Introducing Envision Assistant Beta

By Karthik Mahadevan, 17 April, 2024

Forum
iOS and iPadOS

Hi everyone,

This is Karthik from Envision. About 6 years ago, I came to this forum to post a beta of a new app I was working on called Envision AI (link to my original post). We got a lot of love, but more importantly, we got great constructive feedback that helped shape the future of Envision AI, which now has more than half a million users worldwide.

Six years later, I come before you again, with another beta of a new app that we’re trying to build. Its working title is Envision Assistant (not catchy enough, we know). It’s a conversational, personal and ubiquitous assistant built from the ground up for accessibility.

It’s still in its very early conceptual form, but we would love to invite you to try out a beta of it so you can help contribute to our vision for it. I’m leaving some useful links below that will help understand our vision for the assistant better and link to the form to sign up for beta:

  1. You can read all about Envision Assistant here: https://envision-assistant.com/
  2. You can watch this short video of a beta-tester using this at CSUN Conference: https://youtu.be/U01onUkc-o0?si=U9pJMYuedyaMEPfH
  3. You can sign up for beta using this form: https://beta.envision-assistant.com/

The beta is currently iOS only, but we will be launching on other platforms (Android and Web to begin with) shortly. I am excited for you all to try this and look forward to listening to your feedback. If you have any doubts, clarifications or questions, please let me know here as I will be monitoring and responding to the comments.

Options

Comments

By Brad on Monday, April 15, 2024 - 09:43

I've removed the app, not because it was bad it's just that i don't think I'm the right person for it.

I think the one thing i've realised when using these types of apps is that I'm not a photo taker.

I don't go out enough to get anything really out of images so I hope others can get a spot now I've removed myself and can enjoy the app.

The voices are interesting. I played around with them and it does seam that 11labs or whoever you've gotten the voices from do use one sentral voice and branch off from there, I say that because some british voices pronounce some words in an american way, having said that; it might just be a bug with the app where you change a voice and it actually doesn't change, you'll have to see what others say.

As for the personality, it's interesting; if a bit, well, artificial. I told the app that I liked Harry Potter, food, and am trying to lose weight and it uses that stuff all the time, I did say it could pick its name and personality but I don't think the AI is that advanced just yet.

So I think i'll not be using these picture taking apps for quite a while if at all, they don't serve much of a perpus for me in my day to day life.

By mr grieves on Monday, April 15, 2024 - 09:43

I think the visual side of things becomes a lot more useful when paired with a wearable. I am not going to walk around with my phone outstretched. So whenever I get one of these I just take a photo of my dog and see if it can figure out. This one did recognise her as a border terrier but I did say I owned one so I guess it had an educated guess.

I presume the Envision glasses is where this is ultimately heading which would make sense.

I think getting used to talking to something instead of using an app with swipes and taps is something I struggle a little with. I remember when I first got an Echo I felt weirdly shy talking to it until I got used to it. But I have that with the AI assistants. I just don't really know what to ask. So as above it becomes a bit of a plaything. In a funny way it's almost worse that the voices are so life-like. I've got used to barking out commands to the Echo but now I feel like I am supposed to be striking up a conversation.

I watched the video last night and thought the sarcastic comments were incredibly annoying. But interesting in that I wonder if this is positioning itself as if you were carrying around a friend all the time rathe than using a tool. I'm not sure I see AI that way and was pleased that I could make my Envision AI Assistant as boring as I am!

Well, I guess the main thing is if you are trying to actually get information, having something constantly wise-cracking would make it much less useful. But having said that it is cool you can do that.

I seem to have collected a lot of these assistants and still haven't figured out to do with any of them. I was trying Pi last night and was asking it when the football was on and how I could listen to it and it just constantly lied about it which rendered the whole exercise a bit useless.

I'm still waiting for the penny to drop. But it is fun messing about.

By Brad on Monday, April 15, 2024 - 09:43

I think the issue, at least for me, is that the tools say something like, it's a pizza with blah blah blah, in other words, they go on way to much.

I just want to be able to ask, does the pizza have olives on it and hear a simple yes or no but I don't think we'll get that with these moddles these app devs use.

It's great that some people want and can get all the details and I would too if it wer a video game or something I actually cared about but my room or a pizza, no thanks.

By Enes Deniz on Monday, April 15, 2024 - 09:43

There are multiple reasons for that, some of which may not be equally valid for everyone.
  1. I live in TĂŒrkiye and we use the Turkish lira as the currency. The problem though, is that just one US dollar is worth over thirty two Turkish liras so I can't really subscribe to anything. I know there are countries where exchange rates are much higher and it is practically impossible to purchase anything for sale primarily with the American or European customer in mind. To be honest, I wouldn't be willing to test an app that I wouldn't be able to use anyway. I can guess that Envision Assistant will remain free at least for a while, but knowing I might lose the opportunity to use an app to which I have contributed quite a bit, doesn't feel good.
  2. Envision AI has also recently become free and it used to be paid so the company is indeed able and willing to fund its business through other sources of revenue like the Envision glasses.
  3. The beta-testers who will actually have given feedback will likely not make up a large portion of all the users, given that the app is already expected to have support for languages other than English soon, which will rapidly increase the number of its users as the company, the Envision AI app and the Envision glasses are already known among the so-called blind community.
  4. I think I can really say I have already provided a considerable amount of feedback and will continue to do so, and I will recommend the app to others as I test and use it. In fact, I have already recommended it to a friend who previously didn't know about it.

By Gokul on Monday, April 15, 2024 - 09:43

I guess if apps like this are useful at the end of the day is very subjective. But primarily, where I see these making an impact eventually is in making workplaces accessible. most workplaces that one could think of are primarily designed with visual info at the core (and I'm not talking about the physical infra alone here). So once visual info becomes accessible to a degree, it is sure to lead to better workplace acceptability, better incomes and hence better control over ones life. I say this because I work here in kind of a leadership role here where any visual info I have adds to my power and control over the team.
So I've got this CCTV here in my chamber with several live feeds which otherwise would have been quite useless to me. But yesterday I decided to try and take a photo of the monitor with PiccyBot and as it happened, it gave me info which I could comprehend along with my context awareness, which then lead to me having more control over the stuff that was going on. this is why I keep saying that hopefully these devs focus on increased community involvement, using existing capabilities to tackling day-to-day practical problems, and of course, on workplace accessibility. Because, that kind of thing can have a multiplyer impact on the lives of blind people, their families, and on how blindness is understood and imagined. And it should work as a good business model as well given companies, governments, and players of that magnitude will be capable and ready to spend on these solutions which will hopefully help the devs to keep the tech free for needy intividuals.
And as for the amount of info that we have around us to look at, I guess it is more or less determined by our imagination.

By peter on Monday, April 15, 2024 - 09:43

perhaps it is possible to customize the type of feedback you get using the personalization settings. If you can ask for "humorous or lighter feedback", "detailed feedback", you might be able to ask for "short overview or summary". Try experimenting and see what happens. If this works, it is a great feature that will enable users to customize what type of output they prefer.

--Pete

By Karthik Mahadevan on Monday, April 22, 2024 - 09:43

Greetings everyone,

I'm delighted to share that the response from AppleVis has been exceptional. Within just 24 hours, over 200 individuals signed up through our beta form, far exceeding our expectations. Thank you for your overwhelming enthusiasm! We have already issued around 150 invitations.

Due to the need to scale our services to manage this increase in activity effectively, we are temporarily slowing the distribution of further invites. Rest assured, we will resume expanding our beta testing invites gradually, which may take a week or two. For those who have recently submitted the form, please be patient. If you haven't yet signed up, I encourage you to do so as the invites are being sent on a first-come, first-served basis.

By Karok on Monday, April 22, 2024 - 09:43

hi how do we know when we will be accepted? what am i looking for? just wondered where the email would come from. was hoping to be selected now as need this for work if on the web soon. thank you though for all you are doing. Will

By Sara on Monday, April 22, 2024 - 09:43

Hello everyone! 👋 I have a suggestion for the beta testers: what if we create a WhatsApp group to chat and share our feedback with the developer? While we have a Slack workspace, some of us might find it easier to use WhatsApp since we're more familiar with it.

Of course, I don't want to speak for everyone, but I think this could be a great way for us to communicate and provide valuable insights to the developer. What do you all think? And more importantly, what does the developer think about creating a WhatsApp group for us beta testers? Thanks for considering my suggestion! 😊

By Ash Rein on Monday, April 22, 2024 - 09:43

I really think that the assistant sounds like a snarky jerk. I won’t go anywhere near it if it continues to respond in that way. I believe there needs to be at least some way to set tone. I personally would prefer a professional clear response to my inquiries. I don’t like my assistant saying things like “unless they ordered the entire menu” or dystopian or anything else that comes off as sarcastic and arrogant. Please tone it down a little bit. At the very least, give us a choice so that we can get the type of answers are actually looking for. It wasn’t even funny the first time. Although I know that some people on this website are going to say that they laugh at that. When I’m using an assistant, all I want is the facts.

This could be really promising in terms of some kind of smart classes and being able to just look around and getting the information. For now it’s cool that you can use your phone camera. I just want something that I don’t have to pull out of my pocket every two minutes to get information.

By Brad on Monday, April 22, 2024 - 09:43

If you go into the settings then personality then type something else in, it might help. I don't know though as I've removed the app.

I do agree that it can get a bit annoying, although it is an AI so doesn't truly have a fluid personality yet, perhaps one day they will and when that day comes I'll be very interested but we're a couple years off from that.

By mr grieves on Monday, April 22, 2024 - 09:43

The voice in the demo totally put me off too. But when you start the app for the first time you are asked for a bio about yourself, then you get to describe the personality of the AI, then choose a voice. I can't remember what I put in the personality other than I wanted it friendly, informal but to the point. And so far it's been spot on and nothing at all like the video.

By miguel3025 on Monday, April 22, 2024 - 09:43

Hi.
Regarding personality, I think you need to experiment with a kind of trial and error until you get it right. For example, the first time I had very rich and overly fanciful descriptions for a fantasy book fan like me, so I tried to alter the text to improve and perfect it, and now I have something nice and just right. My favorite voice is, without a doubt, Fin's. It's so natural and human that it gives me a feeling of actually having a conversation with someone, despite Eleven Labs having that issue of bugs in emotions and starting to randomly shout or whisper. But for now, my feedback is more than positive.

By Lee on Monday, April 22, 2024 - 09:43

Guys for those who have and are testing this is it working? Worked fine for 2 days then yesterday it simply stopped speaking. Neither photos nor voice questions are now being answered. Tried different voices and even uninstalled and reinstalled but no luck. Totally silent!

By Dave Nason on Monday, April 22, 2024 - 09:43

Member of the AppleVis Editorial Team

If you’re on the Slack space, you’ll see that a few of us reported this in the Bugs channel. Karthik has acknowledged it and said the team will be back to work to get it fixed on Monday.

By Datawolf on Monday, April 22, 2024 - 09:43

The different personallity traits are something that I had hoped for years by now and I could imagine Siri also getting these, although with me beeing the goofball that I am I would probably turn her into a all sassy black woman beeing snarky about everything possible :D.
Just signed up for the beta, hope that you guys increase bandwidth sooner or later again.

By Lee on Monday, April 22, 2024 - 09:43

Sorry can't get my head around slack so posting here. A new release came out which was meant to fix the lack of speech. Well for me at least it hasn't still silent. I'm in the UK if that means anything.

By Brad on Monday, April 22, 2024 - 09:43

That works for some people.

Also I believe they're using 11labs so the speach might go down there so it goes down on their side too.

By Ollie on Monday, April 22, 2024 - 09:43

I think I signed up for the beta a couple of weeks back but have heard nothing. I'm starting to think my sign up failed. Is there any way of checking?

By Brad on Monday, April 22, 2024 - 09:43

The email should be somewhere in this thread.

By Karok on Monday, April 22, 2024 - 09:43

hi when should more testers receive invites?

By Brad on Monday, April 22, 2024 - 09:43

They're good at responding.

They might have enough testers for now.

By Bl on Monday, July 15, 2024 - 09:43

Okay, I signed up and I got an email but I got the android beta link instead of the iOS one can somebody just give me a test for Betalin because I am not receiving another email after I registered again and I am confused they are not replying to me on the slap channel so can you please give me the link?

By Missy Hoppe on Monday, July 15, 2024 - 09:43

I've been beta testing this app for a few weeks now, and I absolutely love it. In general, Piccy Bot is more reliable and responsive, but I like that there are far more voice options available for Envision Assistant. I'm still not quite sure how to configure its personality. I mean, I know the steps, but am not quite sure what info to write in order to creat it's personality and things like that. It just seems like I get more vivid, detailed descriptions from the Envision Assistant, although I'm very impressed with the info I can get from Piccy Bot as well.