Transcript

00:00:00 - Speaker 1: You start with some seeds of an idea. Basically, it might be some sketches or a picture of a whiteboard you took or a voice memo, and what you want to end up with is, say, an essay, and there’s several steps to the creative process. And one of the things that’s exciting about the new text feature plus blocks is you can see it as a trellis for that growth.

00:00:23 - Speaker 2: Hello and welcome to Meta Muse. Muse is a tool for thought on iPad.

But 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 with Mark McGranaghan. Hey, Adam. Mark, I’m pleased to report that Metause has broken into the top 200 charts on the technology category in Apple Podcasts. It’s a per country breakdown. I’ve been using a little thing called PO status that essentially sort of charts your position over time. Some countries were there pretty consistently, other places like Germany where I live, we kind of pop in and out at the whims of the algorithm essentially. That was quite surprising to me in a lot of ways, cause I just still think of this as, you know, me and you were having a chat sometimes with guests, just people we like to hang out with and seeing our logo alongside these what I consider to be kind of giants of the podcasting world like Cortex and Accidental tech and so on is kind of a thrill actually.

00:01:20 - Speaker 1: Yeah, I continued to be really pleasantly surprised by the reception we get to the podcast. It’s actually just at a family event about a week ago, and people would come to me and say, Mark, hey, it’s great to see you. By the way, I love the podcast, like, whoa, OK, I didn’t know you were listening to that, but that’s cool. So yeah, it’s been fun.

00:01:35 - Speaker 2: Nice. Yeah, and I also want to maybe make a little request. First of all, a huge thanks to all the people who have tweet recommendations or a lot of folks tell me that they do more kind of in person. Reminds me a little bit of our episode on social media where we talked about something going viral slowly kind of through word of mouth, sort of the ideal thing, and I think there’s a little bit of that here, which is great.

But actually, if you haven’t had the chance to recommend us, you can actually help by leaving a review on Apple Podcasts. They make it a little hard to do, but if you go to the new podcast page and go to the Apple Podcasts, I’ll link that in the show notes. You scroll down to the bottom, I think you can tap write a review if you’re on your iPhone. I’m sure Spotify has a similar thing. We only have a few reviews. In a lot of countries, sometimes none, so even just taking a moment to drop in a star review in one sentence of what you think you like hearing these weird guys talk about, if nothing else, will soothe my vanity.

So our topic today is text. Now that word even is so rich for me and many of the reasons I got into computing and tools for thought, and so on. The impetus here is we’re just now releasing into beta for all our pro members a text blocks feature, so essentially changing our text cards today, which are kind of these Post-it notes things, pretty basic, to something that is a little more inspired by the notion Rome craft world of things. And maybe we’ll describe a little more of that vision later, but of course I always like to start with the absolute fundamentals. So Mark, I have to ask you, what is text, and I mean not the dictionary definition, but what comes to mind for you with that word.

00:03:11 - Speaker 1: Well I’ll give you a very marked philosophical answer, which I’m sure we’ll hear echoes of in the rest of this conversation.

Now, if you think about conveying information, there’s sort of a necessarily most primitive form, which is a string of 0 and 1’s, you know, you can’t reduce the dimensionality beyond a line and you can’t reduce the base beyond two, right, or else you have no information.

And then in the case of human acceptable information, it’s perhaps a string of human readable characters. So in some sense it’s the most basic fundamental primitive way to communicate information. So that’s one of the reasons why I think it comes up so often in Tools for Thought, but we’ll talk more about that throughout the podcast.

00:03:51 - Speaker 2: For me, the word text, I think, makes me think of plain text or files that end in .txt and for a very long time, that was my whole knowledge management system was a folder full of text files.

Maybe at some point I did mark down or something like that, but plain text is just one of the most fundamental formats on a computer. It’s how code is usually represented, it’s a very durable and long-term format, it’s very flexible, you can do Aski art and things like that, yeah.

But I guess going back even before sort of the digital side of things, I really think of writing things down in any way at all as the original tool for thought.

And in fact, it feels like almost all of the things that build upon that are essentially variations of ways to write things down.

Ways to externalize thoughts from your mind. I think we’ve talked before about even something as simple as using a stick to draw on the dirt and I don’t know, cavemen drawing a picture of a horse on the wall of their cave, and certainly you have this whole history of, I guess there’s sort of written language, which of course is an extension of or a mapping of spoken language.

And that leads you into the whole world of alphabets, but actually even before alphabets, you have logograms, things like hieroglyphs, you know, you have a picture of a duck, and that means the word duck, for example, and then you have all these technologies for mediums for writing on, mediums for writing with, for reproducing those things, clay tablets, styluses, papyrus, paper, pens and pencils, printing press, whiteboards, posted notes, etc.

And then language, which we typically represent in modern times with alphabets which stand for sounds roughly, but are actually very abstract, you know, they’re pretty far removed from the pictures or diagrams we once had, and I think that leads us to one of the dualities I wanted to talk about or I’ve been thinking about a lot in terms of this muse product direction, which is the duality or the spectrum of symbolic versus spatial and visual.

00:05:46 - Speaker 1: Yeah, this could that something really fundamental, which is formats that are optimal for conveyance and translation and reproduction and storage, which I would say plain text, especially for that category versus formats which are optimized for matching how our minds work and think, which is closer, I would say to the music model of it’s multimedia, it’s free form, it’s kind of messy, and so.

Forth.

And so there’s constantly a tension, I think, between having a tool that better represents and better works with how we tend to think versus having a tool that, for example, can persist that data over hundreds of thousands of years, or just a few years in the case of today’s software.

And I think kind of grappling with that, not to mention just the complexity of actually building such a multimedia canvas is a lot of what the tools for Though space is about.

00:06:35 - Speaker 2: Yeah, I would argue that at least the current tools for thought space, which is maybe a little bit overindexed on building sort of Rome clones and variations, and I think there’s probably much wider space to explore.

I’m certain there’s much wider space to explore, but those, of course, yeah, Rome notion, they’re in the same vein of, yeah, plain text files, marked down, Emacs org mode.

Very symbolic oriented and of course symbolic representation, yeah, mathematics, written language, of course, even programming is just an incredibly powerful way to do things, but this spatial and visual side, I mean, we talked about this with Anne Lohr back in the episode on thinking in maps and sort of diagrams and literal maps in some cases, as being a spatial and visual way to represent things. Or there’s something like data visualization.

Of course we can’t go in an episode without mentioning Brett Victor, and I was just rewatching some of the humane representation of thought, where he talks about the invention of the modern chart or data visualization. I think it’s in the, maybe the 1700s, 1800s by William Playfair.

And this idea of creating a chart where you got time on one axis, and some thing that goes up and down, like money or population or some other thing on another axis, turned out to be a really powerful way to tap into our spatial reasoning and a much better way to get overviews and see patterns in data, but that would be invented just like everything. So that’s an amazing tool for thought, I think, an example of in that sort of spatial and visual side that we think is Maybe under explored in the digital realm or right at this moment in the digital realm.

00:08:16 - Speaker 1: Right? And I also think there’s an element of time and process here.

So there are some use cases where you want a very visual and spatial end product. A map is perhaps the canonical example of that, but I think much more common is a case where the process along some of the Steps asks for such a format.

So for example, if you’re eventually going to write an essay, the final artifact is going to be plain text, essentially. But I find at least it’s quite hard to start ideaating and brainstorming and sketching an essay, like basically in a text editor, you know, maybe I’m going for a walk or I want to be giving myself a voice memo or I’m sketching some ideas in my notebook, or I’m drawing some diagrams, right? And so you have this process where often I find in the beginning stages of ideation, brainstorming, sketching, outlining, you can really benefit from this freeform spatial multimedia model. But then you have the issue of if that’s step one, but the final step is plain text, how do you navigate that jump basically? Do you jump tools? Do you have some kind of conversion step? I think often what people do or often what people have done in the past is they just kind of punt on it. And they find a tool that’s like flexible-ish enough to do some of the brainstorming and like presentable enough to do some final publishing. I think notion is actually a really effective example of that where it’s really nice to do the whole process in one tool, even if it’s not ideal for either end. Or sometimes people, I think they jump tools like they do some sketching in the notebook and they have an outline for their essay, and then they go type it up, they basically got to retype up all their notes. So I think there’s an interesting potential for tools that allow a more gradual and continuous process where the shape of the content evolves from a messy sketch to a typed up essay, for example.

00:09:57 - Speaker 2: Yeah, and there may even be some value to the transcription process.

So in the coming up with what you’re going to write for a long form prose piece, let’s say you’re a journalist or you’re just writing a blog post for your own personal thing, of course, in the end it is going to be this linear one dimensional starts in the top left for left to right language readers and ends at the bottom right and flows very linearly, but when you’re figuring out what to say, maybe you draw on your whiteboard, you’s catching your notebook, you use index cards that you can move around on a table.

And maybe you find what you want to say, but then actually sort of transcribing that fresh into, I don’t know, your writing tool or whatever works.

Maybe that works OK in a lot of cases.

There are some examples in the digital realm.

I think we’ve spoken about them or linked to them before, but this company Literature and Latte makes one of the maybe best known kind of dedicated long form writer’s tool, which is Scrivener, I think it’s really intended, especially for novelists, fiction novelists. But they also have another tool called Skel, I believe it is. I’ll link that in the notes, which is kind of cards on a canvas, Post-it notes, desktop thing, maybe it’s not as sleek and modern as Muse, but I’ve run into people who use Skel in the same way you might use Muse, and notably they’re both from the same company. And they’re intended, but they’re just for those different stages, precisely like you said.

One is this ideation, you’re figuring out the arcs of your character and maybe even want notes that aren’t even going to be sort of in line to the story. There’s backstory about a particular character or a place. You’re not necessarily going to just have a paragraph where you say so.

So it was from here, they’re this old and what have you, but as a writer, having that floating around on the edges as a reminder while you’re near your other kind of plot elements can be useful. So that’s another example of the nonlinear free form, and then eventually you somehow collapse it down to this one dimensional long form prose format.

00:11:48 - Speaker 1: Yeah, I do think there’s just a ton of potential in that type of tool, and I feel like we’re only just beginning to explore it as an industry. We kind of take for granted that you’re gonna be starting with like a text buffer and you’re gonna type stuff in and that’s that. And I think it really impacts our ability to develop creative ideas, and I do believe that as we develop more tools that are more aligned with how we’re thinking, they’re more multimedia, they’re spatial, they’re free form, we will in fact have better ideas. That’s one of the big bets of news.

00:12:15 - Speaker 2: Yeah, thinking about the linear kind of top to bottom flow of text or symbolic language, I also had me thinking a bit about terminals, or reppel, sometimes they’re called.

You know, when you think of computers, text is like really foundational, and then the terminal, at least for Unix folks like us, that’s a place where we spend a lot of time, it’s a venerable, it’s sort of a way to have a conversation with the computers the way I think of that a little bit, a little bit of the gliders, man, computer symbiosis, but I always find it interesting if you dig into why does the terminal.

Work this particular weird way with a lot of control characters and whatever and then you get into this TTY thing.

What does TTY stand for? Well, that’s short for teletype and these like teletype things date back to the 1800s, their stock tickers, that sort of thing. They were essentially these ways to have again the computer or some automated device produce a linear stream of symbols.

I think later they were adapted to mainframes, maybe in the 1960s or something like that. And then even when you go to word processors, whether it’s like the WordPerfect and whatever in the 1980s or Xerox PARC and there what you see is what you get word processor, and even today Google Docs still has this quality. You start in the top left, you go top to bottom. That’s kind of it. Except, I do think a breakthrough or, and maybe this is less at the symbol level, but it is at the overall corpus level, is linked. Let’s see you had wikis first, I think the web with hypertext, you could argue is clearly the biggest and strongest example of that.

And a lot of the excitement also around tools for thought right now, notion first, you know, sort of a modern wiki in many ways, Rome with its backlinks, lots of others have focused on the linking elements of things, and so now you do get a graph of your knowledge and so.

The individual documents are still these linear streams of text, but you can kind of pop around between them and the web, I think, takes it even further and Notion I think does this reasonably well also, which is letting you put multimedia elements in images, video, that sort of thing. Although in the end I think notion is still very much inherits that sort of top to bottom, typing into a word processor kind of thing, and the web has more free form, but of course if you really decompose it, you know, you hit view source in your browser, in the end there’s a top to bottom linear document made out of characters.

00:14:36 - Speaker 1: Yeah, that’s interesting. This is clarifying for me that there’s a couple of dimensions at play here. There’s the dimension of what’s the datum type, and by datum I mean like the atom of information, which could be a text paragraph, it could be an image, and then there’s a dimension of call it like interactivity or freeformness, if that’s a word.

00:14:54 - Speaker 2: Unstructuredness maybe.

00:14:56 - Speaker 1: Yeah, and the examples that you gave are interesting because when we think of text, we often think of the straight up plain text buffer and sublime or whatever, but you gave the example of a wiki. Reppel, I would include social networks like Twitter in there.

There, the datum is still text, but it’s very rich and interactive in other ways, right? And I think that that’s compelling because there’s all these nice properties of text as a datum, but there are a lot of limitations with plain text as a pure linearization of text datums, right? And again, this kind of gets into what I think we’re trying to do with Muse, where we really like text as a datum, and I think it’s really important, perhaps the foundational one. But you don’t want to be limited to putting just text just in a line.

00:15:38 - Speaker 2: Yeah, Twitter’s a great example. I hadn’t thought of that one, where of course as text first, it’s even famous for that 140 characters, now 280 originally started with essentially sending SMS messages before they kind of adapted to the mobile client world, but I would argue that its value and richness does come not just from the text, but from the links, from the images, from the video.

And furthermore, that there is this atomic unit that can contain all of the above in a particular container, which is the tweet, which often is represented with this card, you can invent that. And actually that’s a nice note for what we’ll get onto later, which is sort of the block text concept, and that is an atomic unit, and then of course, news has its cards, so I think all of those sort of relate a bit.

So in thinking about the importance of text and what we might like to do for use again, we love text, we believe in it, we just think it’s so well handled or supported, or that’s where a lot of the interest, innovation focus has been on computing tools and we saw the iPad and it’s particularly. The pencil, which as we talked about in the iPad episode here recently, was kind of the thing that starts to make it potentially a different type of computing tool that’s unique and so we wanted to take advantage of those making another kind of text first tool on the iPads it felt not quite right.

But having done that, having invested heavily in the spatial and visual side, now we think, OK, we have these text cards, they’re pretty basic. Text is still really important, what can we do to bring that in? And in some ways I kind of draw the spectrum or something like that on digital products as you have the increasing number of I don’t know, tools for thought or something like that that again I think are very influenced by the notion and Rome side of things but are again lineage going back to Emacs org mode and work flowy and all that sort of kind of stuff tend to be text first. Yeah, you can do some multimedia, but the multimedia is in line with the text. That’s the focus. It’s on desktop computer, you’re using a keyboard, keyboard shortcuts, etc. And then you have the world of, I’ll call them digital whiteboards, which I think people often do want to categorize news as, that’s kind of what it looks like at first glance. I don’t think it’s actually quite right or that’s not at least our long term vision, but you have something like Mirro, for example, or fig jam, or even like these sketchbooks like good notes, I think you and I have both used quite a bit. That essentially allow you to do very free form stuff. It’s spatial. You can drop in images, you can sketch, maybe you can do so collaboratively, but notably I think all of those that latter category, you can put text in, but it’s not fun. It’s basically the same as putting text in photoshop. I mean, imagine trying to write a blog post in Photoshop with a text tool. It’s miserable concept.

00:18:27 - Speaker 1: It kind of reminds me of when you need to annotate PDFs for like legal forms, you gotta go like annotate, insert text, and it gives you the chunky text box, you know, you can do it, yeah.

00:18:37 - Speaker 2: Yeah, exactly. And I actually had the opportunity to speak to product people from both Miro and Fig Jam recently, and they confirmed, yeah, text is not something that’s important to us. We think it’s useful to drop in the equivalent of a Post-it note or a little title or something, but it’s just, that’s not what we’re doing.

So I think or hope that the hypothesis we’re exploring here with this beta is that bringing these two together, richer tech support. And this visual spatial sketchy environment could be something really powerful and maybe a more useful thinking tool than either of those apart.

Although I’m very conscious of the, let’s say the opposite side, which is uncanny valley, right? Sort of like not very good at editing text, but the text gets in the way of the free form stuff, that would be the downside of that, and that’s why I want to explore it through this beta. Now it’s Ben, our colleagues Yuli and Leonard, both of whom have been on the podcast before, they’re really driving the vision on this, doing some incredible work, as always, and I know you’ve been more heads down on the sinking side of things, so I think you got a chance to try the text blocks beta recently. What was your reaction as someone who is coming in a little bit cold or a little bit fresh to the idea?

00:19:52 - Speaker 1: In the most recent beta of the text blocks, I think I got a glimmer of something really special.

It’s this flow where you have a series of blocks of text, sort of like a to do list or a brainstorming list, and you can move the blocks around, and the list automatically reforms, and I really like that because before I used to do this sort of thing with the pencil because then you can do lasso select and you can move stuff around, but it’s always so. to use the pencil for everything, or I could do these lists on a desktop, but then, you know, you’re sitting at your goofy office chair and stuff, and it just doesn’t feel very creative.

But with this latest beta, I think we’re getting close to feeling like the text tools plus the blocks, plus how they are manipulated is a really natural extension of how you’re thinking, where you’re saying, oh, this idea I should move down or move up, and you can basically do that in the app. So I’m pretty excited about it.

00:20:45 - Speaker 2: Yeah, and you hinted there at um one of the core ideas, which is why we’re calling this text blocks, which is the blocks concept, I think. I don’t know where it started actually.

My first real encounter with it was notion, although I feel like the computational notebooks that I used well prior to that, like IPython and Jupiter had a model like this more for the purpose of kind of almost calls back to the reppel type environment or this top to bottom execution flow, but this idea of stacked cells, it was almost like a spreadsheet that only went in one dimension, worked surprisingly well and of course the thing that makes it a notebook. is that it’s not just code, but you can drop usually mark down cells in there and you can move those around and have the explanations in line. It works out to be a pretty natural way to work to have this structure of the top to bottom blocks, and then within that you can have, in the case of freeform pros, you can basically just type however as you would in a normal text buffer.

And so then I think notion, as again, my first real encounter with that, took that idea of essentially each paragraph is a block, and that by itself, it was just text. There’s some things that are cool about it for like reordering lines and stuff, but I think it wouldn’t be worth the hassle because it can be confusing when you switch between character select mode and block select mode.

There’s still lots of ways that something like notion, roam, or craft behave in a way that’s quite different from the Google Docs word processor, them, whatever lineage of text buffer editing, where I think it starts to excel is when you bring in other things that aren’t text.

So, images, video, links. Tables, convent boards, and I think this is part of the power of modern digital computing. You can do these multimedia documents, you can illustrate things, you can drop in screenshots, etc.

Google Docs, for example, I use this kind of an internal memo thing for a long time. You can put images in there, but it tries to make it kind of a giant character. They’re very weird to work with. It’s just it feels wrong. And having these stacked blocks where most of them are texts, but they can be other things like images and Video and so on, somehow that makes the whole thing, even though it’s still a text first environment, it feels much friendlier and more natural to this multimedia world we live in now.

00:23:02 - Speaker 1: And I think we’re also getting a line of sight on this grow a document use case that we’ve been striving for for a long time.

So this is the idea that you start with some seeds of an idea. Basically it might be some sketches or a picture of a whiteboard you took or a voice memo, and what you want to end up with is, say, an essay, and there’s several steps to the creative process. And one of the things that’s exciting about the new text feature plus blocks is you can see it as a trellis for that growth, right, where you start with Line, which maybe is initially just like one or two word snippets, or maybe it has some pictures thrown in, or maybe it has some handwritten notes thrown in. And then as you go through, you’re rearranging, but also sort of expanding each of these blocks. So your little block that says, you know, to do insert paragraph about food here, that grows into a full paragraph on the essay. And likewise, your little picture of a whiteboard of some diagram gets replaced with a nice diagram that you create, right? And that way the essay sort of organically grows and critically it also Happens in one place. So there’s no point where you need to jump from your notebook to your brainstorming tool to your authoring tool. You can sort of do it all in one place. And I don’t expect that you would do the full creative process in Muse. I think you would probably stop at like the sketch or the outline or the draft phase, you would move to an authoring tool and a publishing tool for the final step, but even the idea of growing the whole meat of the essay in Muse is really exciting to me.

00:24:24 - Speaker 2: Absolutely, and I’ve done this kind of writing workflow myself, and I’ve heard from lots of folks writing in to support or just that I talked to casually that they use Muse in this way for their own writing workflow.

Jeffrey Litt showed something like this for his newsletter, previous podcast guest.

We have a fiction author that described a similar way of working through sort of plots and character development and things.

Now I think in the current kind of text cards thing, they’re basically like Post-it notes.

And what I do like about that, and I do this in the analog world as well, I know I have a lot to say on the subject, but I just don’t know the structure. I don’t know the table of contents. I don’t know where to start. So you can just take a stack of index cards or Post-its and just write down the first thing that comes into your head, you know, the seven word version of it, pull the Post-it off the stack, set it aside, right the next thing comes into your head, and do that until you got 20 of those. Now go arrange them on a wall or something, right? And you might start to see a pattern emerge or a narrative arc or something like that. Right.

I have used Muse for that for years now with the kind of post-it cards, but it’s like you said, you stop pretty early because once I have the rough ideas roughly in the order I want, and maybe a few scribbled arrows and highlights and things like that, I go, OK, I’ve made it this far. I don’t want to do any more substantial writing. I don’t want to do any really amount of big investment in the words. It’s really more just the high level concepts, and I want to go over to my writing tool. Now with the text blocks, potentially, you can do more of the long form writing. Again, it’s not a full-fledged text editor, never intended to be, but you can at least get a lot more of the core ideas written down and then basically use your select tool, just grab everything, hit copy, pop over to a writing. Tool like Kraft Ulysses or your WordPress blog or whatever it is, paste it in and then go and do your more substantial wordsmithing. But at that point you’ve got the flow, the order, the structure of the piece, maybe some of the major phrases and opening sentences of each paragraph and that sort of thing. And now you can go and start really putting the meat on the bone.

00:26:25 - Speaker 1: Now we’ve talked about jumping to a full-fledged authoring and publishing tool for writing, which I do think you’re gonna need to do if you have an external audience, you’re gonna want to create a PDF or something similar.

But I’m saying is that often for internal communications within a company, you can get by with just using the ideation and brainstorming tool as the quote unquote publisher. So we see this a lot internally with both muse and Notion. People create muse boards or notion documents and use those as the final artifact that they’re gonna share with their teammates because yeah, they’re a little bit sketchy, but that’s fine.

We recently had a good planning session where we used a beta of the text blocks and it was really cool to see that and it’s actually nice cause it kind of correctly reflects the state of the ideas.

You know, it’s this notion that you don’t want it to seem too rough nor too polished, because the ideas are sort of in this intermediate state, which I think the muse boards with some text cards and some highlighting and some images and stuff that show which is nice.

00:27:23 - Speaker 2: Fidelity, the representation should hopefully convey the level of polish of the ideas as this uh carefully crafted plan intended to be conveyed to a wide audience versus a thing that we sketched together in an hour of a planning session. Right? It’s a really good point on the publishing side of things. And I think that any idea that you want to express, again coming back to that, writing something down as the original tool for thought, externalizing an idea, and you can start on one end of the spectrum is just you want to express it for yourself. And that’s the sketchbook, that’s what M is focused on today. It’s just, let me get this idea out of my head, explore it on the page, see if there’s legs, develop it.

But I never intend to share it, at least not in this form with anyone else.

The other extreme is I’m a journalist writing a piece for The New York Times or I’m a documentary maker and I’m going to put my thing on Netflix and it’s going to be consumed by potentially millions of people. And so there’s going to be a very high degree of polish and a whole long process going from ideation and sketches to drafts and drafts and revisions to Some maybe post-production process to make it really polished.

So those are the two ends of the spectrum, then you have a lot in the middle, and I think internal memo culture accompanies is a very big, I don’t know if you call it a publishing, it is a type of publishing in a way.

So for example, at Hiroku, I think we typically use GitHub and just. As a way to kind of publish memos, email, of course it’s a classic kind of internal memo. I’ve used Google Docs in the past, as I’ve mentioned, at Muse we use Notion pretty heavily, but more and more we use Muse just depending on what the item is.

But Muse makes it hard to publish. We don’t offer a good or easy way to screenshot or PDF, that sort of thing. Again, that’ll be coming in the future, since our focus really has been on that individual ideation point.

But yeah, if you think about, OK, when I write a memo in notion for my teammates, well, I only have 5 teammates, or there’s only 5 people in the muse team, which means I have 4 colleagues, so there’s a total of 4 people who are going to consume this, and I want it to be comprehensible. I want to respect their time. I want them to be energized by the idea. I want them to just understand what I’m trying to say, so it’s worth a little while. to make sure I’ve got my thoughts together and it’s not just a stream of consciousness that no one can follow, but it’s only for people and people I’m pretty mind melded with and we have a lot of context, so it can be pretty rough. And on that same note, we do share muse boards internally even though there isn’t great mechanisms for that yet, and it works well because again it’s sketchy and it’s It starts as this individual ideation, but it doesn’t take a whole lot of effort to cross that threshold to something that I can give to you or Leonard or all four of you, and you can understand because you have this context and we’ve already talked about these ideas before and this is just another iteration on an overall philosophy of what we’re trying to do with this company and product.

00:30:18 - Speaker 1: Yeah, super interesting, and I like that you mentioned spectrums because one of my favorite intellectual tricks is to look at spectrums or grids and interrogate each of the points on it or along it. And now that you’ve talked about spectrums, this kind of connects back to one of the original impetuses from you, which is to speak in terms of spectrum. There’s a lot Software for the extremely populated side of the spectrum. This is like enterprises where you have Lassian wikis and so on, just because there’s such a big amount of money there. And there’s a fair amount of software for the individual side, if for no other reason, that’s kind of the base case and you kind of got to start there before you add collaboration. But we had this hypothesis that a lot of the creative magic happens. In small groups, maybe it’s 3 to 30 people. It’s the whiteboard, it’s the brown bag lunch, it’s talking over dinner at the summit, and what would it look like to have a creative tool that really embraces and supports that. And if you back into the amount of fidelity, it’s probably this intermediate level that not coincidentally, Muse tends to work with. So I think there’s another way to look at how text blocks and the other features around it can support the type of group creative thinking that we want to see more of.

00:31:29 - Speaker 2: And I’ll make a mental bookmark to do an episode some time on our vision for collaboration. I know you and Leonard have been doing some deep sketching on that just recently here and some exciting ideas shaping up, but I think one of the core constraints that makes it interesting is that point in the spectrum where you’re talking about.

I ideation to share with this small group of colleagues and that’s very different from, I don’t know, a presentation you could argue is something that is intended either for a larger audience or maybe you’ve got your keynote deck for your client, you want to press them because everything is super polished.

You don’t need to impress your colleagues or hopefully you don’t need to if you have the right kind of team. Instead, what you want to do is really get those amazing ideas flowing.

As Nicholas Klein would say, get the creative collaboration going, turn my ideas into our ideas. I think there is a big opportunity with digital tools to allow that to happen for these small teams, but on a remote basis. Now, what are the challenges ahead for trying to bring text into this visual and spatial environment?

00:32:35 - Speaker 1: Well, there are some interesting nitty gritty design issues with text.

So one that we’ve been grappling with is text is only sort of spatial.

So think about an image, an image definitely has a two dimensional representation. You can make it bigger or smaller, you can translate it, it basically works, where text, it’s really linear and you need to choose some way to wrap it, and by the way, the wrapping might change depending on your font or even your text engine. So we have a little bit of an impedance mismatch between the very visual spatial original conception of muse, where you just to be concrete, you could basically take a picture of your iPad and that’s how you would expect it to render in all cases, versus text which people expect to kind of reflow basically. And so how do you reconcile those two worlds? I’d say that’s basically an open design question for us. We have some ideas. Another related issue is that text, especially small bits of text, they don’t quite map as neatly to the card block idea that we’ve had throughout Muse, because Again, for something like an image or a PDF or a video, it feels sufficiently substantial that you want a card that has a different background, it has borders and so forth, it feels like a distinct item, whereas if you had a one word item on your to do list, for it to be a whole card, it’s like a bit much, which is one of the reasons that I think our current Text implementation feels a little weird in some cases because you have this like basically huge card for one word, just it’s kind of missized, but then you have these new items, I guess we’re calling blocks which are very related to, but they’re not exactly the same as the cards that we’ve had on boards previously, for example, perhaps they’re transparent or translucent. So figuring out how to evolve the mental model and the design interactions to support that is another tricky design problem.

00:34:29 - Speaker 2: Yeah, those are two very significant and concrete ones, and I think it does reflect going back to this symbolic versus spatial.

The nature of a diagram, the fact that the circle is next to the arrow is very significant information.

That’s important. You can’t put the arrow under the circle and now the same thing will be conveyed, but one value of this one dimensional string of characters is that you can display lots of different ways. The web is very good at this with responsive design. If I open something on the phone, it reflows everything. So that I can read it comfortably in that format, but if I open it on a big wide screen monitor, things look different there. But the text content, if I was to take a screen reader or just read aloud, a particular piece of text content there should come out the same, no matter where it wraps the words, no matter if I have the font size cranked up to make it more legible or something like that. And so now we have this mismatch.

00:35:23 - Speaker 1: And Adam, you also alluded to another big set of challenges with text, which is editing. So there’s two pieces to this. One is, as we introduce these text blocks, you have the issue of text at the block level versus text at the character level. And with a traditional text editor, all the manipulations are done at the character level or they’re basically macros for doing character level manipulations, you know, you double or triple click or whatever it is to select the whole paragraph, but that paragraph is two indices essentially at the character level, right? Whereas we want and use to be able to support manipulations at the block level to be able to reorder whole items.

But then sometimes you’re kind of crossing that boundary, like, what if you want to select one whole block and half of the next block and delete it? Does that actually do that or is it like both blocks or just disallow it, you know, that’s the kind of design stuff that we’re grappling with on the text front. And then there’s the whole like actual editing in the sense of adding and removing.

Characters, which, as we’ve talked about on the podcast before, is quite difficult on the iPad because people have different input modalities, they might have voice, pencil, touch, or keyboard, and even the keyboard is potentially not as high like throughput as a desktop keyboard for various reasons. So figuring out how to allow effective editing that also embraces the different input modalities that you have with the iPad.

00:36:47 - Speaker 2: Yeah, as we talked about in our iPad episode recently, the multimodal aspect, you’ve got a keyboard, a trackpad, a pencil, your finger, voice input, potentially, or some combination of those, probably you don’t have all of them. You have some subset of those.

On one hand is part of what’s exciting about the platform and opens up a lot of new possibilities, but it also makes it trickier in some ways compared to the more known form factor and even posture of the user when you talk about desktop systems. Right.

Yeah, I’ve used a folio keyboard for a long time, but I recently picked up the magic keyboard, which is kind of this thing that the tablet sticks to magnetically and has a little trackpad, and then you can pull the tablet off when you want to do tablet mode stuff.

It’s really great. I see now why I got good reviews. The price feels disproportionate to me. It costs significantly more than one of the smaller test iPads that I have that I use for QA on Muse, so I’m not sure. that’s really something I’d recommend broadly or even we can expect users to have, but some kind of hardware keyboard is fairly commonplace, whether it’s a simple little Bluetooth keyboard or using a folio keyboard or something more like the magic keyboard, but it does mean you’re in typing mode versus reading, sketching and rumination mode.

And maybe that’s OK, but what I found in my own testing of text blocks is that I tend to go into typing mode and I type a whole bunch of stuff, I’m moving them all around the board with my finger, but essentially it’s a bunch of texts, and then I pull the tablet off, grab the pencil, now I’m highlighting, now I’m Rearranging, and I think that works OK, but I would love something a little more fluid moving back and forth, but I think that’s probably out of scope for us.

I think that remains an unsolved research problem to be tackled if the tablet’s truly going to become a new kind of creative tool.

00:38:33 - Speaker 1: Yeah, a super important open research problem.

It’s not gonna be in scope for us for this iteration, but I do think it’s very important for us to solve it for the iPad if the iPad’s going to realize its full potential, which we know again alluded to in our last episode.

I mean, just to give you a couple concrete ideas. I don’t know I’ve even told you about these before. I mean, so maybe this is your first time hearing them, but one thing I would like to see is a tool for much faster and slicker input of voice.

So I always has like this voice thing, but it’s kind of slow and it’s laggy, and it’s like, OK, you’re entering voice mode and then you’re exiting, it’s like a whole thing, right? And what I would like is a tool in our toolbar that’s like the red ink or the eraser, where you press your finger down on the iPad, you say two words, take it off, and then you get a text card where your finger was, where you just said, I think of how good that would be for like building a to do list or doing some brainstorming.

And another thing I would like to see is, I think the pencil is actually potentially very good for editing text in the sense of cutting it and manipulating around. So you can imagine a sort of exacto knife type tool where you can basically cut the paper, and then you can imagine a finger tool where you can move the cut pieces around, right? And the pencil is actually very good and precise for that, maybe even better than the keyboard.

Again, both those, there’s just very little exploration of such input modes on the iPad to date, and it’s kind of a shame, but I do believe we will get there eventually. Maybe it’ll be us, maybe it will be someone else, but I think we need to see that to realize the full potential of text on the iPad.

00:39:59 - Speaker 2: Yeah, I don’t think I’ve heard you mention those ideas before, but I do find both of those exciting partially because there is sort of incomplete or let’s say implementations that exist on the iPad today that hint how those could be good. So in the case of the sort of voice quick input, I do use create a new text card, tap the voice thing, say three words. You wouldn’t want to do anything long form but for more the Post-it style thing, particularly when you don’t even care what the capitalization or punctuation is, and I like it quite a lot. There’s probably a little psychological hump to get over with talking to your computer, but that’s totally a thing you can adapt to, I found. Yeah.

But yeah, you’re right, it takes enough tapping and special things to get in there that sort of nullifies the convenience probably. And then similarly with the pencil scribble, which is a feature in iPad OS that essentially allows you to do handwritten input, and we do have a pencil tool for that in Muse that some folks use, but it does have some downsides and challenges, but one thing it does have is the ability to essentially, well, scribble out words to delete them, and that feels great. I really love that.

Now in practice, I don’t use that mode. Enough to kind of really that be part of my life, but again, you could imagine seizing on some of those sparks of something that feel good and this fast and precise and expanding on that either in a research context or in product development.

Well, I guess that’s how we’re thinking about text overall, so I’m excited to see how this beta evolves. Now we’re trying something not completely new but a little new in the sense that this beta is very rough. It works, but there’s a lot of problems, a lot of quirks, a lot of bugs, but we really wanted to get it out as early as possible to our pro members to try it out and give us input not just on does it work, but how do you see this direction as making sense. Again, coming back to this. This is bringing together text and a visual spatial environment. Is that something really powerful, exciting, new, opens new vistas in your own creative work, or is it a weird uncanny valley where it’s sort of not good enough at text to use much? And so what we’d love is to get you to try it out and yeah, give us your thoughts. We did a version of this once before with what was called the Infinite canvas beta, later became flex boards, and it was similar in the sense that And the infinite canvas work was going to be a huge investment in engineering and design work. I think it ended up taking us something like 3 months of wall clock time to get it all done. So we really wanted to feel a strong sense of product validation, or this is valuable to people, you know, we’re a small team, we have limited resources, we want to make sure we invest them in places they’re gonna be useful, and that worked really well because we gave, I think it was just a test flight kind of beta, but we basically gave this pretty janky prototype. to people, but got this instant powerful response of this is great. I have to have this. This is so much better than what’s there today, even in its rough form, and that gave us both the energy to push forward to the bigger project, but also gave us useful context, things about disorientation and getting lost on bigger boards and stuff like that that fed into the final design.

So we’re hoping to do something similar here and actually if this pattern works. Sharing data is pretty early to both validate the direction, but also really get key feedback from people. That’s something I’d like to make a really regular habit of. So yeah, if you’re listening to this in, I guess, summer of 2021, when the beta is still running, yeah, we’d love to have you try it out, try it for your workflows, and give us basically a thumbs up, thumbs down, is this a worthwhile direction.

00:43:34 - Speaker 1: Yeah, and I’ll be looking to see if the text blocks beta passes what I learned as the Wiggins product management test, which is, you suppose the feature is going to be going away and how mad would customers be about that? Would they really fight for it, because it’s easy to say, oh yeah, this is fun, this is nice, but where the rubber hits the road is, we’re thinking about not really adding this to the product and the reaction that you want to see is absolutely not, you know, I’m gonna fight you basically. That’s the level of excitement we’re hoping for, but we’ll see, you know, that’s why we do a beta.

00:44:01 - Speaker 2: Yeah, that’s right. I think it’s important or one of my values product wise is to keep things streamlined, simple, minimal, just what you need, not sprouting a million features, but of course you also have to try a bunch of stuff over time to find the right things, and so having a bit of a willingness to kill your darling. I think is the screenwriting term for this, which is, you might really like the idea, probably you can tell from our voices we’re excited about the direction, but we want to see that it truly adds something worthwhile to the product rather than just, it’s a nice idea, but in practice it doesn’t pan out.

Well, let’s wrap it there. Thanks everyone for listening. If you have feedback, write us on Twitter at museAppHQ. We’re on email hello@museapp.com. I mentioned at the beginning, reviews on Apple Podcasts and elsewhere are much appreciated, and I’m looking forward to hearing all your feedback on whether text and muse makes sense.

And of course, in general, we’d love to hear your thoughts on the philosophy of text and digital tools and tools for thought and spatial and all those other things. So please tweet at us if you have a reaction. All right, see you next time, Mark.

00:45:09 - Speaker 1: See you, Adam.