Skip to navigation
Why does this page look like this?
Can we design the invisible?
Many designers focus on the clearly visible things. But there are many parts of a website that are invisible. These parts are either neglected, or they are "designed" by developers. Real people would benefit if we paid some design attention to things like structure, error messages, and of course screen readers.
Vasilis: Today, I wanted to discuss the question, can we design the invisible?
Espen: Oh, that's an interesting question.
Vasilis: Mm hmm.
Espen: What do you mean, when you say the invisible?
Vasilis: Have you ever designed things you cannot see?
Espen: No, maybe depends on the definition, I guess.
Vasilis: Yeah. Okay. So. So I was just wondering, are there others? is there other stuff to design than just the things we see?
Espen: Well, when you say design invisible, what I picture in my head is this sort of experiences that happen in the brain of people that aren't on the screen. So if you're writing a novel, you know, I guess what you're, what you're writing that is invisible is the pictures that you generate in between the lines, so to speak. So as we, you know, thinking about design, it would be things like maybe curiosity, or questions or emotions that that occur based on the choices you've made on screen.
Vasilis: Nice shot, it could be that surface design parts of service design, we could say that these are invisible. Right? There's lots of design going on into designing how, for instance, when you order something that the email arrived just in time, and how does the email arrive and things like that? These are the timing, for instance, is not really visible, but it's important.
Espen: Yeah, talking about emails, even the timing of the first email after you order something, sure, but also the whole design of the of the of the follow up. journey, you could say, you know, there's this maybe 15 emails waiting for you in this track of emails that have been designed, and they're invisible to you as a user until you receive them.
Vasilis: Yeah. Okay, most often, if I look at these emails, I think they are not designed. And there's not really put much thought into it. So I get for for some reason I get I think six emails every time I order coffee at someplace. This is just ridiculous.
Espen: That's incredible.
Vasilis: Yeah. I need maybe one email would be sufficient. Yeah. The order?
Espen: Yeah, you got it work. It's coming.
Vasilis: Yeah. Don't worry. Yeah. So but I wasn't really talking about that. I was thinking, Well, I was not think, Well, I think it's really interesting to talk about that. But maybe we should talk about that. In another episode. I was thinking about the the other senses. So we have eyes. Of course, we can feel stuff. But there's not much to feel on the web, I think while you have touched devices, but still, they all feel the same. Right? We cannot change you cannot alter the way they feel.
Espen: You know, in a sense, we we might soon I mean, we have haptics and stuff right on the on the iPhone and think about gaming interfaces, there's rumbles and and shakes and stuff in the controllers these days.
Vasilis: Yep, absolutely.
Espen: Apparently, the new PlayStation five controller has really good haptics that respond to whatever strength and stuff in the controller so there's boundaries and stuff like that coming for your average mobile devices as well, I'm guessing
Vasilis: Haptics are really, really interesting. Yeah. Yeah. Yeah. And then of course, there's smell. There's not much smell to design a few students of mine actually designed a perfume just this year. Uh huh. So and I had never encountered anything with smell in at University before so that's really interesting. So they're really dug into how to how do you design a smell.
Espen: Yeah, I think in in Environmental Design, smell makes a difference. If you you know, as a shop or a centre for something, I guess smell could come up come into it. Certainly. the freshness of the air or the lightness or the mist or the you know, these kinds of things you commented.
Vasilis: yeah, I think even in in product design there's something that if you open the box, and it smells good, yep, the thing, there's even a, some sort of a new car smell, for sure they spray into new cars to make it smell
Espen: Amazing. And speaking of cars, they also put a lot of effort into sound design, you know, with your blinking lights and the way the door shuts and everything like that. Yeah, I'm sure keyboard manufacturers also worry about how their keyboard sounds when you click.
Vasilis: So why don't we have sound on...
Espen: On the web? That's a good question. I think if we open that door, the predatory advertisers are going to get there first. And they're going to ruin it for everyone else.
Vasilis: That's a very good reason,
Espen: isn't it a bit like video, you know, video is there. And we could in theory be using a lot more video. But people get annoyed when there's, you know, intrusive video content here. And there. There's something as being too much when you scroll a website, and it's just, you know, stuff on top of stuff on top of stuff. And if you add sound layer to that. But that's the average, you know, advertising driven websites I've seen or heard experiences online, that are fantastic with sound. There was one I can't remember what it was called. But it was like soundscapes in the city, you could go to different cities. And it was basically like, you know, a big, maybe it was a 3d picture, you could zoom around, but then the soundscape was just someone had been at that spot and recorded the sound. So cars honking people walking by etc. It was really good. But you know, it's a bit of a different.
Vasilis: Yeah. interface noises right, when you hover over something that it just looks blue, and then when you click on it, yeah, and when it doesn't work. Yeah, these things that you see them everywhere. We have interface sounds. I mean, do you remember the the old iPod, the iPod? Classic?
Espen: The white one with the wheel?
Vasilis: Yeah, with a wheel and it had a [clicking noise]
Espen: Yeah when you scroll the wheel for sure.
Vasilis: That was so nice. It was just nice to listen to it.
Espen: Even now, new iPhones, they have a sound on typing, don't they? And actually, it's interesting because games on phones and games and consoles, too. They use sound like a lot. I was just playing Mario Kart there with my daughter. And you know, maybe we will have a whole episode on the interface design of Nintendo because it's bloody amazing.
Vasilis: Oh, yeah.
Espen: So good. There's this. It's so not subtle. Everything is super clear. The buttons are massive. The hover states and the active states. It's just [kissing noise] . But they all obviously use sound as well.
Vasilis: And also yeah yeah, and good sound.
Espen: Yeah. And computer games as well. I've sure I've come across hover state sounds inside a game. Yeah. So maybe it's a game context thing?
Vasilis: Yeah. So there was this... this is film we should look should look it up. This is I think he's a Dutch gamer, professional gamer, who plays games. And he while he's blind, and he plays his fighting game. So there's two people fighting. And he's actually he beats the other guy, just by sound, he cannot see anything. But because the sound design is so well done. Every movement has its own sound. So you can react to sound. It's just incredible.
Espen: That is incredible. I can't even imagine the reaction time needed for that. I think I think certainly the fighting games I've played though, that I imagine there are visual cues that appear sooner before the sound cue appears. Maybe not.
Vasilis: It's it's it sounds like it's the same thing. Sounds amazing. Yeah. Yeah. Yeah, that's, yeah, we should look it up. We should put it in the show notes.
Espen: But you're, you're an artist and you break the internet every day. With your stuff.
Vasilis: Would you use sound a lot? Not never. And I think that one of the reasons, of course, I thought about using sound for interface design, but then the accessibility world. They said no, you're not allowed to. Yeah. And well, I thought, Okay, well, that's a good reason, because it might annoy people. But then on the other hand, by saying you're not allowed to, we cannot research it. And we cannot find ways that actually improve the internet. So I'm allowing it in my know, if you want to use sound in your design, go ahead and do it. And the reason why is because if you look at so people who depend on sound are blind people, of course. And now can we design an experience for somebody who is blind? And then you have to design something for a screen reader? Have you ever designed for a screen reader?
Espen: I've never designed anything. I know you have. And I've took great pleasure in hearing your talk about how you made a laughing sound in a screen reader.
Vasilis: Yeah, that was fun. But that's about as much as you can do. So you can try and hack the voices a little bit, you cannot add noises, you cannot add anything. There's very limited interaction that you can add, because all the keys are just taken over by the screen reader. So there's really not much you can do. So yeah, so what I designed was, I worked with a blind photographer or photographer who turned blind later on in his career. And then he switched to making these soundscapes as you call them from CDs and other environments. And he had his this website with the soundscapes on them, and I redesigned the website for him. And then I asked him, Do you want anything extra on this website? Because it was actually pretty boring to look at. And then he said, I want funny little animations to illustrate each record recording. And I said, well, you do know that you cannot see these animations, right. I'm making this website for you. And he said, No, that's that's okay. But I want to experience them, though. So then I rethought how do you design these animations. So instead of starting with the animation, and then creating an alternative text, I started with the alternative text, I tested the text with my friend. And then when he thought it was funny, only then I started making the animation, which is actually a pretty clever way of working because it's much easier to create an alternative text than to create an SVG animation. So and then, one of the things I tried was to make the screen reader do funny stuff. So I thought okay, let's make it giggle. So, I tried letter combinations, and then they came out [weird laugh]. Which is really good way of giggling in Dutch but it's impossible to [weird laugh] create in English because there's no [ghe sound] . So there it sounds like "ghe ghe ghe", which is it's not it's funny, but it's not giggling. Then, I started working on an international screen reader giggling database but... [both laugh]. So I made the screen reader say things like, haha, and boing boing boing, things like that. Yeah, this is an the, my friend, he just he just laughed and laughed and laughed because it scry there's they're designed to sound as neutral as possible. Yeah. And then when there's really neutral business, like voice, all of a sudden says boing boing boing, then it's the it's just hilarious. But things like that this almost never happened. And I think there's a there is a lack of hooks when it comes to screen, which we should be able to design more for them.
Espen: Whatever is screen writers, if you're using if you're using audio in your interface, like you play some sort of ambient sound in the background, when you load the website, will that be ignored completely and replaced by the silence of the screen reader? Or would would that also add to the you know, audio experience of that website?
Vasilis: Yeah, I guess it should be possible to do something. So you cannot load sound you cannot play sound during loading you have to interact before sound is played, which is a good thing. Of course. You don't want advertisers—
Vasilis: Right. So but yeah, I think you could, for instance, add a noise when you hover something or where you focus something, which I think would make it easier to understand this is a button, is this a link? Or is this something else? Right? Yeah, I guess these these experiences, these are the experiments I think, are really valuable. But you also need, of course, good sound designers, you can ruin it at the same time, but the same goes for visual designers, right? If you have more visual designers, you have poor hierarchy, and the interface gets cluttered and unclear. But but I think we need other stuff as well. So for instance, if you go to a website for the first time with a screen reader, I mean, if we go to a website, we see the hierarchy immediately, right, we get an overview of what's on here, we get an overview of what's the tone of voice just by looking at it, right, we immediately see this is a website about death metal, obviously, this is a website, My Little Pony, right? The differences clear, even in, not that well designed website. But if you go there with your screen reader, you hear the same voice you hear for everything. And then you have to just listen, one by one, step by step, what's on this page? And if that, what last week last time, we spoke about content, if the content is not written very well. How do you find out if it's a funny website, or an ironic website or a serious website? Right? This is really complicated. And we can see the hints we can see it. So I really believe we need more design hooks where we can, for instance, alter the speech. And this is possible, actually, these techniques do exist, but they are just not available in screen readers.
Espen: Yeah, I suppose changing the voice of a screen reader is a bit like changing the font of a visual design, you immediately change the tone of what you're working on. Right?
Vasilis: Yeah. Yeah, I think it goes maybe a little bit further this. So maybe this is why it cannot be done. So you, maybe you even change the complete interface of the computer? Because this is what a computer, the complete computer is that voice for people with. screen readers are the interface is the voice, right? So it goes a bit further than So for instance, if you slow it down, it may get very annoying for somebody who's a high speed listener. Or if you change the pitch, it could. But these are tools that don't exist, but they're very, they would be very nice to use. Right? We can use them to enhance the experience.
Espen: Yeah. There's no one working on this at the moment.
Vasilis: Well, not for screen readers, as far as I know. So there's all kinds of course, there's all kinds of voice interfaces right now. So Amazon and Google, and they're all working on these things. And they're all these very advanced voices are available. And in these interfaces, you actually have control over more control over pitch and tone and, but not in screen readers. So I really think we need to redesign screen readers.
Espen: That's a challenge for you.
Vasilis: You know, this, the problem with screen readers is that they are based... So they are designed. They were designed in the 90s. I think they started designing on them. And they were designed for people who use blind people who use computers in the 90s. So that there were fewer people using them back then and they were nerds. So their—
Espen: computers were different back then to
Vasilis: Yeah. And then they evolved a little bit into the the zeros and then they were designed for nerds using computers in the zeros and then I think they kind of stopped evolving. They'll well voiceover came and evolved a little bit in the 10s the last decade. But that's it and it's all based on how we used computers 20 years ago. Yeah. And it's not based on all the people who just came into using computers in the last decade. So these, these interfaces, screen readers are not designed for normal people who turn blind. They are designed for nerds who know how computers work. So that's why I really think we need to redesign screen readers because there's a whole new population who needs these things. But they don't understand the the... just the basics of it. It's just not made for them not designed for them. So for instance, do you think that people that normal people should know what a heading level is?
Espen: Normal non web people?
Vasilis: No. But when you're blind, one of the first things you hear is heading level one.
Espen: Yeah. I mean, from from a blind person's perspective, the whole interface, like you said earlier is sound right? So there needs to be in a way there's there is a different way of interpreting things because they need to be described to you in some sort of audio format. And I guess, heading level one, is one way of saying it. Main header is another way of saying it, maybe there's a more human way of of describing what's the biggest, most prominent stuff on the page. And what's not so important, what's the footnote was this was that without these very, very technical terms that are identical to the, to the tags, because, you know, normal users don't care about tags, they care that that's the big header. So that's the main thing, this is small text, I'll read that later. It's like, obvious, maybe there is a way also with sound, to paint colours, or you know, obviously, audio, but in the same in the same sense, we, we, the people who can use our eyes, and we can see properly, the elements on the screen, we can use the visual cues to know that something is a button. So all the only thing we have to focus on is the text inside the button, it says buy now. And we know we can click it because you know, it's it's we don't have to think about that. And maybe in the screen reader, if I'm correct, the screen reader will say button first or link or something like that. But maybe there could be audio that describes it better a sound a click distortion of voice, I don't know something. So that the only thing you really need to pay attention to is the word the phrase inside the button rather than describing that it is that it is a button.
Vasilis: Yeah, yeah, I think these kinds of things would really help. Other stuff, of course, would be maybe analysing the facial tone of the page, and then explaining that, for instance, this is a I don't know. Hi, high contrast, brightly coloured. Yeah, I'm not sure if this is really useful for somebody who cannot see, but maybe you could.
Espen: It's, it's hard as maybe not useful to describe in very blunt terms. This is what the page looks like. However, the page looks like that for a reason. So something is calm or something is intense, or something is some other adjective to evoke some sort of emotion to connect with the content in some way. You know on DVDs are also actually talking about games, but DVDs to have these title screens, or they used to at least back in the day when DVDs were a real thing. And often there will be music, there will be a picture of Aladdin and blah, blah, blah, and some buttons to play or go to settings, and there will be some sort of 22nd score that would repeat. That sets the tone for the for the movie. And it's maybe a little bit annoying, because you just want to get to the play button. But the idea of that, maybe again, going back to this thing of audio like soundscapes if for me someone who can you see I see dark colours and mountains and like very stark, bold fonts and I know Okay, it's heavy metal. Maybe there's ways you can use sound to invoke this feeling of Okay, I know I know what territory I'm in now. Just like you do at the beginning of a movie before the before any visuals show up people often you know the sound design starts before the visuals come and you're immediately in that sphere of some sort of emotion or mood. So maybe I mean, that would obviously require us to like, play audio, or force audio audio on on people, or at least, you know, trigger it in some way.
Vasilis: You can trigger it after interaction, you can ask, do you want audio? Do you want the audio experience? And then when you say, yes,
Espen: It starts.
Espen: Would there be a massive performance cost to this, if you had like files or audio files take up a bit of space don't know.
Vasilis: Well, there's audio files, but there's also an Audio API, so you can generate sound in the browser nowadays. So you don't need to send files over the internet, you can just use the button the built in API's in the browser, you have just a complete synthesiser.
Vasilis: You can generate every sound you can imagine.
Espen: So for for blind people, what's the what's the most advanced best interface out there?
Vasilis: Well, right now. Probably it's these conversational interfaces. Yeah, I guess. Maybe Amazon is the best interface at the moment. But I think that's, Isn't that pretty? Just? Is that it?
Espen: Yeah. It's very, it's very limited. It's, it's, it's like having a word processor. To give you any visual interface. Yeah. It's very limited. But then again, what what do I know? Maybe not, I guess, you know, maybe the system third, you
Vasilis: know, the best interfaces, these games, right? The example of the the guy who plays a game? Yeah, well, he cannot see the game. I think that's, that's the ultimate interface. That's how far as it goes. And probably as far as it can get? Well, as far as I can imagine, well, there's some interesting experiments that students of mine did. So one student he made his, the trackpad. So he said, if you if you go with your finger over the trackpads, to your, it corresponds with a mouse going over your browser window. And it will give you an idea of where the interactive items are. So it gives you an idea of the of the hierarchy of the page. And he played with sounds back then it was very interesting experience, experiment. And the blind person we worked with, she was really, really excited about it, because she just she'd never really thought about it. Things are take up space and that they are placed on a grid in that environment. Yeah, yeah, place relative to each other. And so for her, it's all always linear. Hmm. So that was a very nice experiment. There was the others that was just just super clever, super creative guy, he created a, so he worked with a blind sculpture. And he, he thought, okay, so you like 3d things. So and, the blind guy, he was, he's also an expert on environmental. Well, environmentally building, something like that. So a nerd in that way. But he, it means that for his work, he has to read tables, data tables, lots of tables. But tables are just really, really hard to design. And they're really hard to design for blind people, especially. So what he made he made a with his phone, he turned his fourth phone into this three dimensional mouse like thing where you can just navigate through a table, it became more of a work of art. But the the thing was so interesting. It was so so novel, and so did you could move the your phone in 3d space, and then you just float through the table and it started reading data to you. I mean, these kinds of experiments. Well, they're not really possible with current screen readers, I guess. But these are really interesting.
Espen: Yeah I think it's super interesting because it flips the whole idea that everything non visual has to be text. Whereas in fact, it doesn't have to be you know, I what you said there about the phone. It makes me think of VR. As well, obviously VR is the epitome of a visual interface. But in theory, it doesn't have to be you could be tilting your head. And then, you know, the interface could tell you. It could tell you navigation content, this that whatever. It could give you sound cues. When you're getting close to an element you can interact with or not. You can you could include haptics in this experience, as well imagine, you know, a track mouse tracker pad or something like that. It could give you tactile feedback. Yeah, obviously, you know, that's an interface that would have to be completely redesigned and rethought out it's wouldn't wouldn't translate to the way we design visual interfaces, but still, you know, toggling between items.
Vasilis: I think we have something like that. So if you look at the iPhone, for instance, now the Apple Watch, I used that, experimented with it a few times. It has a navigation on it. Way Wayfinding. And it, it gives you a single click, I think so haptic click to go to the left, huh? Double click to go to the right, something like that. Also, sound of course, is an augmented reality, right? It's an extra layer on top of what we already hear. So if you add sound, for instance, go to the left. Yeah, right. That's, it already exists. Yeah, but we just always think about augmented reality is something visual, but it's also Audible, of course. And haptic.
Espen: Oops, just bump my microphone there. Speaking of haptics.
Vasilis: So yeah.
Espen: Is this your next course then for the next batch of students coming in next year, after the summer, audio design.
Vasilis: No, after the summer, it's about mostly teaching them about designing with CSS, and HTML. So there's a little bit of teaching them about screen readers. But the more experimental courses, l starts in February, again, I say So then, yeah, we'll be designing with individuals with different disabilities again, and Yep.
Espen: It is a really really interesting field. It's like, there should be it should be, you know, maybe there are I don't know, but blind designers working on, on this to see what's what's possible, because it sounds limiting. But like I said earlier, I don't use screen readers. So I, you know, I wouldn't be qualified to say that this isn't good enough for this is it should be like this, it should be like that, who you know, who better place than than the people actually using these interfaces? And to come up with new ways of communicating without visuals?
Vasilis: Yeah, so what I see is that there's not much design going on there. It's more than just using best practices. Yeah. And I believe that these best practices are actually not really that good practices, then. They're, they're old, very best practices, based on ancient insights. And if I look at the well, my current research, I think now we need more, much more experimental work here and playing around. And I also believe that if once we say, this is a field where you're allowed to play, you're allowed to experiment, then all of a sudden, it becomes an interesting field for designers. Yeah. If I tell you no, you're not allowed to do anything here. Yeah. Then why should you come over and try anything because nothing is possible.
Espen: Then also, if screen reader technology works in a very, very limited way, and there are no hooks or no holes that you can plug in your experiments, then? Only only the rarest of people like yourself will actually be inclined to try and experiment with it.
Vasilis: Yeah, yeah, in the end, it's very limiting. So it's really it's it stays in an experimental phase. We cannot go any further. We cannot publish these things because they they need a special interface. We need to disable the screen reader for instance, to emulate a better screen reader It's all just what if experiments, this, which is sad, but
Espen: maybe it helps. It's where it all starts?
Vasilis: Yeah, probably.
Espen: I think the modern computer was a what if experiment at some point.
Vasilis: So would you like smell in your interface?
Espen: Tell you what, on on the average website? No, but this is my office. And this is my work interface. And right now I have a smell. I have a to my right. Little, what is it called the mist diffuser thing with drop of some sort of oil. And it's, it's actually their sound, too is a very, very subtle little sound. And there's light coming from it. And there's like this visual smoke cascading down my—
Vasilis: that's what I was seeing!
Espen: And there is smell as well. So okay, you know, it smell is great. But again, do you want advertisers to have access to some sort of perfume dispenser at the bottom of your screen?
Vasilis: There was this Belgian startup in the 90s that actually worked on it. So they've worked on. Smell, smell, outputs, device. Yeah. For your computer.
Espen: Don't they have this in like, what's it called 4d Cinema or whatever it's called? Like, is the water splashing and wind and movement and smells and stuff? I've never tried it. But I imagine in it helps with the immersion.
Vasilis: It could be Yeah, what I heard, but I only heard it from somebody who went to a very early one of these experiments or experiences. He said, Okay, so I was at sea, and then water comes over me it's not... it wasn't that great.
Vasilis: Yeah, very good. Definitely out. But I think smell is pretty complicated. But well, so sound, so is visual design.
Espen: Yeah, it's just that maybe visual is so much more accessible to much larger subset of the population. So that's where we put all our money, so to speak, and we're way more advanced there than anywhere else.
Vasilis: Yeah, yeah. Sound design, I think sound design is pretty advanced as well.
Mmm. Yes, but yeah, specific kind of arts though. Right. Like movies and music. Obviously. Web design not so much.
Espen: Hmm. Oh, my God. Mario Kart.
Espen: Speaking of sound Did you watch a movie called The sound of metal? I think it got the Oscar for Best Sound Design.
Vasilis: Okay, no I didn't.
Espen: About a heavy metal drummer who goes deaf. Absolutely amazing.
Vasilis: Oh, wow. I should see it
Espen: Really used to sound incredibly well in the movie to get you inside of this horrible state of mind that he finds himself in.
Vasilis: Oh wow
Espen: Really really good.
Vasilis: Okay, going to look at it. You're going to watch it. Nice. Do you have anything to add to this invisible design, invisible design question I asked you?
Espen: Only just say what you call it desound?
Vasilis: [laughs] Desound
Espen: No. I don't know how to design the invisible. I've never really done it. I think there's a there's a whole spectrum. Like I said at the beginning I think what happens inside of anyone's brain when they interface with something is goes way way beyond the whatever visuals and we just happen to have a population where the eyes are like super, super important and over represented so to speak. So we become really adept at making visual things, movies, and we add sound to this experience in many different art forms. Not so much on the web. But we're only dipping our toes into the full set of senses right? Haptics are coming rumbling controllers VR, sound smell lots of sure. But you know, I think I think it's we're we're only just starting to, to touch some of these things. And I have no idea how to do any of them.
Vasilis: Interesting. That's to learn then.
Espen: Yeah, I To be honest, I don't even know how to edit this podcast.
Vasilis: Me neither.
Espen: There you go.
Vasilis: Okay Espen it's been great talking to you again.
Espen: Likewise. Thank you for such a weird and strange question.
Why does this page look like this? #
This is a simplified visualisation of what a website looks like for someone who uses a screen reader. Screen readers use a neutral voice. This means that every website has exactly the same neutral feel to it, no matter if it is a website about hilarious movies, or if it is a website about depressing histories. Hence the sans-serif font. This means that good content design is necessary to understand the mood of a page.
A lot of noise is added to each page as well. Heading levels, for instance, are literally announced as heading levels with their number. And the navigation is announced as such. Sections and articles are announced as groups. Which always makes me wonder: do we really expect all our visitors, normal people who visit our website, to have a thorough understanding of web semantics? Then why do we expect all blind people to have this expert knowledge?
This is not an accurate visualisation of a website read out by a screenreader. If you fully depend on a screenreader you can only hear one item at a time. Which means that to get an overview of the hierarchy of a page, you depend completely on you own memory. Those of us who can see do not have this problem, we can still see the content surrounding the thing we are reading or interacting with. We can clearly see how much content is left, based on the position of the scroll bar. And we can still see each piece of content in a broader context, even if it is designed horribly, like this page.
I hope this horryfying design will inspire people to start designing for screen readers. And I hope it will inspire people to start creating a new kind of screenreader, especially designed for normal people, not nerds.