, video content, or imagery, or spatial images, or spatial audio, these things that are so rich and so dense in their fidelity to, I mean, almost beyond reality requires platforms through which people can actually consume and interact with those things.

So my point in question is actually, are there enough developers, one, who have the ability or the knowledge to develop for Vision OS in a way where the apps are compelling? I would posit that the answer is just like the iPhone in 2007 when most people were building web apps and you couldn't build native apps. And it took a long time to move away from the sort of design paradigm of desktop apps into mobile. Well, in a similar way, developers have to go through the entire onboarding and education process and re-education of learning how to build spatial apps. And I got to say, like, Apple's been putting out a lot of content to that end. And it's impressive to see.

But I got to imagine that most developers are still trying to figure out how to, you know, integrate OpenAI and LLMs and are not worried about spatial computing because the audience is not there. And that's why it's got to be a 10-year prospect where kids are coming out of school, knowing how to design for spatial computing and not knowing how to design for the web. And that's going to be the inflection point where someone says, my first computer is going to be a vision X, X, whatever, and that that is the computing thing that happens. And so that's, you know, I think to build on John's point, why it's going to take some time. Yeah. And, you know, and I think Apple had that patience, for example, with the iPhone.

You know, they knew that a U.S., one carrier in one country was not going to set the world to fire. And multi-touch was a huge, huge change. Right. But they, you know, it give us a couple years and we'll, you know, this thing will spread to more carriers, more countries around the world. But it's, there was no way, there was no possible way that the original iPhone, I don't think. I think it was about as successful as it could have possibly been. And Jobs' goal was in June 2000 or maybe it was June in WWC, but they wanted to sell 10,000. million units by the end of 2008, and they did hit that. They sold like 15. They sell 15 million iPhones like in the first five minutes after they go on sale in September now, maybe more.

But the first generation just has to work like that. And I think we're seeing that with Vision Pro. We have to be patient. But there's so much more potential for this device than anything else Apple has ever made. And I could also envision a future where the Mac doesn't really disappear, but the Mac becomes an advanced thing within the Vision OS world, where your Mac is no longer a device. That's what I see. It's a mode within Vision Pro that most users won't need, but if you're a developer, that's what you do.

And so it's weird to me that you still can't make iPad apps on an iPad. I think that's sort of a condemnation of Apple allowing the iPad to thrive and be what it could be. It is a very strange computer that can't make software for itself. I could see in the future where Vision does that. You can write full-fledged vision software

in Vision OS, but you're using a Mac within Vision OS to do it. A virtual. A virtualization. Yes, exactly. Yeah. Where you get the command line

and all the X code and all this stuff that you can't do in Iowa. Well, you have a bunch of AI assistants sitting around you that are just suggesting ideas to you, and then I'll go write this code for you, and then it comes back and solves it for you. Right. Yeah.

The first time I used the meta quest, whatever version it was, I was like, can I just, like, the thing that was most appealing to me was the, oh, I want to be in this virtual world. Can I go read a book in this virtual world? So, like, that's almost what I want. Is like, could I code in this virtual? Yeah, yeah. Okay.

I want to transition real quick. before we have to go. I want to, John wanted to talk about this, but also, since Chris hasn't been on the show for six months, largely, as people have heard, you know, the hashtag has been adopted in quotes by threads and meta. It's not been adopted.

All right. So Chris go first. if you want to say what you want to say, and then, John, please go ahead and jump in. And, and, and, but Chris, since you haven't had the platform, yes. Let me go. Let me go first and just say that I don't remember when exactly you did. But I, again, in a way, this is one way that you and I are, I think. Links forever.

Link forever is there's, I can't tell you how frequently now people come up to, or not come up to me on the street, but come up to me virtually and say, I can't believe you're the person who invented Markdown. And I'm like, yeah, I invented Markdown and wrote this back, wrote the first harser. That's my baby. And they're like, I had no idea. And same way, Chris invented the hashtag. I guess, I'm going to guess it was like 2009, 2007.

I was that early. But people don't know that. And I think that, you know, certainly my son is 20. I think people of his generation probably just can't imagine a world without them. social media, you know, and that it was invented. You know what I mean? It's, I don't know, I don't know if everybody listening to the show knows this, but Chris just literally came up with the idea that you could just preface a tag and put a hashtag in front of it, put the pound sign in front of it, creating a thing that you could search, because that was the only,

that was the whole idea. And then you'd have the string of characters you could search for, and then all the tweets that had the same string would show up in the search. There was no integration from Twitter itself. So they weren't hot links. You couldn't click on them. They were just text. They were stupid-looking words. Right.

Just stupid-looking words. And but then, you know, and the same way somebody else, I forget, I think it was not quite invented, but similar to at-username was a way of referencing somebody. And retweets. Right. Also retweets. None of the, yeah, you'd type the letters, RT. RT-Pos-us username and then.

Yeah. All of that stuff was invented by. users and then became adopted by the platform and they did like, oh, well, let's do it properly and actually make it a thing people can click. Well, no, actually, I mean, again, and the funny thing is like, you know, Twitter resisted it in the beginning. They hated it. They didn't want it.

And the only way that it came to Twitter was as a Trojan horse through acquisitions because third parties like Summys, which then became the Twitter search engine, had adopted it for their product. How do you know that they resisted it? Who was the people resisting? Oh, I mean, they talk shit about it forever. Come on. old enough now.

No, I mean, Ev, especially Biz. There was actually a little mini fake trailer for like a Twitter movie or something years ago. And there's this like moment where like the actor playing Biz Stone comes around the corner and he's like, hashtag. And he's like so angry. I was literally like the vibe that I got from people at Twitter. And I had plenty of people on the inside that would tell me this, you know. And so look, you know, it worked out.

It's fine. I just like, I guess I will always have a little tip on my shoulder because It's sort of like one of those things where it's like, look, I'm open to better ideas. I'm open to better solutions. You know, I'm open to, you know, evolutions. But in terms of the problems that we needed to solve at the time, it was the best possible kind of worst solution of all of that. Well, think about like a cruise ship, right?

There's 5,000 passengers and there's, I don't know, a couple hundred crew members. And the crew decides where the ship is going and what you're going to eat and what's going to happen and all this stuff. But it, and platforms like Twitter, especially social ones, tend to think of themselves as being the, you know, we're the crew. We steer the ship. But social platforms are different than a cruise ship, right? You didn't need, we didn't need a mutiny to do it.

Well, it's a bi-directional. Yeah. User behavior en masse steers the platform. And there was nothing. Twitter not adopting it and sort of wanting to ignore it and wishing it would go away didn't change the fact that over it more and more people were getting, oh, I get what these things are. I'm going to use them too. I'm going to put them in my tweets. I'm going to search for them. We're going to form communities where our community will use this hashtag so that whether you're following the people or not, you get these de facto groups that the user behavior in large steered the platform. And the resistance was then like, hey, this is our platform. We're the ones that are supposed to be steering this. That was constantly the case with Twitter also.

I mean, old old school Twitter. Yes. I think Twitter in particular had an obstinate culture. Absolutely. Can I, Chris, give you the platform now to express why you don't like how threads has, you would say not adopted. Yes. Go ahead.

So I think it's, let's see. You know, I've written a couple posts like about these things. I don't write as much as I used to. And so I wrote about Blue Sky because Blue Sky wanted to get rid of hashtags and move on to something else. Macedon sort of has changed their adoption and implementation of hashtags to some degree. In many cases, there are very valid complaints about hashtag abuse. And look, you invent the gun, you invent the murder. That isn't what the gun was invented for, but yes, maybe it was.

But in my case, that's not what the hashtag was for. It was to bring preferably offline communities together using social media, whether that was going to South by Southwest or whether it was for the Bar Camp community. The idea was to make the social web reflect more of what was happening in the real world. And so to bring those things together. And so the reason why I have fought for consistency in the design and execution of the hashtag is to build a broader social web, a social fabric that connects all platforms that connect people together.

And then you can build filtering and you can build reputation systems. You can prevent spam or you can do whatever you want in those cases. Look, I worked in Google Plus, and that was the first time where I had to fight for the hashtag because, you know, like many engineering cultures, they're like, look, I know how to do this better. You guys are, you know, the people who came before a bunch of idiots, you know, we'll figure this out, the whole NIH thing. And so Google Plus, they wanted to have spaces in their hashtags where you'd have a pound symbol and then you'd have this space is space, my space, hashtag found symbol.

Great. Okay, fine. Like, that's not very different than what threads did. The problem is that when you send that content, you know, and as someone who's been in the blogosphere and, you know, thought about RSS and, you know, came up with the predecessor to activity pub. The idea is that when you syndicate this content, you need to have some things that are consistent. You need to have a JPEG format. You need to have a GIF or GIF format, even if you don't say it the same way. You need to have a ping format.

All of these things allow different clients on the broader social web to render these things consistently. And so when you break the convention of the hashtag, you break consistency and you break community conversations across the web. And so when threads came out with their approach to hashtags, they allow you to lop off the pound symbol prefix. They allow you to add spaces. And they have not made it clear what happens.

when those tags get federated to the social web, the Fedaverse, nor do they make it clear that the prefix will be added back so that people actually know what is a tag and what is just a normal word. So, for example, I've exported my data from threads and where I have Franken tags, as they call them, on threads, there is no identifier that says this is a tag. Whereas if I mention at John Gruber, then I can find all the mentions of at John Gruber and that relates to a specific person. It doesn't get rewritten as John Gruber or John Smith. It is specifically a username on a platform. And so the loss of cardinality and specificity is something that I think is a great loss, or at least at risk of being a great loss on the social web. And so that's my critique so

far in a nutshell of the threads implementation of Franken Tags. And here I am in the very unfamiliar to me place of the proponent and defender of meta. Oh boy. But I am. I'm a big fan of threads has really opened my mind to re-evaluating my opinion of the entire company. My take on this and why I'm happy, I like the way that they've, honestly, innovated in this area where they're not hashtags.

That's the main thing. They're just tags. And so in other words, instead of hashtag, hashtags is a way of tagging where you use the hash mark on your keyboard. Yes,

they have a hash. And then a string of characters with no spaces because it's, it's not just convention, it's part of the whole thing. And theirs is more like hypertext where like

in a, you know, just the way that you wrap, if you know HTML, an A tag around text, and it can be around anything, multiple words. And I'm happy to see them trying that, because to me, the original idea for hashtag, the brilliance of it is that the idea worked within Twitter as it was. It wasn't Chris saying, it wasn't a manifesto from Chris saying, if Twitter did X, Y, and Z, we'd have this great new feature. No, it was just, hey, if we the users will do this in our tweets right now, we'll have this. We'll have this. If enough people do this, it'll work. It sounds like Web 2.0. Well, so, and I've used this analogy. And again, I do think one of the things that why I love, why I jumped on the chance to come on the show and talk with Chris about it is I feel like too

many of us, everything is so, everybody wants to be angry all the time. And so we've lost the ability to have a good-spirited disagreement, right? It's not even that. It's like, you know, like you could be a Red Sox fan and, you know, like be upset with their pitching. You know, so look, I will say, I will say positive things about threads and their Franken tags. I like the way that they look.

So you have a tag with spaces in it. I think what we, I think it's disingenuous not to identify some of the compromises that come from that design. So for example, how do trending topics work in that system? Have you tried looking for search results for these tags? It just lumps everything together, whether there's a tag or whether there's not a tag. And so you get a very, it's not like Twitter. Like I used to be able to go to Twitter and search for a hashtag and see just the results that had the tags.

That is not possible in threats. And so we lose the ability to have coherence in these emergent spaces, which is what hashtag is provided. And so I agree with you. I'm happy to see attempts at positive innovation. It's just that the way that they implemented it, because when you use the pound symbol to bring up the interface for creating a tag, the default is to remove the hash in front, which as far as I'm concerned from a default's perspective is neutralizing the likelihood that any hashtags will exist.

They also added the restriction of having one tag proposed. And that seems insufficient to cover the needs of, the needs of many posts. Like, I'm happy with maybe three. A thousand is not, you know. Yeah, I'm not sure. And I think they could, they could add to that limit. Of course. Let me say this. Before we run out of time, I just want to use my analogy, which I've used debating with Chris online, which is to file name extensions, where if you have, here's my file.t.txt. It's this, the dot txt is metadata. It is a way of saying this file is a plain text file, but it's not in a field devoted to what type is this file. It's stashed into a field that

is just for the name. And hashtags are very much like file name extensions in that way, in that they are metadata that is not in a dedicated metadata field. They're just- Reveals the intention. Right. And what the data is. And there's also, I think the analogy works a lot, is the other thing. And so the classic Macintosh, classic literally the old one, didn't have file name extensions, didn't use them, and it actually had a separate metadata field for the type of files. And it was far more graceful and elegant and also made cross-platform compatibility. You can do Windows to Mac sharing.

Right. And you could use spaces in your file names. And so Read Me files on the classic Mac were spelled capital R, read space, me with no dot, no text, and they opened in the text editor. And then if you copied that file to a DOS floppy disk and put it in a PC, it was like, I can't even read the floppy disc. It's such a good example. Like, and I just want to like, yeah, go ahead. So again, with this federation angle with threads. And it is weird. This is, and so it's like on the one hand, threads is if threads was doing what we would have thought a year ago meta would do, which is build its own silo with no connection to the outside world other than their own like Instagram meta product, then their own, this is our way of doing tagging on posts would be fine.

The part where I'm as a proponent of Thread's way of doing tagging, the way that I most see the problem is, but yet you're saying you want to federate. And by all appearances, they are, they're working on it. You can already, there's a handful of users like Adam Masseri, who you can actually follow on from Mastodon. On Mastodon. Yeah. And so at that point, it's sort of like, you know, this is why this is why MacOS 10 went with file name extensions. Like Mac, MacOS actually reverted in terms of gracefulness from a platform that had a much more elegant metadata system for type to a system that was crude, but worked

everywhere. And that's that's the, that's where the problem hits with thread style. of tagging. The limit of one, I can't help but feel, is fueled by their experience with Instagram where... Oh, 100%. Thirsty Instagrammers are, you know, there's... I will just add to this point, like, the differences that when you add a comment on Instagram, you are not restricted to characters, whereas on threads, you've got, what, 500 characters? And so the same type of hashtag abuse that you experience on Instagram is far less likely. And so that's also where I'm kind of like, well, which is it?

Which do you want to be? Are you preventing the type of bad stuff that happens on Instagram? Or are you trying to for it? And it just, you know, maybe the greatest irony of this is that when you actually see things like Frank and Tag like tech threads, I can't even say the word. It looks awkward at the end of a post because when people are attempting to tag their posts, it just looks like sort of a strange word in the middle of a, you know, sentence or something now or at the end of it. So anyways, you know, what I will recommend to listeners is you can still create conventional hashtags, just hit a double pound on the keyboard, and then type your thing, and then you too can demand hashtags just like I do. But I will, I can't let that

slide without pointing out the way that the way that hashtags in and of themselves are kind of ugly. Oh, I'm not going to disagree with that at all. Right. Yeah. So you can have that one. That's fine. But the greater benefit and the greater good, look, I mean, interoperability with Windows platform was probably necessary for the max assent. So you know what I mean? You got to like take what you can get. Right. Look, I almost did a show with the 99% invisible guy about the, it's the numero sign. Like it's everything's ugly. Yeah. Right. Okay. Well, you know what? Let me just add this. One of the other little benefits of Chris's invention of hashtags is it's actually helps. helped clarify what people call that symbol.

I think that you would find between... I'm on Wikipedia right now. People will literally call it the hashtag. What else is there? The bang? Or no, no, the bang is the exclamation mark. That's the excellent. But there's other names for it.

The number will sign, the pound sign. Right. Do we say that? Yes. Numeral pound. Everybody now calls it the hash. When I look at... When I look at Wikipedia, Chris,

the photo they have for you. No. Can you edit that, please? I can't edit my own Wikipedia pages. Anyway. Okay. What we need to do is close us down. John. Me go long on a podcast.

Oh, my God. shocking. Shocking. John, Chris, thank you for coming on to do this. But Chris, unless you want to promote anything, John,

we know you have, you know, Daring Fireball and your various podcasts. But please, I will give you the floor to talk about whatever you want. I'll just, I'll throw out a shout out to my dithering podcast that I do with my friend Ben Thompson at dithering.fm, which is two episodes a week, 15 minutes, every episode, not a second less, $5 a month, no ads, two episodes a week, 15 minutes. So if you like short podcasts, unlike this one or my other podcast, because dithering is the one for you. By the way, I got to ask you, like, just in terms of the Spotify paywall, because that's using Oath, like, how is that going? And is it working?

I think it's long story short, that's Ben's problem, not mine. But I believe that it is working very well. And I, you know, and I was resistant to Spotify's attempt to sort of take over. Yeah, but I feel like it's another great example where, you know, these big platforms have move towards the open web. And so, you know, I just wanted to get a sense for. Well, and going where the users are, right? And I've never gone wrong in my career building Daring Fireball by listening to what people want. I mean, there's the whole thing where I started selling RSS sponsorships. Because I was doing password-protected RSS feeds for members in like 2005. And then Google Reader came

along and didn't support passwords, but people wanted to use Google Reader. And I thought, well, I'll just give the full feed to everybody and I'll put sponsorships in there. And now it's like half my income. That's amazing. Awesome. Listen to the users. Great. Yeah. Yes. So, you know, whatever, Daring Fireball. By the way, John, I'm going to give you this. You and this will be a little not great, but Adam Carolla were the two podcasts that got me into podcasting. You and Dan Benjamin back in the day, and so I've made my living for six years now as a podcaster. So I appreciate you tremendously. Well, that is a thrill to here. Honestly, I bet no sarcasm intended. I mean I can remember going to a Jersey mics and listening to you on an actual iPod and be like,

that's where I could do that. I love Jersey mics, by the way. Yeah, me too. All right. Chris, be well, John be well. Thanks for coming on the show, everybody.