Audio
Wearables compared, ChatGP, a launch, and news of David
Assistive tech developments - including Meta Ray-Ban vs ARX vision wearables.
Vision Australia's Access Technology Officer Damo McMorrow talks with Stephen Jolley about latest tech developments from a blindness and low vision perspective.
Vision Australia has a range of products and services to assist Australia’s blind and low vision community. For more information call 1300 847 466 or browse online our Vision Store range of products.
You could also email our program.
In this episode:
- An experience of using Chat GPT to access a PDF user manual;
- Discussion and comparison of the Meta Ray-Ban smart glasses and the ARX vision wearables;
- News of David Woodbridge joining Humanware Australia;
- News of the Australian launch of the Monarch Tactile Braille display.
The picture on this page shows an imaginative idea of high-tech vision wearables. They don't really look like this!
00:33 S1
Hello everyone. Welcome to Talking Tech. This edition available from October the 8th, 2024. I'm Stephen Jolley. Great to have you with us - listening maybe through Vision Australia Radio, associated stations of the Radio Reading Network or the Community Radio Network. There is also the podcast. To catch that, all you need to do is search for the two words Talking Tech and an icon will come, usually on a Tuesday afternoon just after it's been produced. Another option is to ask your Siri device or smart speaker to play Vision Australia radio talking tech podcast. Vision Australia radio talking tech podcast.
With me, Vision Australia's national access technology manager Damien McMorrow. Hey, Damo.
01:18 S2
Hi, Stephen. How are you?
01:19 S1
I'm very well - and enjoying the technology season that we have at the moment, with so much happening - not just in the Apple world, but other things happening. And one of those is in ChatGPT, and you've been playing with that a bit lately...
01:32 S2
Yes I have. I had a situation a week or so back where I had bought a handheld VHF marine radio for my boat, and the manual, as tends to happen with these things, wasn't particularly accessible. You could OCR the PDF, but anywhere where they referenced a button, it just read it as a graphic or not at all. So you would hear the to adjust the volume, press the button. And I thought, well, that's not particularly useful. So, I discovered that we could upload a document to ChatGPT. So you open open it... in a web browser, you may be able to do it from the app, but I just did it from the, from a browser, and next to the edit box where you normally can type in and ask it questions, is a button that allows you to browse for a file. And I just browsed to the PDF and uploaded it, and then I was able to ask questions in the usual way.
So by typing into the edit box, how do I adjust the volume? How do I enable dual watch mode? How do you charge the radio? Those sorts of things. And I was able to get, you know, quite intelligible and sensible answers from it. So basically what it does is, it processes the document and then allows you to ask questions about that particular file or that particular document. So really handy if you've got, you know, a mainstream product, perhaps where you've downloaded the user manual from somewhere and there's all kinds of annoying graphics that representing the buttons, and they don't work very well with the screen reader. It gives you a different way of being able to sort of interpret that information and, and get the, the things that you need.
It's also good because you can just ask it for the information that you want, rather than having to read through things in multiple languages. I had an instance yesterday where I'd bought a new coffee machine and every page had the same instructions in German, Italian and English. And so again, reading that with a screen reader was a bit of a nightmare.
03:45 S1
So you did this from a web page rather than from the ChatGPT app, is that right?
03:51 S2
That's right. Just ChatGPT. Com the easiest way to find the button is to find the edit box, where you normally would type in your questions and then shift tab back once there's a button there that allows you to browse for a file.
04:07 S1
So do you sign in via your account to do that?
04:09 S2
Yes. You do, yes. So you still need an account, but it can be a free account. It doesn't have to be a paid account to use that facility. Yeah.
04:17 S1
Very interesting. Let's move on. Now... our conversation last week was about the the Meta Ray-Ban smart glasses. And you've got some further thoughts about that.
04:30 S2
Yes, I've had a few questions during the week. You know, people sort of saying, well, there's the Ray-Ban Meta glasses. There's the ARX Vision headset. Which should I go with? What do they do? Well, which one would you choose? All of that kind of thing. So I just thought it might be worth having a little bit of a discussion around that... given that the number of questions of that sort that I've had. So the ARX Vision headset is, for those that haven't come across it yet, it looks like a bone conduction type headset.
So... the headband goes around the back of your head and there are two earpieces which, rather than covering your ears, actually sit just forward of your ears and it conducts the sound through the bones of your face. And then on the right hand side of the headset is a camera module, which is a small rectangular box. The camera points forward and there are three buttons. There's a triangular button, a square button, and a round button to control the headset, and it plugs in via USB-C to your phone. You run the vision app and it has a number of different modes. So it has a scene mode where it pretty much gives you sort of a running commentary of whatever the camera can see.
It has a short text mode and a document mode, which people who use seeing AI will be familiar with that. It sort of will scan the document and read it to you. It's essentially just an OCR function. It has a barcode reader, and it also has a what they call Q mode. And it's a bit like what we were talking about before with ChatGPT. So you can scan something like a menu, it'll give you guidance as to where to where to hold the page. And then you can ask it questions like, you know, how much are the chips? Or what are the what's the cheapest thing on the menu, that type of thing.
There are some similarities, I guess, between ARX vision and the Ray-Ban Meta glasses, but there are also some fairly significant differences. I have had an extensive look at both and tried them in different situations. In terms of which one would suit or which one I'd recommend, it really depends on what you're trying to do. So I think that the the Vision is much better for document reading because the... Meta glasses always want to summarise rather than read the document from beginning to end, which can be a little bit frustrating.
The scene mode is quite nice in that it does give you a sort of a running commentary. I was walking along the, um, near my my home and it said... a man riding a bicycle along a sidewalk. And I thought, Oh, I wonder what that's actually saying. And then three seconds later, a guy on a bike shot past me and rang his bell. So it was actually correct. So having that information there at your fingertips, as it were, is quite, quite nice. It can be a little overwhelming though, because it does give you sort of a constant stream. Whereas with the Meta glasses you have to you have to prompt it and ask it, you know, hey, Meta, look and tell me what you see.
So it sort of depends on whether you like that constant stream of information about the visual environment or whether you just want it when you when you query it. So it kind of depends, I think, on whether the AI capability is more important to you or whether you really just want the document reading capability. But the other part of the equation, too, is that the scene mode with ARX vision does interpret. So it'll describe what it can see, but it will also read text. So you don't have to sort of prompt and say, Look and tell me if there's any text. Look and tell me if there's any signage. Basically if the camera can see it, whether it's text or other objects, it will read it to you.
So I think it kind of depends on what you what you want it for at the moment. At the time that we this goes to air, the ARX vision is only Android compatible. We will have an iPhone adapter very soon for it, but at this point in time it's Android only... and it is a cable connection. So that's something that might be a problem for people, having to... have that wired connection to your phone.
08:34 S1
One plus with the Meta glasses is the capability of having a friend or an agent through Be My Eyes talking to you and reading what you're seeing.
08:47 S2
That's right. And being able to do that through things like WhatsApp and so on. The only other consideration is that obviously the Vision is a specialised device and is therefore NDIS Consumables eligible, whereas the Meta glasses would be regarded as an everyday item, so you wouldn't really be able to get them funded through NDIS. So that's something that might also be a consideration. So my advice would be to, you know, if you're if you're not sure, think about what it is that you're you most wanting to use it for. And also whether the wired connection is a problem for you, whether you want just the AI capability or whether your main use case is document reading and have a look, have a good look at them.
Pop into a Vision Store. Take the Vision for a test run, head into your local sunglasses shop and have a look at the the Meta glasses and choose what works for you.
09:42 S1
I think that's very helpful what you've been explaining there over the last few minutes. A shout out to David Woodbridge, who was my partner in crime for this program for 12 years. David is now with Humanware Australia as a blindness product specialist, so we'll be seeing him around the traps quite a lot, and we'll be talking to him from time to time. But he'll be a Humanware authority on things like the the Monarch, the Brilliant range of devices, the Victor Stream and the Stellar Trek devices. And the Monarch is actually going to be front and centre this week.
10:19 S2
Yes. So there is a launch event happening on the 10th of October. That's at 11 a.m. Australian Eastern Daylight Time, because of course we're in Summer Time now here in Australia [NB: Not in Queensland]. Humanware in conjunction with Quantum are running that one. And the Monarch has been a sort of a long-awaited product. We've heard lots about it over the last 12 months, and so it'll be nice for it to be officially launched here in Oz. For those who... aren't familiar with it, it is essentially a Braille tablet device, so it's a multi-line Braille display that can also display tactile graphics.
And if you're familiar with the sort of Kingsoft suite of applications that exist on the Brilliant and the Braillenote touch products and those kinds of things, a lot of those applications are present on the Monarch, but we do also have the Tactile Graphics viewer. The unit itself is about the size of a large gaming laptop. So sort of a 15 inch gaming laptop and around about the same sort of weight. But... it is an exciting product, I think, particularly for... people in doing those sort of Stem subjects - science, technology, engineering and maths and those sorts of things, or people who just want to be able to read more than one line of Braille at a time.
11:36 S1
It's going to cost a bit.
11:37 S2
It is. Yes. The the last figure that I had heard was in the order of $27,000. So certainly not cheap, and not something that you would purchase lightly. But I think, you know, for those who need that sort of functionality where they're dealing with high level maths or graphical concepts or, you know, wanting to take a lot of... Braille material with them on the go. It may be a good option, particularly in the education space I think. Mm.
12:08 S1
Yes. So that's this Thursday, the launch of that device in Australia. Does that mean it's going to be available for sale straight away?
12:15 S2
Yes I believe so. Yes.
12:18 S1
Great. Yeah. I want to have a quick word with you about a very useful Vision Australia service, the Help Desk, which comes under your portfolio.
12:26 S2
It does. It's one that's quite close to my heart. So the Help Desk is there to provide phone and email support for access technology queries. So it could be anything from What's a good accessible phone option? to My Jaws has suddenly stopped speaking. What do I do? All of those sorts of access-tech-type questions, general advice, that type of thing. And the team is fully staffed by people with lived experience. So I have two people on the team who are full time screen reader users, two who are full time magnification users, although all of them are familiar with all of the products. And their team leader Jim Piepszak has been in the access tech space for a very long time as well.
So there's a lot of a lot of expertise there, a lot of passion in terms of, you know, using the tech in their in their daily lives. And they're also very, very diligent. If they don't know the answer, they will do the research and come back to you if they can't solve it straight away. But we obviously try to solve as much as we can, with the customer on the phone.
13:33 S1
So the way to reach the Vision Australia Technology Help desk is...
13:37 S2
To ring the Vision Australia main number, 1300 847 406. That's 1 300 VISION - and just ask for the Access Technology Help Desk, and it's available. 8:30 a.m. to 6 p.m. Australian Eastern Daylight Time, Monday to Friday.
13:56 S1
1 300 847 466 ... 1 300 VISION. Before we go, a reminder of where there are details of what we've been talking about in this and previous editions of the program, you can just go to VA radio.org/talking tech. VA radio, dot org slash talking tech and to right to the program.
14:19 S2
Damo... damo.mcmorrow@visionaustralia.org ...
14:27 S1
Damo dot McMorrow at Vision Australia dot org. This has been Talking Tech - with me has been Vision Australia's national access technology manager Damo McMorrow. I'm Stephen Jolley. Take care. We'll talk more tech next week. See you.