He tried 200 to-do apps so you don't have to

He may have tested every productivity tool in the world, but in the AI era, which ones actually solve your problems and which are a waste of your time? That's this week on Platformer. This episode is brought to you by Jira Buy Atlassian, where teams and coding agents get the context they need to do the right work. Try it free at Jira.dev. That's J-I-R-A dot D-E-V. Welcome to Platformer. I'm Casey Newton. Last season, we talked about what AI means for jobs and the risk that huge numbers of jobs might soon go away. This season, we're turning our attention to what you can do to keep your job. What tools and strategies can you use to keep your advantage in a world where AI capabilities continue to advance? There's nobody better to help us map that landscape than my friend David Pierce. David is an editor at large of The Verge, co-host of The Vergecast and the author of Installer, the weekly newsletter where he tells the world what to download, watch, and try. He spent his entire career test-driving the tools that the rest of us use to organize our lives, and today I'm going to ask him what's real and what is just productivity theater. First, though, as always, we begin by checking in on the state of AI and the economy, and that means it's time to bring in Platformer Fellow and Gen Z AI Correspondent Ella Marcianos. Ella, how are you this week? Um, I'm wonderful. I just had a beautiful 4th of July weekend. I went to a nice cookout and also to a more heterodox celebration of Canada Day, which is apparently on July 1st. And how do Canadians celebrate Canada Day? Uh, there was only one Canadian at the gathering, interestingly. Um, but I think a lot of kind of strangely flavored Canadian sodas were consumed. What is the strangest Canadian soda that you encountered? I really cannot remember the names of any of them. They, like, flew through my brain. Well, perhaps something for our listeners to investigate on their own time and discover the dynamic world of Canadian soda. In the meantime, though, Ella, I'm wondering if you have any news to bring us from the realm of AI and jobs. Yeah, so in fact, I have some news from both the U.S. and Canada about some people who really don't want to use AI at their jobs. So the Digital Education Council, which is a consortium of global university educators, did a survey that surveyed 18,000, over 18,000 faculty members across 35 countries. And one thing they found is the percentage of faculty members in the U.S. and Canada who plan to use AI in teaching has gone down 9% from 2025 to 2026. Wow. So even though the technology has been getting better, the number of Canadian teachers that plan to use it as part of their work is going down seemingly significantly. Yeah, American and Canadian. Oh wow, okay. Yeah. And in the rest of the world, this actually, like, isn't what we're seeing. Like, they have three other regional breakdowns. There's like Latin America. There's a group that's Europe, the Middle East, and Africa. There's Asia. And all of those groups are hovering around 90% plan to use AI somehow in their teaching. And that's about the same as they were last year. But in the U.S. and Canada, it's only 67% of professors, and it's down from last year when it was still in the 70s. And what do we make of that discrepancy? Did they sort of follow up with the American and Canadian teachers and ask them why they felt this way? Yeah, unfortunately, we do not have, like, direct responses from these teachers right now. So it's hard to tell. I keep on, like, trying to think of wild theories, and, like, none of them actually fit regional breakdowns I can think of. Like, data centers wouldn't be that different in, like, or energy use wouldn't be that different in how big of a deal it is in terms of, like, actual data center buildup between regions. I don't know, they're like, this has, like, in some sense, stumped me, but also on a vibe level, it makes a lot of sense. Like, yeah, just like when I talk to mostly the population I talk to about AI is AI researchers and AI policy people. And, like, in general, when I talk to AI researchers and AI policy people in the U.S. and Canada, they just are way more negative about AI, like, across the board in, like, a really strong way where it, like, almost seems like this, like, very pervasive cultural effect. Right, and we know that when we look at the surveys of Americans of how they feel about AI, its popularity has been rapidly declining. And so it sort of makes sense that teachers are just sort of part of that American populace that doesn't like what they're seeing. At the same time, I can also imagine that many of these teachers have now just had a few years of experiences with AI in the classroom, and they're not liking what they're finding. They're finding students are cheating. They're finding students are not learning the material. They're finding that their place in the classroom feels threatened in a way that it didn't before. So part of me feels like it would be surprising if the adoption of AI were going up in American and Canadian classrooms. But what makes you think that it ought to be going up? Yeah, so I don't know. Like, AI knows a lot of stuff. Like, I was a university student, I don't know, like, last year when the models weren't even that capable. And while I would not necessarily want to be getting lectures from an AI lecturer, like, you know, as a CS student, while I was not vibe coding and I was not allowed to vibe code, and I'm actually kind of grateful for that. I do think I learned stuff creating antiquated artisanal code in C. I, like, I essentially, like, you know, used it as a replacement for Stack Overflow. It, like, told me a bunch of random stuff that I was forgetting. And that, like, was a little bit faster, I think, for, like, memorizing all of the, like, syntax stuff I needed compared to, like, if I didn't have AI. That said, like, I do think, like, most of the places where I've seen myself or students I know actually finding AI useful is, like, when they seek it out in contexts where they're allowed to, like, customize towards what they actually think they need as opposed to, like, teachers intentionally doing some sort of AI integration in the classroom. Right. That makes sense. I mean, I do see a world where a teacher could think of AI as a partner to them in the classroom and maybe it's able to do some one-on-one coaching and tutoring with students that are falling a little bit behind. Maybe it can help with grading or evaluating papers, although, of course, you know, that could be very fraught. But it's easy to imagine a world where, like, if it worked, a teacher might like it. But again, I think all those other concerns that they have are real. Any other takeaways from this study? Yeah, I don't know. It's just, like, there's a really big difference between these groups of people. Like, a roughly 20% difference in, like, how much they want to use AI in the classroom. And, like, people in these different regions have access to the same tools. They're teaching different curricula to a certain extent, but not completely. And so I think that this, like, really does show us that, like, the extent to which people are willing to adopt AI is just, like, not some sort of standardized objective sense of how useful the tool is that everybody will agree on. It's, like, very contingent. It's very culturally mediated. And I think we should all, like, keep that in mind, like, whenever we feel super confident about some assessment we have of AI in a specific context. I think it is a great point that the opinions do seem to vary widely around the world. Well, it's an interesting study, and it speaks to a question that we're going to be asking a lot this season, which is, what kind of tools do you want to use in the AI era? And which tools do you not want to use? And so I feel like today we learned at least what American teachers seem to not want to use. Speaking of that, I think it is now time, Ella, to bring in someone who can tell us a lot more about tools. Ones both that he likes and does not like. After the break, my conversation with David Pierce. Jira buy Atlassian is where agent speed meets team intelligence. Assign any work item to your favorite coding agent and see exactly what it's doing without leaving your flow. When something stalls, you're not digging through logs. You can see what's running, unblock what's stuck, and stay in control at scale. Don't let your agents start cold. The teamwork graph feeds them context from across your entire stack, delivering 44% more accurate results with 48% less token usage. So the work they pick up is the right work, done right, the first time. Try it free at jira.dev. That's jira.dev. My guest today is David Pierce, editor at large at The Verge and co-host of The Vergecast and Version History. I've known David for about as long as I've been a tech reporter. He worked with David Pogue at The New York Times, was a senior editor at Wired, was the personal technology columnist at The Wall Street Journal, and editorial director at Protocol before returning to The Verge. And since 2023, he has written my literal favorite newsletter, Installer, which is The Verge's weekly guide to what to These really beautiful to-do lists and never look at them. Just never. And then I'd be like, Well, that's weird. I don't do anything on my to-do list. Maybe I'll get a new to-do list. And so the biggest change for me has been figuring out sort of what I need to be successful. And the problem is I've become very clear on that. And now there's basically this incredibly annoying thing happening where there are a handful of apps that all have every feature that I want minus one. And then they keep adding things that are like the thing that I want, so I go back to that try that app and then realize it still doesn't have the actual feature that is very important to me. So I leave for one of the other ones that has it. And I'm just, I'm in this cycle basically between five. It's two to-do list apps. I basically am switching between Todoist and Reminders at all times for a variety of reasons. And then I'm bouncing between four or five notes apps depending on which particular pain point I am willing to put up with at any given time. I switched today, Katie. Like this is not a bit. I switched notes apps today. What did you switch from and to and why? I'm just an insane person. No one should ever hear me say this out loud. So I switched from an app called NotePlan, which I really like. Obsidian, it has become this very popular app by basically just building an app on top of a set of markdown files. I think that idea is very important, right? That fundamentally, the thing at the base of it should be a bunch of things that I own that do not exist in some proprietary format that are not hard to get out, that are not hard to read on some other device. I have a folder of files that I can upload to Google Drive. I can put on a thumb drive. I can print out and hand to my wife. Like, that's very important. NotePlan takes that idea and basically spins a bunch of really interesting, like, task-related stuff on top of it. Obsidian is very good for writing notes. NotePlan is much more geared towards being like a sort of daily notes and tasks app and does it very well. So I used it for a long time. And then Craft, which is the white whale of my productivity applications because it is the one that is the closest to doing everything correctly. It's almost a really great calendar app. It's almost a really great notes app. And it's almost a really great tasks app. And every time they make a tiny little bit of progress in one of those directions, I throw my entire life back into it, run into the edges again, and leave. So the update literally today was that they've done a much better job of integrating tasks around the app. So now you can see all of the relevant tasks everywhere you're supposed to, which sounds like a small thing, but it's just not how it worked before. And so now, like, it integrates with Apple Reminders. So I can say, you know, I can say to my phone, like, remind me to water the grass tomorrow morning. And now that is in my daily note in Craft, which is a very important little bit of the synergy that makes this stuff go a long way. So for now, I'm in Craft. By the time anyone hears this, there's a strong chance I will no longer be in Craft. And when you say you're in Craft, you are also in Reminders. And does that mean you are also in Todoist as well? So Todoist and Reminders are, again, two apps with opposite strengths for me. My sort of running theory of all of this stuff is that captures the most important thing by a mile and none of the rest of it matters. And I think AI is actually increasing that fact, that you don't have to build a system. You don't have to care about where any of it is because we actually have technology that is making it easy to find and organize these things after the fact. The whole idea is just being able to get it out of your brain and into something. It's useless in your brain and it's actually incredibly useful as long as it is somewhere. So Reminders I really like because it has, it's unparalleled in its Siri integration, right? You just say, remind me to water the grass at 7 a.m. tomorrow, and it just does it. I don't have to, like, remember the weird syntax to say the name of the app before Todoist sounds like to-do lists. So that trips up Siri like every three times. It's just one of those things that, like, Reminders is the cleanest way I have ever found to get something out of my head and into a system. And so I always end up coming back to it because it's just so fast. Todoist is a vastly better app for, like, managing products and to-dos, but most of the time I come back to what I actually just need is a list of all of my tasks. I don't need it. I don't need it any more complicated than that. I just need them all somewhere. And Reminders, because it does a good job of sort of percolating around other apps too, is mostly the one I've gravitated to. Got it. Okay. So it's sort of cartwheeling between this, like, native system feature of the Apple ecosystem, a really good to-do app, and then a, like, pretty good at everything app. That is the current landscape of how, David, you are trying to stay on top of your very messy personal and professional lives. There's more. There's slightly more to it than that, but that's the base base of it. Everything else sort of flows around those things. Yeah. When you interviewed your colleagues at The Verge about their own productivity systems, my takeaway was that the lesson you learned was that simplicity and a single source of truth beat the perfect system. Why are those two elements so essential for getting work done? I think it's the... How should I say this? I think having a single source of truth solves two problems at once. And we don't actually think enough about either of these problems. It's knowing where to put it, and then it's knowing where it is. And I think that is, like... That's the whole thing, right? Like, I have something in my brain, and I see this with people all the time because I have made myself public as a productivity nerd, so people love to talk to me about this stuff. There's, like, you can have the sort of best-fit solution for every individual thing. You're like, well, if I have this link for this thing, I'm going to put it over here. And if I have this thing that I need to do for this thing, I'm going to put it over here. But then if I have this other Google Doc that I need to put... And what actually happens is all of the systems just break down because there's too much friction to put stuff anywhere. So the idea of just saying, this is the place I put things, is actually the single most important thing you can do. And then the reverse of that is, I know where it is now, right? And there's... You don't have to build some elaborate system that will reveal it to you at the right time. You don't have to go and, like, jump from app to app to app to try to figure out where this thing is that you were doing the other day. I think people have this experience all the time, right? You're like, is this in my email? Is this in my Slack? Is this in Google Docs? Is it in my to-do list app? Is it in my notes app? Like, the fewer number of places that can possibly be, the more likely you are to actually go interact with that thing. And then this is the newest part of my theory, and I'm curious how you feel about this. I've become a big believer in mess in those spaces. That I used to think, I don't want to see any tasks that I don't need to worry about right now. I only want to have the most important things in front of me every minute. And then I started talking to authors and very productive people, people who are working on lots of creative projects. And what they discovered is the idea of actually seeing the things over and over and over and over and over again is really important. That I'm forced to wade through my to-do list every day makes me more likely to accomplish the things on my to-do list because it actually plants it back in my brain. It helps create connections between things that actually you should be confronted by all of your stuff as often as possible. And that's another reason that I think having fewer places for that stuff is just really valuable. Right. And this is the logic of, like, seeing your tasks in an app like Craft because you're going there in the morning, you're starting a daily note, maybe you're writing a little bit in your journal, you're looking at your calendar, but all the to-do lists are right there and you're not just sort of hunting all over. So if folks take nothing else away from this conversation, it is that simplicity is your friend, a single source of truth is your friend. And now David and I will spend the rest of the conversation talking about how we've worked to needlessly complicate our lives and spread all of our writing and content across as many apps as possible. That's correct. Can I give you one really stupid example before we move on? So there is this guy named Jeff Wong who wrote a blog post in, I think, 2022 called My Productivity App as a Neverending .TXT file. This is, like, a formative blog post in the David Pierce theory of technology. And basically what this guy did is he started an incredibly long text file and he would have the date and then kind of a record of all the things he had to do and all the stuff There is the way that many CEOs believe that AI is making them massively more productive, but their workforces tend to be much more skeptical. At The Verge, I feel like you sort of straddle kind of the line of like, you're a worker, but you're also like in leadership there. What's the dynamic at The Verge right now? Are y'all AI curious, enthusiasts, skeptics? We kind of run the gamut, I would say. I think a kind of foundational truth of The Verge is that we evaluate the products, right? And I think it is like, Nilay, our editor in chief, always likes to say, that's the source of our power, is like, you can give us all the bullshit you want, but if the thing doesn't work, I'm going to find out, and then I don't care what you say anymore. Because the thing doesn't work. And I think we've just run into this weird thing with AI over the last three and a half years now, where the gap between what people say and the ground truth of life as a user of these products was so enormous for so long that everybody just kind of lost the benefit of the doubt. And now it is genuinely smaller. There are a lot of things that we have been promised for a long time that actually now kind of work. A lot of them still don't, and a lot of them don't work all the time, but some of them do. Some of them really earnestly do the things we were promised back in 2022. But I think... Go ahead. Well, I was going to ask if there is anything that AI does for you well enough at work now that you would hate to live without it. Oh, that's interesting. Um, no. No. I don't think so. I think at this point, the number of things... Oh, no, that's not true. I have a good one for you. It is very good at drawing connections between a bunch of sources, which is actually a thing I have come to appreciate. I use Notebook LM from Google for this a lot in a way that I really like. And to just basically sit down and read a bunch of stuff, and it's really helpful for finding people to talk to about certain things. I'll read a bunch of white papers and be like, Is there anyone who was quoted in all of these? And like, often there is. And so it's like, Oh, I should go talk to that person. And so as a way to sort of take a bunch of stuff and reduce it to something much smaller, that's very helpful. And I think especially in the stuff that we cover, part of the job, frankly, is to be kind of a mile wide and an inch deep. And so it's like, to be able to be that wide very quickly and then be more selective about, okay, here's where I'm going to go really spend my time and try to find the smartest people the fastest and try to go do the best work. That doesn't feel irreplaceable. It just would take longer. And there would be a certain kind of thing that I don't go read as a result. You know what I mean? That makes sense. I think for me, a big one is like transcription, right? Which, like, I think maybe most people don't even think of as AI really. Because maybe it doesn't feel quite the same as like Notebook LM, which feels way fancier. But my God, the hours that I used to spend transcribing interviews that I just like, don't have to do anymore. You're completely correct. That's also, like, that actually is the only one I can't live without. Because I used to, I mean, I used to spend an hour interviewing somebody and then four hours transcribing it. And now, shout out to TurboScribe AI. I just upload it and it spits out a not quite perfect, but like functionally useful transcript in 10 minutes. And I don't, God help me if I ever have to go manually transcribe anything ever again. That's wild. I don't have to listen to my voice anymore. It's so great. And that's a dream. I know. Just not listen to David's voice. No, we love David's voice. We love David's voice. I will also say, and this, here's where I feel like this gets a little edgier. And, like, some journalists, like, I think would say, like, I would absolutely not do this. But I've been using it for podcast prep. You know, like, when I have a pretty good, I did it for this episode. I had a pretty good idea of what I want to talk to you about. I wrote a long prompt and I walked away and I came back and I had a shell. Some of the questions were stupid and I deleted them. And a lot of them were, like, directionally right. It's like, I do want to ask about this. I do want to ask about that. And I feel like it saved me time in a meaningful way. And then, you know what else I did? In the same conversation, I said, summarize these topics so I can send an email to David to kind of let him know what I want to talk about so he can prepare for this podcast. And I'm telling you, this stuff took, like, three minutes and that would have used to have been, like, an hour and a half or two hour, you know, process for me. So I'm always trying to find, like, where am I actually getting more productive as opposed to feeling more productive? And that is a space where I just actually feel more productive. Well, okay. Talk this one out with me because I've been thinking a lot about this recently. So I also host this podcast called Version History. You've been on it. And the show is, it's part sort of litigation of interesting products over time. And it is partly just, like, straightforwardly a history show, right? So it is, like, it has to be deeply researched. It has to be full of, like, interesting reporting and lots of stuff. And I spend an enormous amount of time researching those shows. And, like, on the one hand, it's very fun. And I sort of set it up as a show because I like doing that work. But I've also found myself wondering, like, okay, is there a beginning of this process that I could offload? Yes, absolutely. So, but every time I've done that, I think the show is worse because I think I, like, the less I know the material, and I think I know the material less when I'm not doing as much work. Like, I really do believe the idea that there is a linear relationship between how much of the work that I'm doing and how well I know the thing at the end. Like, reasonable people could disagree on that, but I firmly believe that. There's a great, like, Andre Karpathy quote that's just like, you can't outsource your understanding. You know, like, there's a lot that you can hand over to AI, but, like, the more that you hand over, the less that you are going to understand. And that does really matter when you're hosting a podcast. I think for me, yeah, go ahead. So I think one thing I've been thinking a lot about is, like, I will go and I will read, you know, for hours and hours and hours and I will talk to people and I will end up with this giant, massive notes. And one easy thing to do with AI is to basically say, hey, I have all of this. Put this in chronological order. Another thing AI is fabulously good at, right? Like, here's a bunch of dates. Here's a bunch of things that happened. Can you put them in chronological order? And I'll go through and check at the end. But, like, I wrote all of this. Can you put it in the correct order? I don't think I have an intellectual problem with that being an AI thing, but like... And I don't think I also have an intellectual problem with being like, hey, can you recommend, you know, the 10 most cited things about XYZ? Like, if I were to, if I were to start researching Nest, where would you think I should start? I don't think that is outsourcing your understanding. But it falls down such a slippery slope into, I've actually done none of the work and now I have an outline for a show that I've never read. And I'm just, I'm so terrified of getting to that point that I don't quite know where to cut it off. Yeah, and I think it probably is worth a little bit of exploration. I do think that, yeah, there is a version where you just sort of snap your fingers and it's podcast prep. And there are podcasters that just read things that their producers hand to them and they basically are seeing it for the first time when they sit down in front of the microphone. I think those podcasts are mostly not great. I don't want to make that kind of podcast myself. I know that you don't either. And yet I can also imagine a world where, like, you know, version history follows a certain format. You could take the first 20, 40, however many episodes you've done, sort of feed that into an LLM, say, like, this is kind of the format that our show follows. Our next episode is about Nest. Like, go out and do some research and, like, create a prep document that is kind of in this format. And then, you know, maybe that does save you a couple hours. I think the tension is, the LLM is now going to choose things for you that it thinks are the most important moments. And you're sort of outsourcing, you know, to the LLM, its judgment of what is important. And you might have made different choices had you not done that. And so now you maybe And you go through and it's like, menus exist. Do you know what I mean? Like, the experience I had with ChatGPT was as if there were no menus anywhere in Starbucks, and you walked up to the barista and they spoke to you a list of every available drink at Starbucks. Like, we've figured out better ways to do it than that. And then when you say, you know, I want an iced coffee, she goes, okay, well, let me read you the possible milk options that you can add. And then you can, and it's like none of this works. And then I heard from a bunch of people in and around Starbucks who were like, oh, I don't know if you know this, there's one button in the Starbucks app that just says reorder. It's like, we have solved this problem. We do not need to AI our way through this problem. We have pressed the reorder button through this problem. And I think there's so much of that stuff right now that everybody is trying desperately to like reinvent the wheels, and what we're actually doing is just coming up with needlessly elaborate, needlessly brittle ways to do everything. Like, I'm not here to pretend that Airbnb is the greatest interface in the world, but it is so much better to use it yourself than to watch ChatGPT try to use it on your behalf. Absolutely. My answer to this question, for what it's worth, and somewhat to my surprise, is deep research, which seemed so cool when I first started using it, but then I just realized, like, all I'm doing is generating 10,000-word documents that I am not reading, and whenever I zoom in on anything, it either feels like filler or something that has a questionable source. You know what I mean? So you sort of are just better off doing the actual research than the allegedly deep research. Yes, I think one of the things that I've thought about AI since the very beginning is that it is a very useful way to do things that don't have right answers and that are pretty low stakes, right? Like, my first favorite thing to use LLMs for was movie recommendations. You just input, like, I like this movie, this movie, and this movie. What movies should I watch? It's actually pretty well-suited to put those things together in a useful and hard-to-replicate way to be like, here is a movie you will probably like based on what I know about those things. Does it need to be the absolute exact platonically perfect answer to that question for me to get value out of it? No. Does it need to even, like, get the movie right every time? It's like, no, if it hallucinates a movie, who cares, right? Like, I just don't watch that movie, and then I go find something else. So that's the sort of thing that it's like, back to the sort of plumber example, right? Like, I do not actually require the single greatest plumber in the history of Alexandria, Virginia. I just need one who's gonna be good, because I'm going to get three quotes, and I'm going to go with the one I like best and is the cheapest. Like, I need you to start this process for me in a way that does not require you to be completely correct about everything all the time. And the problem you run into with deep research is we're operating under the assumption that you are giving me lots of correct information, and if I have to go redo all of your work, what the hell was the point of any of this? So you're sort of, you're like, you're just giving the thing a task you can't actually expect it to correctly complete, and that mismatch just still feels so real to me. Right. There's a version of the future where AI doesn't reduce the amount of work we have to do at all, it just raises expectations for everyone. Everyone writes more emails, generates more slide decks, produces more slop for other people's AI to summarize. The workday stays the same length, the volume of content triples. Does that feel like a realistic scenario, and are there any tools or practices you've found that sincerely give time back to you? Not really so far, to be totally honest. I think the thing that is true, and the thing that I find hopeful, is that in the same way that technology has always taken busy work away, this could happen again, right? Like, I think the history of spreadsheets is really instructive in the history of AI, that it didn't obviate accountants, it didn't change the way that we think about business, even. It just changed the tools with which you could do the job. And in fact, what it let you do is do the work much faster, and it let you forecast more into the future because you could operate with more numbers, changing them more quickly, and you could see the outcome of those numbers instead of doing it by hand. There's some very good, Steven Levy wrote a great thing many, many years ago about spreadsheets that I encourage everybody to go read because it's a really interesting cultural study of what happens when suddenly a thing can do the work for you. And everybody thought we would never have accountants again. And then what it turns out is Excel is really hard to use, and one of the skills of being an accountant is now proficiency in Excel. I think there's gonna be a lot of that with AI. And I think the thing that I am hopeful for, and frankly, already grateful for in the technology, is that it can do a lot of the busy work I don't want to do. It can go in and clear out the thousand emails from Land's End for me every morning. It can relatively successfully surface the tasks that I have today. It can collate a bunch of information from a bunch of different places and put it all into a single text message that I get every morning. Like, that's the kind of thing that is useful. It's not creative. It's not even thoughtful. It's just administrative. And I think that kind of task, there are a limitless supply of administrative tasks in our lives, and so I'm not worried about people who do administrative tasks running out of them to do. But I think the hope is, I don't think anyone who tells you that we're all gonna stop having to do work because of all this, and we're gonna live on universal basic income and just live lives of leisure, is ridiculous. But I think there is some possibility that we're going to get to sort of cut off the bottom pyramid of what we do at work and spend more time doing more interesting work, which is why everybody should stop firing their workers, actually. You should hire more smart people to go do smart things because that's going to be the thing that, like, raises the ceiling of all of this because sitting around turning a bunch of paragraphs into a PowerPoint deck is not useful work. Like, no one needs to be doing that job. We can all find better things to do than turning some Excel files into bar graphs in a PowerPoint. But there are people who do that for a living because we have to. And so I think if we can automate that out of existence, I am actually extremely hopeful for what we can all go find to do instead. You wrote the review that doubled as an obituary for the Humane AI pin, and you were similarly unmoved by the Rabbit R1. Two years later, we're seeing a similar promise minus the hardware, agents that do things for you while you do something else. How close do you think we are to living in the future promised by devices like the Rabbit? And what do you think the role of AI hardware might be in the next couple of years? Input. It's the whole thing. I think the idea of having a device that is easier to get some kind of information into than my phone is really powerful. I'm kind of out on an island on this one. Like, there are a lot of people I work with and in the world who are like, you're an idiot. Your phone is fine. I actually don't think that. I think the friction in I have to get out my phone from wherever it is, unlock it, open an app, tap on a thing, and type, is actually quite a lot of friction. And if we can get it down to I talk into my ring. You know Eric Mijakowski a little bit. He has this thing called the Pebble Index that is literally just a button on a ring. You press it. You talk to it. You put it away. That is vastly less friction. And if they can build that into a thing that's like, oh, you just set a task. Let me put that into your task manager. Or, oh, you just set a note. Let me put it in your notes app. That is like a powerful new productivity tool. You know the other thing that has going for it is when you get out your phone, you also have to navigate through a thicket of notifications and distractions, right? How many times have you opened up your phone to do something and you end up on TikTok? And you're like, was I going to write something down? Right. Or even worse. Yeah, you open up your phone and then 10 seconds later, you have no idea why you opened up your phone. There's never a moment where I feel worse about myself than that moment. Completely. And so I think where I'm at is trying to figure out how all of that is going to actually practically work, right? Because what we have right now is this run of devices that are like, we'll record your meetings and get out some of your useful information and action items. I think that'll eventually work fine. I think that has a bunch of weird societal problems. Like if you and I sit down to have coffee and I'm like, I'm just going to turn on my AI pin so that we can hang out more efficiently. Like, that feels bad. And it will change the way that we are together. But that Not a replacement for my brain or my family, and I think you're going to be fine. If you take nothing else away from this, don't make AI replace your family. Listen, man, you'd be surprised. No, it's a legitimate thing to say because some people are trying. Now, it's also true that some people have terrible families, so I'm sympathetic to the folks who are looking for support elsewhere, but what I'd say a lot is if you read a thing about AI and you just find and replace the word AI with the word software, everything gets a little more understandable and a lot less scary, and I think we should all maybe do it that way. Like, it's just software, and that's okay. It's sometimes it's very good software, but you should treat it like software. I like that. Well, speaking of software, David, last question. I can't let you get out of here without giving us a recommendation. If a listener is in the mind to turn over their life to a new piece of software tomorrow that does anything in the realm of making them more productive, what might you tell them to check out? There's an app called MyMind that I really love. It's spelled like it sounds, M-Y-M-I-N-D, and they developed it as a really beautiful sort of visual reference guide. So the idea is, like, I'm a designer. I don't know, I guess designers, like, read magazines all day. I don't know. I like wander through beautiful museums and I take pictures of them and I upload them, and it does a really interesting job. This actually, I think, is a fascinating and great use of AI. It automatically categorizes everything that you put in. It sorts it by color. It sorts it by type. It sorts it by, like, all the pictures you take of cars, it understands are of cars and it will find all the cars for you. It is just this sort of endlessly reorganizable set of stuff. And again, when I come back to, like, you just need to know where everything is, MyMind is a really great place to put stuff. So the way I've been using it is it is just an endless compendium of things that I like. If I read an article that I like, I save it into MyMind. If I watch a movie that I like, it goes into MyMind. If I read a quote in a story that I like, it goes into MyMind. If I take a picture of my kid that I like, it goes into MyMind. And so now I can go through and be like, oh, this is, I have just a running set of my favorite pictures of my children. And I, did I read a great article in the last three months? All of them are in MyMind. Like, this is, there's a very old idea of a commonplace book that people have been using for forever, which is essentially this thing. You write down little snippets of things that are important or relevant to you. This is that, but in a way that is sort of searchable and remixable, and you can find lots of stuff, and it's also just beautiful and lovely and available on every platform. It's expensive. I think it's like $15 a month, but it is maybe my favorite piece of software that I use every day. So there's my recommendation. I love that. That's a wonderful recommendation. I've also tried it and can confirm. It is a very fun and cool app. David, thank you for the recommendation. And thank you for joining us today on the Platform Podcast. Thank you. It's been a blast. Platformer is produced by Lindsay Chu and edited by Fitz Harris at Story & Sound. You can watch this whole episode on YouTube at youtube.com slash Casey Newton. My email is casey at platformer.news, and we'll see you next week. Jira by Atlassian is where your team and your agents work from the same context. Try it free at jira.dev. That's jira.dev.