Transcript

00:00:00 - Speaker 1: The entrepreneurial drive or in this case the entrepreneurial drive of feeling agitated, there’s like a problem or a thing that just seems wrong or weird or annoys you and you just can’t stop thinking about it and you look for solutions, but none of them really quite seem to fit.

Hello and welcome to Meta Muse. Muse is a tool for thought on iPad. This podcast isn’t about Muse the product, it’s about Muse the company and the small team behind it.

I’m Adam Wiggins here today with Mark McGranaghan. Hey, Adam. Well, we’re entering a new year here, and for me, the holidays is a time to reflect and think back on what’s learned. And for me, one of the biggest surprises of 2020 was this podcast. We got started on it. It was, I don’t know, just an idea that we wanted to try and then to my surprise, here we are now 21 episodes later with thousands of listeners and something I really enjoy doing with you every two weeks. Do you have any reflections, Mark, on this epic podcast journey so far?

00:01:07 - Speaker 2: Yeah, it’s been great. I remember when we recorded our first Uh, demo episode, episode zero, if you will. I was using my AirPods in this Airbnb in Arizona, I think, and, uh, it was really fun, and I didn’t, didn’t quite land the first time, but we felt like there was a good spark there. And sure enough, we followed through with it, and over the course of the last 20 or so episodes, we’ve gotten really great reactions from our listeners, and that’s been the most rewarding piece, I think.

00:01:33 - Speaker 1: Yeah, actually, in that same spirit of prototyping we might use for product work, we sort of prototyped the idea we had to do a podcast by recording, I don’t know, 20 minutes I think it was, yeah, in the freezing attic of this Airbnb we were staying in in Sedona, just recording into the voice memo apps on our phones, and I kind of just edited together with Ferri, which is this little audio editor for the iPad, and even I think found some stock music and just shoved it in there just to get the feel, and then we listened to it and said, OK, does this.

And the audio quality wasn’t good, it wasn’t long enough, it wasn’t, you know, we were still finding our groove, but as you said, is there a spark here? Is there something worth investing in? And we felt like there was and that kind of caused us to roll forward with figuring out how to do the real thing.

So I figured that since we’ve been at it for a little while, as well as the palate cleanser of the new year, I thought it’d be good to do something we wanted to do for a while, which is a listener questions episode. So happily we ask for kind of input feedback from folks at the end of every episode, and we got plenty of those over the last 9 months we’ve been doing this or whatever, sometimes by private email, sometimes from former colleagues or friends, very often from folks we don’t know at all, sometimes. Twitter, etc. And then I recently just posted a thread on Twitter asking for folks to contribute questions. Lots of great stuff there. Thanks everyone for sending things in. And, uh, yeah, I just thought we’d spend this episode walking through a few of our favorites. We can’t answer all of it, obviously. And certainly many of these things, I think probably just serve their whole own episode. So you can give us feedback on the feedback episode and tell us which of these you’d like to hear us talk about more. Do you want to uh start us off by reading the first question?

00:03:12 - Speaker 2: Yeah, so to pull the first listener question out of the mailbag here, we have a question from Gary Zoo, and we have to add a blanket caveat here about the pronunciation of names, apologies in advance. I hope we got that right. So the question is on designing advanced gestures. Says I am working on designing a new writing tool for myself that combines some of the stuff I love about Figma into writing. However, I struggle with designing with the right amount of fidelity. It feels like mockups are not enough to truly express the idea, but I don’t want to go into prototyping too early and lock in the design space. How do you design the interactions from Muse and deal with tensions while prototyping?

00:03:52 - Speaker 1: And one reason we picked this, so we’ve gotten a number of variations on these happily we have a solid following of design-minded people and yeah, the traditional design approach is static mockups, you know, you basically have these storyboards.

Here’s a screen that you tap this that flows over to here and then even the prototyping tools, for example, prototyping built into FigGMA, as well as standalone prototyping tools like I used Balsam.

For many years and this kind of tap through screens or click through screens is a standard mechanism, but if you’re designing something with really sophisticated gesture space, for example, or lots of sophisticated transitions, um, that kind of doesn’t quite fit that storyboard thing and I feel like that’s a little bit what Gary touches on here with this mockups are not enough to truly express the idea.

And yeah, my experience with this, we’ve tried a lot of things and we definitely don’t have an answer, but certainly my experience has been that you just kind of wing it depending on what the, what the thing is.

So sometimes we do do a lot of stills and just kind of classic flows, often with kind of textual descriptions that say when you do this, this should happen, but getting more into, you know, slightly more creative approaches, for example, Julia, I remember, did a great one quite some time. Back by sketching out the screens on paper and then recording a video of herself kind of manipulating the papers and so then things like sliding and other, you know, you could see the hands and you could see the gesture and she would kind of like you know puppeteer essentially.

Something that’s a nice way to express the idea and we’ve also toyed quite a bit with, for example, I’ve used Framer a bit sometimes I just use HTML. Leonard has been doing stuff with origami Studio to again be more gesture and motion oriented and things you can even run on the actual device on the iPad. I don’t think one of those feels like a silver bullet or a catch all solution.

00:05:42 - Speaker 2: Yeah, I don’t think there’s a silver bullet here or a single answer, but I do think it’s really important to make it real.

Adam, this is one of your rules for creative work that I’ve learned over the course of many years working with you. And here, you need to make it real because you can’t really evaluate gestures which are so physical without something there to actually be touching or at least seeing.

But at the same time, implementing a full on gesture system that handles all the edge cases and is integrating into the app is super complex.

So what you want is a way to cheaply evaluate and de-risk gesture decisions. So a lot of what we’re doing here is ways to kind of isolate the gesture problem or attack it more cheaply.

So another example to compliment the ones that Adam gave, we have a now sort of internally famous idea of pink cards on a board, and this is a demo app, which is basically a blank screen and you have Hot pink cards that represent muse cards and you can implement new behaviors with them.

And so the idea is you’re not dealing with any of the other complexities in muse, you’re just dealing with how do you manipulate these cards because we know once we have the cards, we can put an image in them or a text in them or a PDF, but the actual manipulation of the cards with gestures is quite complicated. And so in that way, you can fully evaluate how did the gestures work, how does the implementation go, but you don’t need to deal with all the other complexity and therefore cost of integrating it into. A full on app.

Another example that we’ve done here recently is prototyping in the web.

We’ve been doing this for our infinite canvas experiments where for various reasons, it’s just much quicker to prototype stuff on a web page. You can just do a single page HTML JavaScript thing and quickly learn how does it feel as stuff gets dragged around. Now, there, you don’t have the fidelity of you’re not actually touching something, but you are seeing, OK, these things are moving, how are the content indicators changing and stuff like that. And again, you’re really cheaply and quickly learning.

00:07:29 - Speaker 1: I think another piece of the prototyping is maybe what you described there is almost the reductionist aspect of let me just pull out this one piece rather than trying to put it into a full sophisticated app codebase, and that could even be, yeah, it’s often the case that something like HTML and JavaScript is a faster way to prototype compared to kind of mobile development environments, but Yulia does tons of standalone. prototypes where she just says, OK, we need to explore a new way that we do the transitions because we were changing the visual model and therefore I’m just going to make this iPad app that’s essentially not much more than a hella world of moving these squares around on the screen, but I can try this one idea in isolation, get it to where it feels right, and then I can go to the kind of back porting that into the main app.

So another cluster of questions we have is around zooming UIs and sort of the navigation within a board, and then there’s the navigation between boards, which of course you do through pinching to zoom in and out, for example, hey bang bird, again the pronunciation caveat there says I’d love to hear your thoughts on zooming you. or Ricardo Medina asks, thinks this is one of our more interesting design choices and wants to know why it is disorienting and intimidating potentially to have the infinite canvas. So we wrote about that a bit in our memo, but maybe this is a good one to expand on a little bit. So Mark, tell us about the visual model and why we’ve made the choices that we have there.

00:08:56 - Speaker 2: Yeah, the zooming UI question is really interesting because a lot of people when they first seem, they say, oh, it’s a zooming UI app. Like that’s the thing that it does, it’s so distinctive and so prominent.

But that isn’t necessarily a top level goal of Muse. We introduced a zooming UI because it solves some particular design criteria we had around how you navigate your content.

So for example, I think it’s really important when you press your fingers on the screen and you go to do a gesture and you start the gesture. That it immediately does something and that it’s a continuous experience from you start the gesture all the way to you finished it. And assuming UI is a pretty natural way to do that. Like as a contrast, a lot of apps navigate with like page changes, like you just kind of teleporting around, and that is quite disorienting because you don’t know where you came from or where you’re going, and it doesn’t have the same naturalness that this gradual and incremental transition has, especially when you’re doing a continuous physical gesture. But there are potentially other ways to get that, and there are ways that we’ll consider in the future. You can have a crossfade or a slide in or slide out, things like that. So it’s not so much that we want a pure ZUI per se, it’s that it gives these nice properties, especially in the context of a gesture system.

00:10:06 - Speaker 1: Well, one of the original kind of core ideas or philosophies we were trying to explore back when we were doing this in a research context was how to tap into the user’s spatial reasoning. Because there’s a very powerful part of the human brain we evolved to navigate spaces in the real world and to the extent that computer interfaces can tap into that and make things feel place, feel like a place and access those parts of our brain, the more we can orient and move around comfortably.

And I think in a lot of ways the computing industry, particularly with mobile, has gone that direction a little bit. So for example, standard iPhone apps have first of all usually. Some kind of a light spatial thing. If you tap an item, it kind of slides to the right, then when you tap a back button, that slides you back to the left and you have kind of this sense that the thing exists in a larger 2D space.

I would also argue something like actually the iPhone core interface for the home screen is a very lightweight zooming UI, particularly with the swipe up gesture and the zoom kind of transition, you get this feeling that when you’re looking at an app, you’re sort of zoomed up and then you swipe up and you toss it away and it flies back to the grid and you can tap another one to zoom into that, right? So the part of what I think makes the iPhone and mobile interface is generally much more accessible and much more natural for most people is precisely because it’s tapping that spatial reasoning. So we wanted to see if there was like a power user version of that, essentially, right, exactly.

00:11:32 - Speaker 2: And as for this question of zooming potentially being disorienting or intimidating, we’ve been talking about this in the context of when you’re in a single board. And if you can arbitrarily zoom in and zoom out, and especially if you can also pan in any direction without any boundary, it’s kind of hard to know where you are.

So when you start, you know you’re at zoom, you know, 100%, and you’re in the middle.

But if you do some gestures, like say you zoom into some blank space, it looks blank, just like it did before. So how zoomed in are you? Well then, do you need like a zoom level indicator and then you’re into the whole Chrome thing.

And then if you want to zoom out, do you snap to 100% or is it if you’re close, you snap, or otherwise, you know, there’s just so many degrees of freedom that you introduce, which sure gives you flexibility, but also makes it hard to orient, especially to some kind of standard landmarks that you would want.

So if we’re thinking about our flexible canvas feature and other related ideas, we’re trying to strike a balance between, you have the flexibility to move around and to zoom when you need to see something closer, but also you don’t feel like you have no grounding in your content.

And I keep coming back to this idea of drawing on the, the physical world, and here, like with your desk, you can zoom in, like lean in to see something with more detail, and you can kind of lean back to get something with more context, but there’s this natural balance point that you return to, which is like you’re sitting neutrally in your chair. And that seems really, I don’t know, obvious or straightforward, but it’s a really important idea of like you literally know where you’re sitting, and it’s so easy to lose that in a digital world where you don’t have all these physical cues.

00:13:01 - Speaker 1: a trade-off between oriented, grounded, and freedom, openness, spaciousness. It is extremely challenging. We talk about one aspect of it in our infinite canvas memo that I’ll link in the show notes, but this is an ongoing challenge for anything we ever do with the visual model, which is you want freedom, that’s part of the value of these digital tools. You don’t run out of pages in your notebook, but that actually comes at the expense of too much freedom and you’re just lost.

00:13:32 - Speaker 2: Yeah. So speaking of organizing your muse content, we very often get questions about that. So one recent one, for example, from Akash was how do you organize your boards by ideas, by articles or something else. So how do you approach this, Adam?

00:13:46 - Speaker 1: Yeah, I usually have the answer of however you want or don’t worry about it, which sort of is unsatisfying.

In practice, when, you know, of course, anything you put in use is private, we can’t see it, but I have had the opportunity to kind of watch over someone’s shoulder either virtually in a screen share session or sometimes in person and see how people organize their stuff.

And there are folks who do a very careful, OK, here’s my work board over here, and here’s my philanthropic activities over here, and here’s some stuff I have to do with my kids over here. And they basically have, you know, subboard kind of like organizing folders on your file system or something like that.

For myself, I find that I tend to use it in a more messy way, and this is especially true for me because Muse is really about what’s on deck. What am I thinking about now? Um, it’s less about the long term kind of archival of um project material and more about what I’m working on right now.

And so I tend get a thing where I have a couple of work-related boards on one side of the screen, maybe kind of in the upper left, and I have some personal stuff that’s maybe more to the right, including maybe some journaling, but also some personal admin, life admin stuff, and then I have like kind of a little archive space down the lower left, then I might, if there’s one particular project that’s really top of mind for me in the moment, I might make that a little bigger and put it towards the middle.

So then I do end up with like a board per Not quite per project, but per like chunk of thinking about a project, and those kind of migrate to my archive as they’re not relevant anymore. I’ve moved on to new things. There doesn’t tend to be like a deep hierarchy where it’s neatly sorted and taxonomically correct. It’s more just I have a board for stuff that I’m thinking about, you know, by topic, and I just kind of put them roughly in clusters and then eventually those move to an archive and that’s kind of.

00:15:38 - Speaker 2: Yeah, I have a similar approach where everything comes into a sort of inbox or desk, and it’s very much a messy working set. It’s basically whatever I’m thinking about today. And then from there, any structure is sort of emergent. So if I have a few things on my desk that are related and I want to keep, I might make a board for that and put that board on my desk.

But An active project I’m currently working on. And then as I complete that project or it becomes colder, I might shuffle it down further into the hierarchy. And that way, the hierarchies that I might need sort of emerge organically. So I don’t try to pre-plan anything. It’s just once my current space gets too many items on it, then, OK, I need to basically rebalance the tree and push some nodes down.

Adam, you also mentioned kind of a mix of personal and work. One question that we didn’t pull up explicitly, but I’ve gotten a lot is, you know, how do I make different workspaces or can I have different users in Muse? We made a pretty explicit design decision in Muse that the way you separate your top level context shouldn’t use a different mechanism from how you organize things at a lower level. Like we want to build these general purpose primitives that you can use to organize at any level of your workspace. So for example, I have a big top level board for personal stuff and for work stuff, but those are just regular muse boards. We don’t have a separate concept of like muse workspaces or muse users or muse accounts or something like that, um, because we want you to be able to reuse these primitives and flexibly recombine them however you want.

00:16:59 - Speaker 1: The next question comes from Chris Corella. He says, I’d love to hear the projects and use cases people are using use for.

And this one is for me is a personal, not sure what you call it back burner project or just a thing I’ve wanted to do for ages, which is have a gallery that shows sort of the type of person or says a few things about the work they do, but also shows their boards and we did a little of that in the newsletter sometimes users would send us screenshots, by the way, we love screenshots of your boards and of course they’re personal and often people feel vaguely embarrassed because they’re messy because thinking is messy. But it’s really great, you know, when you’re comfortable to share that, giving us a little peek at how you use it really, you know, pictures worth 1 1000 words, as they say.

But in any case, for a while, I would ask folks for permission sometimes if we could put that in the Newsletter kind of dropped off on that. Love to do like a user gallery or something like that because we have really some pretty interesting people doing some pretty interesting things and I get a real kick out of that. Obviously we’ve got our, I don’t know, kind of core audience of tech people probably came out of the ink and switch days product designers, engineers, computer scientists, product managers.

But then additionally we have all kinds of interesting things. We have a few restaurant owners, this was an interesting one to me, but actually it’s great because you’ve got the bring in the menu as a PDF and you can kind of mark it up and then you’ve got some photos of the food and then you got like some screenshots of reviews of your restaurant and some competitors and there’s a map and there’s, you know, you kind of mark all that up. Another one I think you uh recently did a user interview with was a pilot, and these folks often take the, what is it like an iPad mini or something and strap it to the knee.

00:18:36 - Speaker 2: Yeah, it’s an iPad, so they put some Velcro on their thigh and then put some Velcro on the back of the iPad and stick it on their thigh for the flight so it doesn’t move around when you hit turbulence. And apparently iPad is the sort of universal standard for private pilots because that’s where the best private piloting software is, I guess. And then from that and from the weather reports and from the radio during your flight, you’re collecting all this information and it’s PDFs and images and handwritten notes, and it’s nice to have a way to bring that all together in one spot.

00:19:02 - Speaker 1: Also seen some attorneys using it for casework.

There’s one fellow that, or I should say one person who periodically sends in screenshots usually connected to bug reports or whatever, but turned out as a like a board game designer, so there’s really interesting card artwork, very visually rich, you might say.

And then we also have a number of authors, but a really interesting case that I got to meet through our sports channel last week is a really prolific fiction author who has something like 50 published works and told a really nice story of using Muse to essentially work out a thorny point in one of their plots that they were trying to figure out for a piece.

So it sounds like there’s going to be a book published soon that Muse at least had a small role in helping with the, of course not the writing of it, but the thinking of it, which ultimately is our place in the creative tool chain.

So yeah, I would love to take the time to secure all the permission and get some screenshots and maybe get some photos and some quotes or something and just have a nice little gallery on our website, but haven’t quite gotten to that yet.

00:20:02 - Speaker 2: So our next question is from Edward Lee. He says, you guys clearly have a mission and you’re building towards that, but there’s also the reality of making sure you’re delivering a unique value prop over other products. So, and to paraphrase here, who do you think of as Muse’s competition, and how do you reason about that?

00:20:20 - Speaker 1: Yeah, I think competition is really important for understanding how product fits into the marketplace, and certainly when folks come to me for, I don’t know, just input or advice or whatever on their products or their startups, one of the first questions I like to ask is competition because I think that helps clarify things.

For Muse, I would roughly group three categories. One is the, let’s call them other multimedia canvases, mixed media canvases. That’s Miro is probably the biggest player there, but Millaote is one I like a lot. Miro and Millaotte are both kind of web-based real-time collaborative. They don’t have the iPad component. And then one actually that our guests on the last episode mentioned using is Skel, which is this kind of targeted more at authors, but again it’s like a cards on a board type of thing that works on the desktop app.

The second category I think of is this kind of broad knowledge workspace, and probably notions that maybe the hot up and comer there. Evernote actually is one that’s really well established. We hear about that one quite a bit. And then you have more specialty ones like liquid text, which is kind of focused on PDF annotation and excerpting, or actually one I’ve taken also a lot of inspiration one which is called Devonthink, which is this pretty old school desktop software you collect up a lot of material and you use it to essentially find connections and generate new ideas. And in a way I don’t necessarily think we’re directly, actually feature wise we’re not very directly comparable, but just in terms of a digital space where people go to kinda collect a bunch of stuff together and think about it, I think the Enos and notions of the world are comparable tools. And then the third category is what I would call the iPad native digital sketchbook. So all the products I’ve mentioned so far are either web-based, maybe with an app, maybe not, but they’re really native to the web or maybe they’re native to the desktop, as opposed to iPad native. And you go iPad native, there are lots of interesting choices, but the digital sketchbooks would be good notes, I think it’s a really nice one. Notability is another one. Those are really pure, I know sketching is quite the right word for it. You can, for example, Goodotes, I think is a great piece of software. It’s one I used prior to Muse, I think you did as well, but it’s really, it’s really about laying down ink on a page. It is a digital transliteration of a sketchbook. They have a lot of nice inc options. It’s good for making sketches and exploring things and writing out stuff. Can add text, but it’s clumsy, you can’t really add links. I think you can put an image in there, but that’s just not really what it’s made for. But in terms of being fast, private, and in particular that tablet posture and stylus use, you know, they’re very inspiring. So I think Muse tries to take some elements of each of those three, the mixed media canvas, kind of classic knowledge. Workspace and the iPad native sketchbook, but it also doesn’t have a lot of those features. And so for many folks when they come in and they say, well, I’m thinking of switching from any one of these products I’ve just named to Muse, but here’s the features that I need. And in some cases we have made deliberate choices to leave out, for example, really sophisticated inking options because we just don’t want to be a sketchbook as an example.

00:23:25 - Speaker 2: Yeah, and I think actually a potential 4th category would be straight up paper notebooks.

00:23:30 - Speaker 1: Ah yeah, of course.

00:23:32 - Speaker 2: Yeah, we’ve seen in our user research that a surprising number of people still swear by paper notebooks. They’re very reliable. They have low latency. They never run out of power. There’s lots of great attributes of them, and so people keep going back to them, and in many ways we’re trying to do with Muse is bring iPads up to the level of the humble paper notebook.

00:23:52 - Speaker 1: Yeah, great point. And actually if you further expand that to be all analog ideation products, sketchbooks, but also whiteboards, Post-it notes, index cards, if you count all of those, I’m pretty sure that, you know, if you just pick a random creative professional and ask them what they use to think through their ideas. is far more of them. I think that category that you just named, that kind of analog thing to would dwarf all of the others, right? That’s the real incumbent that I think we’re looking at is whether digital has something to offer against these very tried and true and rightfully beloved products for thinking. Yep. So our next question is from Brian Zimdler. This is about how we understand user needs. They ask, how did you conduct the user interview process for Muse?

00:24:38 - Speaker 2: Yeah, well, I feel like this has evolved over time, so maybe we can work backwards. So now we have quite a few users, many of whom are very active, and so we’re able to engage with them.

Sometimes they just come to us and volunteer, sometimes they have support requests or feature ideas and engage them on that basis. And we say, OK, like that’s an interesting idea, you tell me more about your use case, maybe we can get on the phone and chat. That’s. Avenue that we have.

And then going back in time before we had products, we would have to be more proactive about finding these and there we would try to identify the archetypes, the types of people that we would want to learn from and reach out to them. And often these people in our network or people who we had some connection to, so there was some level of trust already, and we would ask them about their creative process, what tools they use, things that they were happy about and not, and so on.

00:25:21 - Speaker 1: Yeah, I really enjoyed the pure ethnographic research that we did in those days.

It’s tougher when you don’t have a product because it’s kind of you’re sort of asking someone to tell you about how they work just because, and you can say I’m associated with this research institute or, you know, we’re doing the survey, but it’s a little bit less tangible.

If someone’s using your product or wants to use their product, they’re trying it out, and particularly as you said, this support becomes in a way a kind of a lightweight, steady way to get a sense of, you know, what people need and want.

And you make a great point there as well, which is folks often write in very reasonably with a kind of transactional question, which is, oh, I would like feature X, and we almost always answer with something like, OK, well that’s interesting, and we might say something like, yeah, lots of people have asked so that we’ll put you on your list. We’ve got a little database basically where we just kind of tag people with what they’re interested in, but what we really want to know is why, like what’s the reasoning? Why are you driving that in the product manager lingo is use case, which I sometimes.

But it may be for normal people it’s just how are you using this? Why do you want it? And so I try to dig in on that, like what kind of projects are you doing and what drives you to need this and why are there other ways you could do it not suitable and the real good information is in that.

But to my surprise, sometimes folks are a little bit, not that they’re withholding but more I think they maybe just assume that we’re most interested in talking about our product. Which is actually a very reasonable assumption, but I’m much more interested in you. Like, what do you do? What is your creative process, what are you trying to accomplish? How does this tool fit into your life or not, you know, how do you aspire for it to fit in, but maybe you can’t do that today. That is much more what I’m interested in.

00:26:59 - Speaker 2: Yeah, especially with a product like Muse, because people are doing so much wild stuff with it. It’s not like, for example, uh Photoshop, it’s like people are editing images and there’s some variants on that, right? But with Muse, there’s so many projects that people are working on in so many different domains, it’s super interesting to learn about what people are up to.

00:27:14 - Speaker 1: Absolutely, and maybe it helps that maybe all of us and certainly me are speaking for myself. I just have a real passion for creative processes generally. One of the reasons I enjoyed working on Hiroku was because it was very much about the developer creative process, the software engineer creative process, and we’re getting a glimpse into a different side of that through working on news.

00:27:34 - Speaker 2: Yeah, so speaking of creative process, we have a question here from Tim Lombardo. How do you recognize good ideas to work on?

00:27:42 - Speaker 1: Yeah, I think this was a follow up to episode 12 of our podcast with Andy Matzek. We’re talking about idea development, and I had spoken about how I use for me there’s a sort of ritual to starting a new muse board and particularly giving it a name, where I’m saying there is a problem that I’m facing in work or life or just an idea I want to develop that I think is interesting enough that I want to make this blank canvas for it.

And yeah, I think Tim was thinking of this as, is this a meta skill to develop, which is knowing which ideas to chase, which is really interesting. I hadn’t thought about it that way before. I certainly have a strong hunch for what things to go after, which I think of partially as like an entrepreneurial drive. Maybe it’s something that is Sort of an effective experience when you’ve been doing maker work for a long time or for a multi-decade career, you have a good, particularly if you do proper retrospectives, hopefully you have a good sense of which ideas panned out and which didn’t, and then you have a good idea of which ones to invest in.

But I don’t know, I want to think about that one more. Mark, what’s your answer to that?

00:28:50 - Speaker 2: Yeah, so one angle for me is getting a lot of different reference points and potential connections with new ideas that come in. And so that’s about having a lot of raw material to work with, raw intellectual material. So a lot of reading, a lot of time discussing ideas with other people, so that when a new idea comes in, you can kind of reference it against all these other. I ideas you have in your corpus, that’s gonna happen with things that don’t even seem directly related. That’s why I think reading about science and history and the academic literature and the industrial literature, those are all really important to do, and it builds up over time.

00:29:22 - Speaker 1: So curiosity is upstream of having good ideas.

00:29:27 - Speaker 2: Yeah, exactly. Another thing is I lean quite a bit on people who I think have really good radars or radars that are in tune with mine.

For example, Adam is one of these people who just seems to have a really good intuition.

Well, thank you. Sometimes he comes to me and says, yeah, you know, Mark, I have a feeling we should, I don’t know, start a podcast, let’s try it. Oh, OK. I didn’t exactly thought of that, but I trust you guys, let’s do it. And there are a few other people who I have a similar level of trust and respect with. Actually, Twitter has been a pretty important vehicle for this.

There’s some people that you can find and follow there and Again, it’s not about credentials or popularity, it’s more about individual people who have a specific articulated way of thinking, and on the basis of that you can come to more heavily weigh their input.

00:30:08 - Speaker 1: I’m actually reminded of a talk I saw many years ago by Paul Bhey, who’s the creator of Gmail, and this of course is in thinking in terms of product opportunities or what’s a good product to build, which is obviously a small subset of what counts as good ideas but happens to be the one I tend to operate in.

And he talks about the entrepreneurial drive or in this case, the entrepreneurial drive of feeling agitated. There’s like a problem or a thing that just seems wrong or weird or annoys you and you just can’t stop thinking about it and you look for solutions, but none of them really quite seem to fit for him.

I think he was telling the story of being annoyed that All email was slow and clunky and required you to do a bunch of organizing work that you didn’t really want to do, and losing track of long threads and just the nurturing of that agitation over time turned into like a good understanding of the problem and the drive, the initiative to go and do something about it.

00:31:10 - Speaker 2: Yeah, I call this my problem. So at any one time, I have 01 P problems.

And this is things that I just can’t help thinking about. I’m thinking about, you know, why are they wrong? What are potential solutions, what are potential angles for it? It’s in the shower. It’s, you know, when I’m taking a walk, it’s when I’m talking to people, I can’t help but bring it up. And that’s something that you basically can’t really control directly. It’s a function of your environment and what you’re reading and so on. But when I get that, and especially when it stays the problem for a long time, I’m really inclined to go work on it because I have a lot of energy for it.

00:31:41 - Speaker 1: Yeah, one that I had a front row seat of was what eventually became your article on slow software or the article titled Slow Software, which is a look at why software is slow, why it’s important for it to be fast, a bunch of the human factors. But yeah, I remember when you were. that and a lot of it came in the form of just being constantly irritated and wanting to seek the base of that irritation, and it would, as you said it, because it’s the one thing you’re thinking about, it pops out in other places. I remember you having internal memos, you know, within in and Switch that were about maybe like a project and it was sort of related, but there would be several paragraphs that was basically ranting about like it’s got to be high performance. And it sort of, oh, OK, and like eventually that just grew and grew and grew and eventually took shape as this excellent essay that then was widely shared. So this next question is about building news with MUS. So this, unfortunately I couldn’t find the original tweet. This was from around the time of our launch back in late August, early September. There was a real fire hose going on at the time. We were just getting hundreds of messages through all these different channels Twitter, hacker news, support. Yeah, the adrenaline was running high, but I remember I thought it was a really interesting question, and I said, oh, I should actually do a whole Twitter thread on this. I should post some screenshots and talk about it. And then I don’t know, there’s a million things going on, so I just lost track of it and I couldn’t even find the tweet looking back. So apologize, we can’t to the author that we can’t give you credit.

Yeah, I think this is a great one or sparked my imagination because the answer is, yeah, absolutely. I think that is just so obvious. So much of the work we are doing and in general with product and business work, so much of it is developing ideas and that precisely is what use is for. And to me it was a real milestone in the business for me at least, this piece of software crossed from a research prototype to a product with legs in the real world was when I started to be able to use it to do ideation on our business. I remember one of the first real kind of major boards I had was our first team summit in Dublin there back a year and a half ago, and you know, we had the different sessions, we’re gonna talk about the structure of the business. I had one board for some sightseeing we were going to do there in Ireland, and another one that was kind of just some of the scheduling stuff and some others that were sort of like freeform conversations that we were having that collection of things was really um helped me. Absorb and work through everything we were talking about there and in fact I think of this as being self hosting a concept from computer science which is if you’re working on a piece of software, let’s say you’re working on a programming editor, there’s a point at which in the very beginning it’s not good enough to work on its own code base. But at some point, you know, I assume that the eye and sublime text and VS code and so on, the authors writing that editor also are editing the future version of the editor in the current version of the editor. That sort of self-hosting thing is a good sign, a similar thing for programming languages and. And so on. So in that sense, the moment for me it was a big moment when news became for me self-hosting in the sense of the ideas and product roadmap and design work and so on. At least that I was contributing to the business, I was doing a lot of that in Muse itself.

00:34:57 - Speaker 2: Yeah, I feel like as Muse gets more capable, we’re able to move more and more of our work into it, which is a great sign.

And I also do use Muses a lot for, let’s say designingm. A lot of work that I’m doing these days is product architecture.

So that’s like, what should the pieces be and how they fit together? What should they be called, what are the nouns and verbs, what are the core primitives, how do you power the core engine and stuff like that.

And that’s a very contemplative activity. And so it’s really important that you have an environment that feels conducive to that.

So I like to take my muse and go on my couch and scribble some stuff and maybe do some reading, and over a course of a few days, maybe develop an idea.

This actually leads nicely into questions of roadmap. So, the first one is from Nicole Carrasano, and it asks, I was wondering if there’s a Mac version or if you intend to make one.

00:35:49 - Speaker 1: Yeah, so this is a bigger rabbit hole than you might think.

The short answer is yes, for sure, we need to do that for a bunch of reasons.

The longer answer is more complex because we think it’s long been our hypothesis for why the iPad and tablets generally are underinvested in platforms because very often you have transliteration of either phone or desktop apps to the iPad and a mouse and a keyboard or a trackpad and a keyboard are just not the same as a touchscreen and a stylus. And so then it’s basically worse than what it would. on the computer and then in the meantime you’re not taking advantages of the unique benefits of touch and we have the same thing bringing it the other way. If you port an iPad native app to a desktop environment with that keyboard and mouse, it’s going to be worse. And in fact we’ve prototyped that out and seen that there’s a lot of ways that the basic metaphor of the cards on the board will work fine with a mouse, for example, but there’s lots of other things that really will need some rethinking.

00:36:44 - Speaker 2: Yep, I do think it’s gonna be really important to get news on the desktop.

We’ve said many times in this podcast that the creative process naturally encompasses a range of devices.

You have phones for on the go, look up reading and capturing, and you have tablets, which is our current focus for reading, annotating, contemplating, sketching, brainstorming, thinking, basically, and then you have the desk. To which is the power tool for authoring and editing. And it’s very helpful to have your thinking work next to your authoring and editing work, and to be able to go back and forth, as well as to do research sessions on your desktop as input into your thinking. I think it’s very important that we get onto the desktop, but it’s not going to be, like Adam was saying, it’s not going to be the same app transliterated. So there’s quite a bit of design and technical work that we need to do to get there.

00:37:31 - Speaker 1: That also points to the other big challenge, which is I think getting a version of Muse running on, for example, a Mac or any desktop computer that has some changes to the interface that will make it feel natural with that set up.

I think that’s kind of a moderate amount of work. The bigger thing there is exactly what you described, which is if you’re moving back and forth between desk. To for one kind of setting and tablet for another, well, that implies syncing and good cross device syncing, uh, in particular that has some of the privacy qualities we’re hoping to be able to achieve is a very, very big technical challenge. Happily, it’s one we’re working on right now and we’re optimistic that it is an achievable end for us. So the next question again is I’m going to do more of a a paraphrasing of something we hear very, very often, which is basically just what’s the muse product roadmap, and I think some additional color on this question or a lot of the motivation is I’ve really come to realize how important it is for people to know not just what the product is and what it can do for them today, but also where it’s going, and that makes a lot of sense in the context. of subscription software, you’re not really buying it for today, you’re buying it for the, you know, in this case, if you take an annual subscription, even if you don’t renew, that’s 12 months' worth of improvements that you’re expecting to accumulate, and that gets even more so if you get into something we’ve talked about in some past episodes, which is in many ways supporting an early product like this. It’s almost more like a Kickstarter you’re saying. I want software like this to exist. I think this team is the right one to do it. I want to support you financially to make that come true, but then that also has the additional dimension of you want to know and trust the team, but you also just want to know where is it going. Maybe the places they want to go don’t match up with what you think is interesting or inspiring. So we don’t have any kind of public roadmap, but this certainly makes a lot of sense to me that people ask about this.

00:39:23 - Speaker 2: Yeah, maybe we can describe this as our areas of interest in areas that we’re being pulled and drawn to both by our vision for the product as well as quite consistent customer feedback and questions. There’s a few areas that constantly come up that people want more in.

00:39:37 - Speaker 1: Yeah, so I think really zoomed out. I think of the kind of three big tiers or sequential areas to work on is starting with the single user thinking tool.

And that’s the mixeded canvas, use both hands in the stylus, it’s super fast. It uses all the screen for your content, really fully taking advantage of the iPad and stylus hardware. And actually this launch that we did a few months back to me was the feeling like we delivered that. There’s much, much more to be done there, both improvements on what’s there and more features, a very long list on that, but I feel like that pillar is well in place.

Then the phase two, if you want to call it that, is what we’re working on now, and I expect we’ll take most of this year we’re just now entering, and that’s the multi-user, multi-device stuff. So that is, I can access it from all my devices, the new experience in each of those devices is appropriate to that device, phone, desktop. And so on.

And then of course the multi-user collaboration, sharing, how do I send stuff to clients, how do I send stuff to colleagues, how do I share in a way with people who are not using news or people who might be using Muse, we can get additional benefits from that.

Huge technical and design challenges there, but certainly a very fun space that we’re starting to get some progress on.

And then the final piece, uh, to me is what I would call the end user programming or the programmability of it. I see we have a question further down the list about that one specifically, so maybe we’ll expand on that there, but that’s just my take. I’m curious how you think of the overall roadmap.

00:41:12 - Speaker 2: Yeah, that’s pretty much how I think about it too, and maybe you just drill down a little bit into the 1st and 2nd phases there.

On the first phase, I agree that we have a quite solid foundation with this tablet native thinking experience. We’ve gotten very good feedback on that. Now there is some more to do, just to give you a couple examples. One is, well, we think it’s important to be able to move ink across boards as well as move it within a board, which you can currently do, and that’s quite a bit of a technical and design challenge.

Another one that I’m excited about is a non-spatial collection type. This goes back a little bit to our zooming UI conversation. Right now, the only way you can collect elements in muse is on a spatial board, and spatial collections make sense in a lot of cases, but I would say not all cases. And so I think we want something more like a set type that automatically manages the position and ordering and sizing and so on of your elements. So if you just have, say, a stack of papers that you’re reading, you can do something that’s more like putting a paper on the stack as opposed to specifically laying it out somewhere on your desk. And then phase two, yes, it’s a simple matter of making it run on everyone’s devices everywhere. There’s obviously a lot under that hood. There’s a big technical challenge in going from all the data resides canonically on one device to it’s distributed across all devices. But once you unlock that, you can do a lot with syncing across one user’s device as well as collaborating across devices. So I’m quite excited for us to be pursuing that. And then we also talked about how you need basically 3 different apps, you need one for the phone, one for the tablet, and one for the desktop. So there’s quite a bit of work to do there.

00:42:41 - Speaker 1: And certainly on the product or design side, I think that while there’s a really good precedent for collaborative tools, including GitHub for code, Figma for design, Google Docs for writing, There really isn’t a collaborative thinking tool and you know the sort of obvious thing to reach for is the live whiteboard.

And in fact, in many cases folks find us or imagine that that’s what we’ll be sort of like standing in front of a white board with your colleagues talking through a problem and scribbling out visual support and indeed we may support a use case very similar to that. But I suspect that there’s a lot of nuance and interesting subcas or related cases when you talk about asynchronous work and the degree to which you want to do things in private and then share them with your colleagues for kind of input or to further discuss versus like starting together with a blank slate. I think there’s a lot there.

00:43:36 - Speaker 2: Yeah, the way that people actually work together creatively or want to work together creatively is very complex.

They want privacy of their information, they want to do a lot of stuff privately to themselves, and then they have a bunch of different granular collaborations.

You know, you and I are collaborating on this podcast. We are collaborating together with the new partners on the business. We also have all kinds of other stuff going on in our lives and it’s kind of formed these overlapping groups and correctly representing that and supporting that in software is a really important goal of mine for this phase two.

And that brings us to what you described Adam as phase 3 of Muse. Jimmy Miller has a question here. He says, I’d love to hear you all talk about end user programming, explorations that bring Muse a bit closer in that specific direction. What is the place that you see for end user programming in Tools for Thought and use?

00:44:24 - Speaker 1: So this is a huge passion of mine, career long passion. I’ve written articles about it reaching.

Back 10 years and more I was one of the foundational inspirations for Hiroku and just to kind of briefly define and user programming, the idea there is that most software today is written by professional software engineers. It’s not something that basically computer code and the programming tools and everything you need to write software is something that you really have to be an expert at in order to do, not even just at a high level but just at all.

And the idea of making that more accessible, such that maybe not every single person that ever uses a computing device, but many more people could have access to that, and the comparison on that is often made to literacy. So once upon a time, reading and writing was something that was a tiny tiny elite. of scribes and alerted people could do and later on through the printing press and education and some other means became something that almost everyone can do.

Not everyone’s a professional author or really strong on English grammar or whatever, but almost everyone can write a shopping list or write a letter or an email to their friend. And so there’s a version of this for programming as well. How do you think about end user programming as related to Muse?

00:45:46 - Speaker 2: So I suspect for end user programming as we typically understand it, that is users writing code like stuff, I suspect you’re going to have the clearest vision for that, and you’ve thought the most about it. I have a couple different angles on this, which I’ll share.

One is, I’m quite excited about the idea of the programmability of data.

So we often think about programming as you have codes manipulating like in memory variables and stuff like that. I’m really excited about the Users having programmatic access to their corpus, read and write, and that being reflected in muse and outside. So you could potentially think of your end user program as you’re writing these little like bots basically that are watching and crawling and manipulating your corpus data and doing automatic things with it. It might be organizing your daily to do list or something like that, or something more complex, doing summary statistics over all of your work for the past year. And I think there’s something very powerful as well as accessible about accessing the underlying data. And that leads kind of into my next angle here, which is open ended computing, which is maybe a generalization of end user programming, and by that I mean users being able to do combinations of things that the designers of the program didn’t specifically need to plan for. So with typical mobile apps in particular, the set of things you can do is basically the set that the product manager said you should be able to do. And if you want to do something that the product manager didn’t think of or doesn’t agree with, well, too bad, delete the app or something, you’re out of luck. This is in contrast to the world of traditional desktop computing where you have the standard file formats that any program can access and manipulate and that the user can directly access via their visible file system. So to take a simple example, if you’re writing a text note and you don’t like the color of the text screen, you can just open up a different editor or you can CP it into a different. There’s so many things you can do and you have so much agency because the files and the programs have this interface that’s kind of general and that different people can access. And that’s something that I think we’ve kind of quietly lost in the world of mobile apps. You have a little bit of it with the share sheets that export standard-ish file formats, but it’s very clumsy and it’s quite limited and oftentimes the programs just don’t let you export stuff easily. So what I’m curious about is how do you bring that desktop sensibility of open-ended end user computing back to the world of mobile apps.

00:48:05 - Speaker 1: I think this point of data being the center of gravity or the thing that matters to users and customers, I think as software creators, we tend to think that the code or the running software is sort of the thing that matters and then data is this kind of thing that’s added.

Onto that like a database is something you attach to an application or you write the software and opens and it reads and writes to a file, for example, but from user perspectives, the data is what they care about.

It’s like I’m using your, I don’t know what piece of software to write my master thesis, you know, they may like the software or not, but ultimately they care about their work, they care about the thing they’re creating. That’s more important to them. And yeah, standard data file formats that I think once upon a time were much more kind of universal and easier to use, and those have sort of faded in utility largely because of the mobile world and finding ways to not necessarily step back to a previous time to get some of that value of I don’t know openness is quite the right word for it, but the ability to manipulate your data in a powerful and free way. I think that’s something missing.

00:49:11 - Speaker 2: Yeah, maybe agency is another way to describe it, yeah, yeah.

00:49:15 - Speaker 2: We also talked about this in our local first paper which we can link to, but yeah, the world of files gave you all this control and visibility and portability, but then they basically Fall on their face when you get into the world of mobile and especially collaborative apps. And then with mobile and collaborative apps, we kind of went all the way the other way and said, now your data is totally opaque and maybe you can’t even get it out, but at least you know, you can see each other in the Google Doc. We really need a third way, and that’s something we’re excited to pursue with Muse.

00:49:42 - Speaker 1: Yeah, this is one reason we made it a priority from the beginning to these Muse bundles that you can export are just zip archives that you can open and they contain flat files and a little bit of metadata. That was something, but that’s really just a very small step in that direction. So I’d love to see us do much more, and I agree that fits together with the the end user programming is ultimately about agency and grasping the full power of computing rather than being kind of at the mercy of the software vendor.

00:50:09 - Speaker 2: Yeah. Another thought I have on end user programming from looking at it in the wild, if you will, is that what’s really important is that people really care about the thing that they’re programming. And then they just need some little window or way in.

So some examples of this one is Minecraft, where you have these Redstone contraptions, and it’s basically a very primitive way to program and it’s pretty gnarly and it’s literally quite bulky and blocky. But people are really motivated to build all kinds of wild stuff in Minecraft.

Another example, you know back in the MySpace days people customizing their pages, and even I would say JavaScript and the web, like, frankly, JavaScript is a pretty gnarly programming environment. There’s a lot of things wrong with it, but People so much want to be able to script their content on the web. They basically found a way to make it work.

And now you have these incredible JavaScript VMs which have wild performance.

So a lot of the work I think of end user programming is not, how do you make like some editor that people can think about clearly. It’s more like having something that they want to program and then giving them a little bit of a window in. And that’s why I think this idea of the users' content and their corpus is so interesting because people care a lot about their creative work and everything they put into their music corpus potentially. And so giving them a way to unlock that could be really powerful.

00:51:17 - Speaker 1: I agree and I think this is one of the challenges of a lot of the HCI research that looks at end user programming or programming environments that are more accessible, is they tend to start with the programming environment. But almost by definition, the people who are not programming today are people who aren’t interested in programming for its own sake, right? They’re interested in it because of what it can do for them. So that means they already have a thing they care about that they want to automate, they want to extend, and so that is one of a few reasons, as a very natural reason why anything in user programming related would come late in the roadmap relative to other things.

Yeah, one little teaser you could get on what that might look like is for the capstone project we did an Ink & Switch and I’ll link the end user programming article we did there, but we had one of our colleagues did essentially a series of experiments with, as you mentioned earlier, the bots. Basically you could have cards that had a little script attached to them that you would write yourself and if you drop the card onto a board, it would then Sort of do its actions. So we had a couple of examples that were something like a sorting card you could drop that on and it would sort of arrange everything in a grid. And there was another one that I think was a journaling card that would essentially put like sort of a title with the current date and time that was intended for a board you’re using kind of as a journal. And those are simple examples, but they showed the kind of stuff that a person might like to do with like a personal knowledge tool or thinking tool, and those were very, very rough and very early, but I think that kind of gives the direction of that.

And furthermore, that also raises this question of end user programming can be sort of automating stuff for your own sake, but it can also be that you make a small plug-in or script or something you want to give to a friend or a couple of friends. I think a good example of this we’ve seen recently is Figma plugins, where people write plugins for themselves, but they may also distribute them a little. More widely, it’s not really intended to be sort of a full-fledged sophisticated application. It just extends or changes or customizes this environment that you’re doing creative work in in some way that makes it suit your needs and maybe you want to share with a few other people. So I’d imagine some kind of set up like that for use.

00:53:28 - Speaker 2: Yeah, this makes phase 2 and 3, if you will, very complimentary. An important learning from our ethnographic studies of end user programming and the other stuff that we’ve read is that there’s very little original end user programming. It’s mostly stuff you learned from a friend and then mutated slightly. So there’s this kind of one. Original spreadsheet and then it basically propagated through the world as people made different variants of it and copy pasted cells and so on. And I think if you have native collaboration and sharing in use, that could be the vector by which the end user programming learning gets propagated through the system.

00:54:00 - Speaker 1: Steal a cool bought card off of someone else’s board by duplicating it into your own.

Exactly. Love it. Well, I can see that this is a topic that is one we could easily spend a full episode on, so maybe we should wrap it there and maybe we’ll put that on the schedule for the future.

Right on.

Well, if any of our listeners out there have feedback, either responses to what we’ve talked about here, or maybe you think some of these questions could be full episodes in their own right, feel free to reach out to us.

We’re at @museapphq on Twitter or hello at museapp.com on email. We always love to hear your comments and ideas for future episodes.

Well, Mark, it’s been a real pleasure doing this podcast with you for the last 21 episodes. Uh, I imagine maybe we could do another 21 more. I hope so.

00:54:46 - Speaker 2: I’m looking forward to it.

00:54:47 - Speaker 1: All right, see you next time. See you.