Welcome to the fifth episode of Access On, the National Federation of the Blind's Technology podcast.
Episode
Listen to the fifth episode of the Access On podcast (Browser).
Or listen on your preferred podcast platform.
Timestamps
The show is segmented by chapter, making it easy to move between segments of the podcast if you have an app or player that supports chapters. Below is what's on the show this week, and when you can hear it.
- Introduction 0:00
- Karthik Kannon from Envision discusses their new AI called Ally 1:09
- Reordering tracks in Ferrite Recording Studio 35:09
- Listener responses to Thunderbird accessibility question 38:40
- Facebook accessibility 41:50
- Ray-Ban Meta glasses feedback 45:34
- Induction cooktop that can be used by a blind person 47:27
- Braille displays and notetakers 49:00
- Live Listen not working 52:22
- Closing and contact info 56:38
Transcript
Jonathan Mosen: Welcome to Access On, the technology podcast of the National Federation of the Blind. Envision has been working hard on a new AI product called Ally, designed with the blind in mind. Listeners respond to a question in a previous episode about the accessibility of the Thunderbird email client. More listener comments on Meta things, including Facebook, Instagram, and the Ray-Ban Meta smart glasses. And our tech tip describes how there are ways to switch apps in Windows, other than the venerable alt tab. Welcome to episode five of Access On, I'm Jonathan Mosen, coming to you from the National Federation of the Blind's Jernigan Institute in Baltimore, Maryland. And just a programming note to let you know that after episode six, which will be published next week, we'll be taking a small break between Christmas and New Year, and we'll be back with you in early January, 2025, as Access On continues.
AI is having a profound impact on the lives of us all. Blind or non-blind people the world over are using tools like ChatGPT or Microsoft Copilot to simplify our lives or help get the creative process started. But there are some unique use cases of AI for blind people, including describing the visual elements of the world around us or photos that we've taken. So what if you could combine all of that into one tool designed specifically for blind people? Well, that's the ambitious goal of Ally, a new product from Envision, the same people who bring you the Envision app and the smart glasses. And to talk about Ally, what it can do now, and where it's going, I'm joined by Karthik Kannan, one of the co-founders of Envision. Karthik, welcome to Access On.
Karthik Kannan: Thank you so much to have me here, Jonathan. And yeah, like you mentioned, AI is having a moment in time. Right now, it's capturing people's imagination, and I'm very excited to see how we can apply that for accessibility, like we've been doing it for almost seven years now.
Jonathan Mosen: So elevator pitch time, what is Ally and why should people be using it?
Karthik Kannan: Ally is a very easy to use conversational personal assistant that can go ahead and help you describe the visual world around, can go ahead and help you read text, recognize faces, objects, and so much more. And the best part about Ally, is that you never have to learn how to use it. You can just connect to it, and then just start talking like how you would talk to a person, and Ally would take care of everything under the hood. If you wanted to read text, you can just go ahead and ask, "Hey, can you read this for me," and Ally would read it out to you. Or if you want to get a description of what's happening around you, you could just say, "Hey Ally, could you just describe what's happening," and Ally would just give you a continuous description.
And the best part about Ally is that it's not just limited to your phone. It can work on the glasses, it can also work on the desktop, apart from the phone itself. And our aim is to bring the Ally assistant to as many devices as possible. And lastly, Ally is extremely personal. So the thing about Ally, is you can, A, give it a personality. You could basically write a beautiful story about a character, and Ally would role-play that for you. And you can also go ahead and give information about yourself, so you could tell Ally parts of your life that are important, and Ally would remember that for you, and then bring that up during conversations, or keep that in mind when you ask it questions. It's conversational, personal, and it's available everywhere, and it can help you do things more easily.
Jonathan Mosen: That latter point is pretty cool, actually. I've been testing this, and it helps deal with an issue which I've seen in other AI products, which is that some people like their AI to be friendly, maybe a little bit frivolous, and other people are very businesslike, and you can really make a difference to the way this thing behaves by giving it instructions. I've actually played with this from the incredibly frivolous and almost flirtatious, all the way through to deadpan serious.
Karthik Kannan: Yeah. When we first included this into Ally, it really felt like a gimmick. I was like, "Hey, we'll just put this in there, and it's going to be a fun thing for people to do," but it has now become a part that people enjoy the most, and so much so, that in the near future, people will be able to create Allies and share it with other people. So you can go ahead, and if you like somebody's Ally, they can share it with you, and then you could use it. And if you want to share your Ally to the world, you could do that as well. And all of a sudden, we'll just have a universe of Allies that you could pick from. And depending on what you fancy that day, you could just pick one and have it be your Ally for the day, or share it with the world.
Jonathan Mosen: So are you able to save those personality characteristics so that you might have one Ally in the office and one when you're on personal time, for example?
Karthik Kannan: Yeah, totally. And you could give Allies different capabilities. You could ask it to behave a certain way if you're in the office or at home, so there's a lot that you can do with Allies, and we're going to be bringing that all very soon, so people will be able to share and use all these Allies.
Jonathan Mosen: What are the benefits of using it over using ChatGPT or Gemini or Copilot, et cetera? What are those defining differences?
Karthik Kannan: One, I think we have put the blind and low vision community at the center of this experience. So right from the time you go ahead and set up your Ally, it's able to do it in a very easy way, that you never need sighted assistance for it. I would say that's the foundation. One particular example I can give you is just being able to get descriptions of people. Now, with ChatGPT, with Gemini, with Copilot, with any of the other big corporate-speak AI assistants that exist out there, a lot of them don't, for example, help you describe faces of a person, or they just outright refuse to give you descriptions if there are people in the room. And with Ally, people will be able to go ahead and get descriptions of people, and you can even teach Ally to recognize a face of a friend or a family member. And we try to do this in a way that's as privacy-conscious as possible, that's one key difference.
Another key difference here also, is when it comes to reading text. Most of these proprietary models still hallucinate a lot when it comes to reading text, so they don't have a specialized text reading AI. But with Envision, or with Ally, we've had a lot of experience building text recognition systems and text recognition tools. So you have the scan text feature on the glasses, you've got the instant text feature on the glasses. And whenever someone asks us to read something, we go back to those specialized AIs that we have built over the years. So the amount of hallucination you have when it comes to reading text is, again, very low with Ally.
And lastly, Ally is available across all platforms you can think of. So we're not really restricted to just one particular platform, we're not like Siri, for example, where we're only available on iOS and the Mac. But today, people can use Ally on iOS, Mac. People will be able to use Ally on the web. We're going to have a WhatsApp bot for Ally, a Telegram bot for Ally. And very soon, you'll be able to actually make a call to an Ally, and if you want to make a phone call to an Ally, you can have it do that. So it's just the fact that it's available everywhere, including the Envision glasses, that makes it a key difference between what Gemini and ChatGPT are offering right now, over Ally.
Jonathan Mosen: All right. Now what's interesting about that, is that if you can make a WhatsApp call to Ally, you can use it on the Meta Smart Glasses.
Karthik Kannan: Yep, that's one thing. And with regards to Meta, we've got other stuff also in the works that should make it easier. But yeah, the aim is you could go ahead and make a WhatsApp call, or even just a regular phone call. One of the things that the other day, one of the users of Ally mentioned, was he'd like it if Ally could give him a wake up call in the morning. Instead of an alarm clock, he would have an Ally call him and deliver a personalized message for him, and stuff. So those are things that you can do, you can do with Ally in the near future, that a lot of the folks building general purpose assistance won't be able to offer.
Jonathan Mosen: Could you tell me more about recognizing friends and family members? I mean, that is a great use case if you can go into a crowd, and say "I'm looking for a family member or somebody that I know." I'm curious about a couple of things, how's that training process work? And second, is it doing real-time video, or is it taking pictures from time to time, to try and get that information?
Karthik Kannan: So the way it would work, is that it's very similar to how you currently have the Teach Faces feature on the Envision app in the Envision glasses. So you could basically say, "Hey, can you help me recognize this person?" And then you just hand over the phone, or you take five pictures of the person you're trying to recognize in person, and then from that point on, Ally remembers their faces. And yes, you can go ahead and say, "Hey, can you find this person for me in a crowd," or "Is this person in front of me," and Ally would be able to answer that for you.
As of today, it takes discrete pictures, so it takes individual images. But in the near future, what we are looking to do is just make this into a continuous video, so as you're walking through, it just keeps snapping pictures in the background, and then looks for a specific person, or looks for a specific piece of text. And then if it's able to find it, it'll let you know. It's a little bit of a distinction, but the video part is ambitious, probably in the coming year, but the picture stuff is something that we can roll out in the near future.
Jonathan Mosen: When I ask this next question, some people will answer it and some people won't, so this is the big transparency test right here. What large language model are you working with?
Karthik Kannan: We work with a bunch of them. One of the key advantages that Envision has is that we're not really stuck with one particular language model. So unlike OpenAI or Anthropic, we're not beholden to use just one model that we make, so we're able to mix and match models. Because to be honest, today, no one assistant, or no one model, is the clear winner. Different models have different specialties, and some of the models, frankly, are not usable in this particular case because they really are heavily, the technical word for it is nerfed, right? They're heavily, I would say, censored.
So when you ask it descriptions of people, for example, or ask it to read specific pieces of text, they have very high rates of refusal. So we use a mix of open models that we have trained and we have fine-tuned over the years, and we use closed models where they're really good at. For example, we use our own models to get descriptions of people's faces, or get it to read text, or get it to detect things for you. Whereas we use closed models, like Gemini and ChatGPT, if we need to go ahead and determine which particular AI model to use in which particular situation. They're really good at routing things to different models, and so we use them in those instances. So it's a mixed bag, and that's how it's always been at Envision actually.
Jonathan Mosen: Can I use it for usual ChatGPT style, or Gemini style queries, like "Compose me a poem," or search the web for certain information, anything I might be working on, like a project, will it do those things, or is it predominantly focused on blind people in their environment?
Karthik Kannan: It can be used as a general purpose assistant. The idea is that it's a general purpose assistant, but it's heavily built with accessibility at its heart. For example, one of the things that we are working on right now, is desktop and a web version of Ally, where people are able to share their screens with Ally and get help. For example, you're going through a presentation and you want to know what a particular graph is, or you want to know what a table is, and there are images that are there, that you're not able to get descriptions of, or whatever. You could share your screen with Ally, and then just get help, like how you might get help from a customer service person, or a sighted person. And that can be used in any circumstances.
People are experimenting with using it to get descriptions of games, or environments in games which are not entirely accessible. And people are using it to work on their projects, or if they're being sent a PDF, or a PowerPoint presentation that is not accessible, they're using Ally to get descriptions of that, and so on. So there's a lot of use cases that you can do apart from just using it as a blind and low-vision person. It can be a general purpose assistant because of the fact that we use all these really powerful models in the background.
Jonathan Mosen: When I start dreaming about things that I would like an Ally like this to do, some of the first things that come to mind, I think, are fairly simple because they involve APIs, and that would be things like integrating with Microsoft Outlook so I can check my calendar and schedule events easily on my calendar, maybe even scheduling a Zoom meeting or a Teams meeting. And then I go a little bit more ambitious, and I think wouldn't it be great if Ally could hook into something like Expedia, or some travel site, and I can say, "Tell me about all the available flights from BWI to some airport or other, between 9:00 and 12:00 on a particular date," and then if I find a flight, book it. Am I dreaming too much?
Karthik Kannan: A little bit, but it's not outlandish. These are things that will definitely happen. For example, the calendar stuff, that's going to happen before the end of this year. You would be able to connect your Google calendar or your Outlook calendar to Ally, and Ally would be able to pull up information about it easily. So that's something that we're working on already, and that's going to happen before the end of this year.
Now, being able to give you information about things like what flights are available, and so on, that's something that I expect to happen probably by the end of next year, give you information, maybe even help you book that. In the short term, we're also experimenting with things like can you book an Uber ride through Ally, right? Just say, "Hey, Ally, I'd like to book an Uber ride to my house," or something like that, and Ally would be able to give you options, and then you can just go ahead and go through that entire experience with an Ally as through a conversation. Those are things that we can definitely do in the short term, but being able to take actions on your behalf, like booking a ticket for you step-by-step, that can happen probably by the end of next year. The way AI is going right now, and AI agents are happening, we definitely can see that happen very soon.
Jonathan Mosen: This obviously takes engineering resources to build, and any blindness company is so much more than some of these really big tech companies, the AI companies that are rolling this out. Do you worry about things moving so quickly that it's going to be difficult for Ally to keep pace with change once people start using it?
Karthik Kannan: Well, actually, the funny thing is that AI is helping build AI. About 30% of Ally's code base today is actually written by AI. So where we would've hired maybe a junior engineer, a couple of engineers here to do this and that and stuff, Ally is helping us, or AI itself is helping us build it, right? And I expect that to at least hit 50%, or more than 50% by the end of next year. So it's kind of phenomenal what's happening, because every engineer at Envision is actually using an AI sidekick to help them write programs. And what we are essentially doing now, as developers, is where we're spending a lot of time reading the code that AI is generating, making some small adjustments, and then adding them to the code base, so that's helping us move a lot faster. I'm more confident today, of these ambitious milestones than what it was a year ago, simply because we now have AI writing AI. That's pretty phenomenal
Jonathan Mosen: In May, OpenAI gave a public demo of an aspect of GPT 4.0, that at the time we're recording, they have not publicly released yet, which was a real-time video stream. And I wonder when you think that might come to Ally, and what it will mean for the product and for blind people? Because I've got to say, that particular demo was one of the most remarkable things I have ever seen, and I'm a bit of an old tech cynic by now, but that was truly remarkable.
Karthik Kannan: I definitely welcome that. I always believe that there's definitely going to be these general purpose tools that people would use, and then there are tools that we will be building that keeps the accessibility part of it at the heart of everything, which means that we'll make different decisions that ultimately, people might go ahead and choose us over a ChatGPT, or things like that.
And a big part of what we're trying to do, is we're trying to work with companies to see if we can help them adopt something like Ally as part of their software suite so that blind and low-vision people don't have to directly pay for Ally, and we could work with companies to do that. So there are a lot of different ways that we can approach this, and a lot will depend on how the field changes as well. But I'm very confident that any advances that happen from here on out, to the foundational models, to AI in general, will actually have a more beneficial effect on Ally, on the work that we're trying to do here.
Jonathan Mosen: What was the rationale behind creating a new Ally app rather than including it in the Envision app, which so many people have, and it's a pretty strong brand?
Karthik Kannan: I think the main idea is that we wanted to reimagine how user interfaces are built. You've been an old-time tech person, like you're saying. I know you're not that old, you're very young person. You're just touching 25, I can tell that to your listeners. But there was the text-based interfaces of the past, where you had to learn these cryptic commands, and you had to keep all of them in your head, and then you had to type them out in a terminal, and it was very nerdy. And then the graphical user interfaces came along, and people started to use their keyboards and their mouses to point and click, and stuff like that. And that's how software has been built over the last 30, 40 years.
But now it's remarkable that all of a sudden, we don't have to learn the language of the computers. Computers are learning the language of the humans, and they're able to understand how we speak and how we react, and it's able to work based on that. And that meant that we had to reimagine how everything gets built from the ground up. When it comes to, for example, being able to search the web, it would be really nice if I don't have to explicitly give some command, and say, "Hey, can you please search the web for me, and read the top five results and summarize them?" You don't want to do that, right? You just want to go ahead and ask a question like, "Hey, when's the next India New Zealand match happening," and then you'd want the computer-
Jonathan Mosen: I look forward, especially after the last three.
Karthik Kannan: Well, we have a T20 series coming up, and that's our strong suit.
Jonathan Mosen: Right.
Karthik Kannan: We'll see. But I think that's basically what is happening, and that meant we had to really realign and relearn everything, and that's why we thought, "Okay, let's start with a fresh start." And I know the Envision app users , when Ally gets up to speed in a few months time, will absolutely love this experience far more than what we have with the app today.
Jonathan Mosen: So you have a demo there, right? You can show us some of this?
Karthik Kannan: Sure. I actually can go ahead and show you. So I'm going to show you Envision on the glasses. I'm actually going to have two different Allies do this as well. So I'm going to go ahead and first use my first Ally, which I've given it the personality of Batman from the Batman series, so he's a very prim, proper Ally. So let me go ahead and bring him up.
Speaker 4: In your endeavors today.
Karthik Kannan: Could you go ahead and read this piece of text for me?
Speaker 4: Let me take a look. Certainly. The text reads "When we go together, what is the simple form we make," Harriet Mullen. Below it, it features Jane Cortez, Jane Hirschfield, Randall Mann.
Karthik Kannan: Okay, this is actually the poetry magazine that I subscribe to. There's a poem here which is in a foreign language, could you read this for me and help me understand what it says?
Speaker 4: A bit of a linguistic challenge, is it? Let me have a look at the poem.
Karthik Kannan: Okay, go ahead and take a look.
Speaker 4: Let me take a look. The poem is titled Yolkuluk, which means journey in Turkish. Here's a translation: "I have neither a home, my God, nor a garden, nor a boat in the ocean. My head is in clouds, smelling of pineapples. My feet in waters smelling of pineapples. I know the cloud won't give what I hope. I know the ship coming with me..."
Karthik Kannan: Okay. Could you give me a brief description of my footwear?
Speaker 4: Let me take a look. Your footwear appears to be soft, cozy slippers with a yin and yang design, perfect for relaxation and comfort.
Karthik Kannan: All right, thank you. Thank you, Ally. Bye-bye.
Speaker 4: You are most welcome, Karthik. Farewell, take care.
Karthik Kannan: Oh, wait. Before you go, I'm actually going to travel to Amsterdam tomorrow, do you think I should carry a jacket with me?
Speaker 4: In Amsterdam tomorrow, the weather will be clear, with a temperature around nine degrees. A jacket would be advisable to keep you warm. Safe travels, Karthik.
Karthik Kannan: Okay, bye.
Speaker 4: Farewell, Karthik. Till we-
Karthik Kannan: All right. Yeah, so that was a quick demo of Ally, just being able to answer questions, read text for you, describe things, and also even tell you the weather if needed. That's one example.
Jonathan Mosen: If I just comment on that, is that the thing that really stood out for me was there's not that much delay between when it took the picture, when it said, "Let me take a look," and when it started giving an answer, so that is quite an impressive feature of this.
Karthik Kannan: Yeah. I think the idea was to keep it so snappy, because with Ally it really feels like you're having a conversation with another person if we can keep this as snappy and as fast as possible. And to be very honest with you, we can make it at least twice as fast as what it is right now. So there's a lot of work that we can do still to improve this, but as you see right now, it's just as impressive as it is.
And a lot of these things with the Envision app for example, or with the Envision glasses, you have to navigate to a particular feature, like scan text, or instant text, then double tap on it, and then activate it, and then read it for you. Here you just have a conversation, and it just seamlessly works out of the box. The other Ally I have, is it's a little bit more whimsical Ally, and I'm going to go ahead and bring it up, and she can tell me more about stuff that's happening. I'm going to ask it some questions around, let's say, current events that are happening right now. Let me just go and bring her up as well.
Speaker 1: Hey, Karthik, how's the curtain call of life treating you today?
Karthik Kannan: Could you tell me what was the outcome of the 2024 India versus New Zealand series?
Speaker 1: Well, if cricket series were theater performances, New Zealand just gave an Oscar-winning act, sweeping India 3-0. Looks like the Kiwi bowlers had India on a shorter innings than a sitcom pilot episode.
Karthik Kannan: Okay, wait, that's a little bit hurtful, but anyway, I'll let it go. Could you tell me if there are any upcoming matches between India and New Zealand for the rest of the year?
Speaker 1: Well, it looks like after the test theatrics, India and New Zealand are planning to switch to a shorter format drama with a T20 I and ODI series starting November 18th, 2024. Let's hope for some plot twists in India's favor this time.
Karthik Kannan: All right, ciao.
Speaker 1: Catch you later, Karthik. Break a leg, both on stage and in the app world.
Karthik Kannan: So that was Ally being able to give you real-time information, look up stuff for you, and stuff like that. And I've given it this one saying, "Hey, my name is Karthik, and this is what I do," and over time, it's also learned that I like theater, because I do theater on the side. So there's a lot of stuff that it picks up about you and just roleplays to the way you want it to.
Jonathan Mosen: How are you going to monetize this? I mean, there's obviously a lot of resources, despite the AI doing some of its own work, but is there going to be a cost for Ally in the long-term?
Karthik Kannan: Yes, definitely. I think with Envision, we've always been kind of clear that we would like to be as self-sustainable as possible, and that's what helped us be in the business for over eight years now, when we've had other competitors come and go. It will be. We still haven't really decided on it yet, so it will be a subscription fee that people can pay, which gives them access to Ally across all the platforms that it's available on. We'll have more information about that to share in the coming year. But right now, the primary focus is to have Ally available in as many hands as possible, and get as much feedback as possible, and really refine the product itself.
And it will be a freemium model. We're not going to go ahead and say, "Oh, if you don't pay a subscription, you don't get to use it." It will come with a pretty generous free tier for people who really enjoy using it for extended periods of time, and want all the bells and whistles and the tools, things it can do. Yes, it will come with a subscription fee, but we'll talk about that later next year.
Jonathan Mosen: Yes, because it would be amazing synergies with other things. We were talking about APIs earlier, I use a tool called Todoist, for example, which is an amazing task manager app. And I'm thinking, "Well, if Ally could integrate with that, that would be good." But obviously, every time you hook into another API there's likely to be a cost associated with it.
Karthik Kannan: Yeah, yeah, there's definitely a development cost, there's a usage cost. But like I said, I'm quite confident we should be able to cover as much ground as we can. And like I said, at some point, I'm really hoping the AI writes all these different integrations, because they're quite run-of-the-mill software integrations, so I'm hoping that we can just on-the-fly, do this stuff in the near future. Even if we don't have an explicit integration with a particular tool, if they do offer an SDK or an API, can we kind of have the AI just write its own code, and then test it, and then integrate itself with your Ally, and then all of a sudden, it'll go ahead and integrate with the tool that you want to use. So I'm just being a little science fiction-y here, but stuff like that would happen in the near future. It's kind of insane to think of actually.
Jonathan Mosen: When does Ally launch officially?
Karthik Kannan: So Ally is right now available as a public beta. So anyone who is listening to this conversation can just go to ally.me, so A-L-L-Y dot M-E, and then sign up for a public beta of Ally. We're kind of sending invites every single day, so when you sign up for it, chances are you'll get an invite in a few days time. And eventually, we'll move out of a closed public beta, to a wider beta where anyone can just install the app from the app store, and I'm hoping we can do that by the end of this year. From that point on, we just do a little bit of more work, and then Ally would be, I would say, production-ready probably by the first half of next year.
Jonathan Mosen: Mindful that it is a public beta and you're looking for feedback, I was using it in a hotel the other day actually, and I did the classic test of "Which is the shampoo, which is the conditioner, and which is the body wash." There were three identical bottles that the Hilton tends to do, and what was interesting to me was that some of the other tools were able to just give me that info right away. And at the moment, what Ally seemed to focus on was the apple cider in all three bottles, and it concludes that all three bottles were of apple cider. So that was interesting, and obviously, those are the things that you're going out to test for before it officially gets released.
Karthik Kannan: Definitely. And I think those are things that we're also kind of catching and working on fast. For example, in the coming week, we're going to be rolling out the live search tool that I just showed you, and also we're going to be rolling out the ability to read text better than what it is right now. So that should be able to help you answer questions around like, hey, based on the text on a piece of object in the real world, can it determine more things? So those are things that it will be solved quite quickly, and that's what we're trying to be doing with the public beta, like you clearly pointed.
Jonathan Mosen: You mentioned the glasses, and obviously, that's your key platform, how's that going? I wondered if you were feeling a bit disheartened about the incredibly quick adoption of the Meta Smart Glasses in the blind community. I don't think I've seen anything quite like this since the Victor Reader Stream came out in 2007. It's just like they've come from nowhere, and every blind person I run into almost, seems to have them now. Is that frustrating for you, given how long you've been in the business and all the work that you've put into it?
Karthik Kannan: Not really. The way I see it, there are two things here. In the short term, the Meta Ray-Ban glasses are still a very early-adopter kind of a product. It was what Envision was probably when we first launched in 2020, the Envision glasses. A lot of things, for example with the Ray-Ban, still needs to be refined for accessibility use cases. So being able to, for example, even set up the app, or being able to get descriptions of people, or being able to read text without it being censored, these are things that overall, the community still kind of finds it lacking in some sense. But on the other hand, we work very closely with Meta, and we've been...
There's some things in the works, which I can't really talk about right now, but I'm really excited, because when this platform opens up, Envision would be one of the many companies that would truly benefit from it, because all of a sudden, we'd have this tool that we can put our software on, and then a whole bunch of people would be able to use it at a price point that would be interesting for the community.
So I think in the short-term, I still believe Envision glasses and a specialized piece of hardware for accessibility ultimately wins out for the everyday user, for the average user. Whereas yeah, these glasses are definitely very interesting for early adopters, and I'm hoping in the long-term it opens up more avenues for developers like us to build on top of. Meta is just one of them. There's just a whole bunch of companies that are currently building similar style glasses.
And this space, we always knew it was going to go mainstream, and the hope was that it will go mainstream. And now it's mainstream at a time where we're seeing phenomenal advances happening in AI, and I'm just thinking that there's just so many ways we can leverage this. And so it's not like it's come out of nowhere, but at the same time, I know that people have this option as well, and so we need to go ahead and generate better awareness in the community about what the differences are, and people should know what they can expect from each type of a device.
Jonathan Mosen: Because there was some concern expressed in the community when Google announced the discontinuation of their glasses product. So you do have sufficient stock, I take it, and there's no threat to the long-term viability of the product?
Karthik Kannan: Definitely. I think we've worked very closely with Google after the announcement, and we have enough stock in place to handle current and future demand. And it's happening at a pretty opportune moment for us with the inventory that we have right now, and with all the advances that are going to be happening in the coming years, I think we're kind of poised to make that perfect transition from the current hardware, over to future hardware that's going to come.
And having said that, personally, for me, I feel a very deep sense of commitment towards people who currently use Envision glasses. It's not that we are going to just wake up one fine day, and we are going to pull a fast one on the community, and say, "Hey, no, we're not going to support this hardware anymore." The main idea with building an Ally, is that all these advances can come to the existing Envision hardware without users having to really go ahead and update their hardware, or feel like they're working with outdated hardware. And like I said, we're scratching the surface of what's currently possible with this hardware, and we'll continue supporting this well into the future.
Jonathan Mosen: I want to thank everybody who attended our recent boutique at the Center of Excellence and Non-Visual Accessibility, on teaching friends and family to use a screen reader. We had a phenomenal number of registrations for that, and it's good to see. And it's time to tell you, or remind you, about the next CNET webinar event, and this is going to be an exciting one. It could be just in time for the geek in your life, whether it be you, or somebody else, for Christmas, because it happens on December the 18th, and it's the Accessible Device Showcase.
It goes from 1:00 P.M. until 5:00 P.M. Eastern, and it happens via Zoom. Whether you're buying for yourself or searching for the perfect gift for family or friends, it can be hard to find consumer electronic devices that are accessible to blind people. The National Federation of the Blind, in conjunction with the Maryland Department of Disabilities, is hosting an Accessible Device Showcase.
Now, this showcase is going to help you write a gift or personal wish list with the latest accessible consumer electronic devices in mind. From smart homes to leisure activities, we'll present video games, smart home devices, mobile applications, toys and games, home appliances, and a lot more. You can register for the December 18 Accessible Device Showcase today, just head on over to nfb.org/cena, that's nfb.org/C-E-N-A, for more information, and be sure to join us for what promises to be a fun and informative four hours, 1:00 P.M. Eastern on December the 18th, head over to Nfb.org/cena.
This email is from Jessica Dale, and she says, "I listened to the first episode of this brand new podcast, and I have to say I'm very impressed with it." Well, thank you, Jessica. "The quality is great, as are the demos and other topics which have been discussed thus far. Today though, I am writing with a question of my own, which I'm hoping someone can assist me with. I recently started a podcast of my own."
"Depending if I'm on the go or not, I either use Reaper on my PC, or Ferrite Recording Studio on my iPhone, to edit. My question today pertains to Ferrite Recording Studio. When using the app with voiceover, how do I move a track? I have three or four separate tracks which need to be placed in a specific order. Often, when I attempt to create a new episode within Ferrite, the tracks will end up in the wrong order. Does anyone know how to fix it using voiceover? I'm running the latest version of the app under iOS 18.2 on my iPhone 15 Pro Max. I also have an iPad Pro 11-inch from 2021 with Ferrite installed and have the same problem there. Thanks in advance for any advice you can provide."
Thanks, Jessica. I would observe that although it should be technically possible, I've not yet found an app for iOS that gives you the same efficiency of editing as several apps on Windows and Mac. What we really need is good quality keyboard shortcuts. Actually, the one that comes closest is the other app from the same company that makes Ferrite. The company is called Wooji Juice, by the way, but I think it's just a single developer who operates under that company name, and he has another app called Hokusai, which has very good editing from the keyboard.
So if you use a Bluetooth keyboard, it's pretty efficient to do editing on there, that is the best that I have found. Ferrite, in my view, is a little bit more fiddly because you've got to make items and delete those items, and do all sorts of things like that. So maybe someone has an answer for you there. If it were me, I think I would still do the recording on the phone if needed, but then bring it into a digital audio workstation on the Windows or Mac, and do the editing there.
But perhaps there are others who are using Ferrite every day for their podcast production. Jessica also points out in that mail, which I slightly abridged, that she's using the new call recorder feature. And this is one very good feature of iOS 18, which is that you can now record phone calls. I use this all the time when I'm calling businesses, and I've been calling a lot of businesses lately in association with move and travel, and they'll often say "Your call may be recorded for quality assurance." Well, I record them for my quality assurance as well, and I've got a record of when they've made promises in case somebody chooses not to keep them. So that's good, but it also has a podcasting use case as well, because you can use those call recordings in a podcast, so it does lower the barrier to entry to some degree there.
You could also contact Wooji Juice itself. In my experience, the developer of Ferrite offers fantastic and detailed replies to emails. And this is often one of the benefits by the way, of dealing with some of these smaller developers who write the product, support the product, do everything for the product, if you get a good one, they can be extremely helpful. And I know that he is committed to voiceover over. Over the years, I have pointed out one or two little glitches with voiceover support, he's always fixed them very quickly. So if anyone has any thoughts on this, [email protected], if you want to be in touch and help Jessica out with her Ferrite question.
Well, it appears Thunderbirds are go. That's dating me, that was a great old series, FAB. But many people may not know what I'm talking about, so I'll just talk about the Thunderbird. Christopher Wright says, "I use Thunderbird 1.33 in Windows 11 with NVDA and it works great. I've disabled most of the extra stuff, like the welcome page that appears in the main window. I can press tab or F6, and shift+F6, to move between the folder list, message list, and other things, like the spaces toolbar and search. I wish I could hide everything except the folder and mailbox lists, but I can't, and I can't figure out where to submit feedback to Mozilla that will be prioritized and taken seriously. The only minor gripe I have is that the keyboard focus isn't automatically placed on the message list when I open the program. I have to press shift+F6 to move there. Once again, I wish I knew a person, or the appropriate channel at Mozilla, so I can report this issue."
If you're not used to the regular releases of Thunderbird, consider the LST channel, just like Firefox, it gets major updates, but once a year. So if you don't want things to change frequently, it's a great choice. The current version is 1.28. And George says to the listener wondering if Thunderbird Nebula was accessible, "I'm using it now with the latest version of NVDA with a Thunderbird G5 add-on from the add-on store. I have to say it's absolutely spectacular and fantastic. Nothing seems to have changed much, and I'm using it to send this email. I've also got a couple of questions of my own. I'm moving into communal accommodation in a couple of months. I've decided to get a talking microwave from Cobalt Speech Systems. I know some people regard it as patronizing, but I don't really have the time nor the patience to guess which button does what. Whenever I face this dilemma, I always have this little voice in my head telling me that I shouldn't buy these products because other blind people will see me as less independent if I do. Is this true?"
That's from George who was in South Africa. George, I got a quick reply for you, and it's this, live the life you want, use what works for you, and if sighted people have a screen which displays information, it seems reasonable to me that an accessible option should exist that speaks that information. In the United States, there are a number of talking microwave appliances, but the difficulty with exporting those, is that they only use 110 volts, and many parts of the world use a different voltage, at 220 volts, and some of the microwaves that talk here in this country, in the United States, are not available in other markets because they just haven't been produced with that voltage. So your choice is a little limited, and the Cobalt Speech System's one has been around a long time. It's got a pretty good reputation, and it will get the job done for you, I'm sure. Good luck with your move.
Abby Taylor is writing in on the topic of Facebook accessibility. She says, "Hi, Jonathan. Congratulations on your new podcast, I've been enjoying it so far." Thank you, Abby. "I appreciate your efforts to resolve accessibility issues with Facebook since the mobile site for PC was discontinued. In case you're not aware, here are some issues I'm having with Facebook, both on the iPhone and PC. On the iPhone. When I get a notification that one of my friends has posted something, when I double tap on it, if the app is already open, nothing happens. The only way I can view the post is to triple tap the notification to bring up the context menu, double tap, preview. Of course, the notification still needs to be cleared from the notification center. On the PC, when I receive an email saying someone has posted something, once I click on the view status link, NVDA often won't read what's on the page, even though with my limited vision, I can see the post is there."
"I'm using Google Chrome, when I tried reading the same page in Firefox, NVDA will read what's there, but only one letter at a time, and each letter shows up as a link. Also on my PC, when I tag someone, or check in, after I search for a person or place, NVDA won't read the results, although I know they're there. I haven't tried this with Firefox. The only workaround for these issues is to use the iPhone app, which works well for these purposes. Most of the time this isn't a problem since I usually have the iPhone handy, but I hope these issues can be resolved soon. Thank you again for your efforts."
Thanks, Abby. And of course, we always must be mindful that there are people who choose not to have a smartphone, and that is a perfectly valid choice, and we want to make sure that their choice is respected and that they can access the public discourse that goes on on social media, with the tools that they have.
And Mike Calvo was writing in with a variation on this theme. He says, "Jonathan, congratulations on launching the Access On podcast with the National Federation of the Blind. It's great to see you bringing your voice to this new platform." Thank you, Mike. "Also, congrats on your move to the United States. In light of your recent coverage of Meta accessibility challenges with Facebook, I wanted to share an experience with Instagram. I set up an account to promote my public speaking work, but due to accessibility challenges with the platform, my sighted executive assistant in another state, managed the account, Meta closed the account citing a violation due to the geographical difference.
Despite several appeals explaining the situation, I have no response. The account is still closed. I am lost as to where I go next. This issue likely affects many blind individuals who need sighted assistance to navigate time-consuming and inaccessible platforms like Instagram. I hope you and the NFB can address this, as it underscores the need for accessible social media solutions."
Thanks for writing in on this, Mike. And clearly, if the accessibility is such that you require sighted assistance to manage your presence on a social media platform, what we're talking about here is an accommodation. And I appreciate that it may be a violation of the terms of service, because what you're effectively doing is sharing your password, and I'm sure that the terms of service will have something to say about that. But as you say, in this instance, there's a clear accessibility use case, so we will add this to our list of items to raise with Meta, and draw this to their attention.
We are going across the pond to the UK for this email from Ian Harrison, who says, "It's great to track you down to your latest podcast. You can run, but you can't hide." Scary Ian, you stalker. "I've had the Meta glasses since October, and have some pros and cons for you to hopefully publish. First, the bad news. One, in the UK, there is utter frustration over whether the AI function is officially available or not. I've achieved it both with or without a VPN, and the service seems to come and go randomly. I get nervous when a Metaview update turns up, in case it breaks my connection. Two, this is a purely personal thing, but the range of colors, styles, and lens types are rather limited on the official Ray-Ban site. I had to go to a third-party lens provider to get the transition lenses with a brown tint, for example."
"Once I changed from the permanent brown tint the glasses came with, I found myself wanting to wear them the whole time. The extra lenses cost another £85, and were relatively easy to change on the frames, and I can now use them indoors with what little light perception I still have. Now, the good news. One, apparently they actually look good on me. Two, the microphone and audio capabilities are great. Being able to have my ears unclogged while listening to VoiceVista prompts when out and about, feels much safer. Three, not having to drag my phone out in the open all the time is a real bonus, and I can't wait to be hands-free when I need to call Be My Eyes in the future. I'm sure there is more to come from Meta and competitors in this growing new market."
It's a pleasure to hear from Larry's Skutchan. Larry, I hope that retirement is treating you well. Larry has made a huge contribution to access technology. He says, "Hello, and congratulations on your new position at the National Federation of the Blind. I think this is a super expedient move for both you and the organization." Thanks, Larry. I appreciate that. "In your last podcast, the listener asked about any accessible induction hot plate slash burner. I have used this one with much success, the buttons are all accessible. Unfortunately, the fan noise is so loud you cannot hear the food cooking. Therefore, I don't necessarily recommend it, but it is usable." And the product that Larry refers to is called the Duxtop, which is spelled D-U-X-T-O-P, 1800 watt portable induction cooktop countertop burner, and it has an Amazon Choice label associated with it.
So thank you, Larry. And this is a wonderful example of what we are seeking to achieve here on Access On, someone asks a question and we have another blind person helping somebody out with an answer that is working for them. If you've got any thoughts on the whole induction cooktop question, do be in touch. What are you using that you find accessible? [email protected] is our email address, that's accesson, all joined together, @nfb.org. And when you get in touch, if you happen to be coming in from a chapter of the National Federation of the Blind, do be sure to tell us that so we can give a shout out to that NFB chapter.
Let's talk Braille displays and note takers. This email says, "Hi, Jonathan. My name is B. Oliveri, and I'm from the Northeast chapter of New Jersey. As a proud federationist, I'm glad you are hosting the technology podcast for our beloved federationists." Thank you B, it is a pleasure to be here doing this. "Jonathan, like two years ago, you spoke about a Braille device complete with Braille display and a Braille keyboard that was around 2,500 dollars, and allowed for the use of traditional apps as well. You said you use it yourself and enjoyed it. What was that device called, please?"
It could be one of two things, B. There is the Brailliant, which I did evaluate on a podcast a couple of years ago, from Humanware, and I'll have more to say about that in responses to other parts of this email. There's also another product that's being promised by Orbit Research and Access Mind, called the Optima. And I haven't heard how this is going, but the idea of the Optima is that it will be a fully-fledged Windows computer with a Braille display. So if it works well, you might be able to discard your traditional laptop, just take your Braille display with you, carry it around on your shoulder, and you'll have the full Windows experience with some proprietary applications as well, and just use that everywhere. Now, I don't know how that's going, what the timeline is for production, but it will be interesting to find out more about that.
B's email continues, "Also, I need something comparable to Humanware's Braille Note Touch Plus or the Braille Sense, for work especially, but also for my personal use, but can't afford their prices. Can you suggest something comparable and more affordable?" I guess it depends on what you want to do with it, B. if you just want to take notes and meetings, then Humanware and Salvis, which used to be called Hymns, have come up with products that are a lot cheaper than the traditional note taker, with some basic notes taker functions. The Brailliant is the range from Humanware that does this, and you also have this new product, which we've just procured for the International Braille and Technology Center here at the National Federation of the Blind's Jernigan Institute, and that is called the Emotion. So you might want to check those out and find out whether they meet your needs.
If you need things like web browsing and email, then that might not be enough for you, but certainly, just for note taking functions, and then you can take the notes back onto your computer and work with those, read documents on the go, that may be sufficient. B says, "Also, are there any grants around for us blind folks to use toward technology purchases, or a credit card for same?"
Thanks, B. Well, I guess any credit card should work if you're willing to go that route. And let's just open it up, and this may vary from state to state, of course. So if there are people who are aware of programs that can help somebody to pay this technology off, maybe it's something that the individual companies are starting to offer now, then do let us know what your experience has been with trying to procure a product like this, that can make such a difference. You can email us at [email protected], accesson, all joins together, @nfb.org. And please feel free to do what B did, let us know what chapter of the NFB you're coming from, if you are in one, and we can give a shout out to that chapter.
This next email says, "Hey, Jonathan. My name is Lexi. I am a junior in high school, and a longtime listener of the former Living Blindfully Podcast. When I heard you were not doing Living Blindfully anymore, I was really sad." Oh, I'm sorry to make you sad, Lexi. "I thought that was the end of your technology podcast career. But one day, I was downloading the remainder of your episodes, when I saw your new podcast trailer. I was so excited, and I waited and waited, and finally, your first episode of Access On came out. Thank you so much for doing this podcast."
Well, thank you for listening, Lexi, there wouldn't be any point otherwise. "I can't figure out for the life of me why my Live Listen is not working. Is there any tip that you can give me to see if I can get it working again? I don't have Air pods, but I feel that my safety is very important, and I don't have an ambient mode feature." Well, Lexi, there are only certain devices that Live Listen supports, and it's not clear to me, what you are trying to use Live Listen with, and whether it used to work before and isn't working now. So not too much to go on here, for me to give you any advice, other than to say give Apple a call and see if they can assist you. Given that this is Live Listen, I think that is touted as an accessibility feature. For those who aren't familiar with this, Live Listen allows you to hear, using your microphone, what's going on around you.
I know that there are people who use this in a meeting environment, and it's sufficient for them to put a phone in the middle of a table, for example, and they can hear all speakers around the room if they have a moderate hearing impairment. For people with more profound hearing impairments, it may be necessary to use a specialized device for this, but Live Listen can help in many situations. So I'd suggest giving Apple accessibility support a call, and see if they can sort it out, particularly if it worked before. If you're saying you had something that worked and now it doesn't, I'm sure that Apple would want to know about that.
We love to get your tech tips, and if you have a tech tip, something that you think maybe not everybody knows about, and it really just speeds you up, makes you more efficient, then please send that in to [email protected], that's [email protected]. And if you are a member of the National Federation of the Blind, be sure to tell us what chapter you're coming in from so we can give a shout out to that chapter. And today's tech tip relates to navigating between open applications in Windows. Most people know that if you want to switch applications in Windows, you can hold down the alt key and press the tab key. So keep the alt key held, down press tab, and you'll be able to cycle through the open applications.
But there are also some other options that you might try. Instead of using alt tab, you can press Windows tab, and this works in any Windows screen reader. When you do that, you'll be placed in a list of applications. So you don't need to hold down the Windows key as you press the tab key, just press Windows tab, you'll be in the list, left and right arrow through the open apps, and just press enter on the one you want, sometimes that's more convenient. There is a downside to this feature that's built into Windows though, and that is that at the moment, at least in Windows 11, you can't use first-letter navigation. That's where a JAWS version of this feature may come in handy. You can press the JAWS key with F10 and bring up a list of applications, and type the first letter of the application that's open that you want to switch to, and then press enter. Hopefully, Microsoft might restore first-letter navigation into this feature sometime soon.
In both cases, with this feature, and this didn't use to be the case with JAWS, but it is now, you can press the delete key to simply close the application. Sometimes we find that a Windows application is hovering around, and when you press alt tab, you don't hear anything. It's like it's blank, but there is an application there, and it can cause some disruption to your computing experience. So going with the JAWS key and F10, and pressing delete on that silent application, can sometimes tidy things up. So there are a couple of alternative ways to switch applications in Windows. Again, we'd love to hear your tech tips, [email protected].
That concludes this episode of Access On, the technology podcast of the National Federation of the Blind. To send in a contribution for a future episode, email us, attach an audio clip, or just write it down and send it to [email protected], that's [email protected]. To keep up to date with Access On, follow us on Mastadon, [email protected], that's [email protected] on Mastadon. To subscribe to an announcement-only email list about upcoming episodes, send a blank message to [email protected], that's [email protected]. To learn more about the National Federation of the Blind, visit our website, nfb.org, or phone us, (410) 659-9314, that's (410) 659-9314. And be sure to check out the Nation's Blind podcast right from where you heard this podcast.