I thought that I would take a minute this morning to write a little bit more about the motivation behind building Talk to Me, Goose! now that the app is live in the App Store. Before I do, go download it and try it out. Give me feedback on what you like and don’t like in the comments below, and give it a rating on the App Store. I need both to continue to make it better.
What do people do when they lose their ability to speak?
When I was first diagnosed with a motor neuron issue and was having the first signs of challenges with my speech, I became deeply concerned about what might happen if and when I lose my ability to speak. As someone who finds that I use my voice not just to communicate but also to work out hard problems by “thinking out loud” with my colleagues, I became pretty fixated in my head on what people do who lose the ability to speak.
We probably all remember Stephen Hawking and the device that he used to communicate using his eyes, so I reflected on that. I assumed that given the time that has passed and the level of innovation, especially in the area of artificial intelligence, that the quality of voice synthesis in use today would be vastly superior to what Stephen Hawking had used. And, it was. But, I was a little bit surprised at what I found, actually.
Is this the best there is?
Now, I will admit to the fact that I probably did NOT do a comprehensive and systematic review of all of the tools and technologies in use across the entire landscape of the speech pathology domain. I was, admittedly, a bit of a man on a mission. Here I was beginning to experience what I perceived to be the beginning of the loss of aspects of my speech, so I felt I had no time to waste. So, when I found what seemed to be the most frequently cited tools and technologies, I went right to them to check them out.
I created a voice clone using the technology that was described as the leading technology provider to the leading solutions provider for people like me. This is what I found.
And, while this does indeed sound like me speaking the words, the overall quality of the speech left me wanting for something more realistic and natural.
In my mind, with the advances available in artificial intelligence, something more realistic and emotionally rich had to be possible. Right? I mean, we’ve all seen the deep fake videos of Tom Cruise from a few years ago. If we could do that, we should be able to do something better than this was what I was thinking at the time. And, I did indeed come across something far better and it got me thinking, a lot.
Inspired by Conan O’Brien. No, really…
This is when I came across a brilliant episode of Conan O’Brien’s YouTube podcast of Conan O’Brien Needs a Fan.
I encourage you to watch the video because, well, first of all, Jules and Maria are clearly just remarkable people and, secondly, because the video just lit me up about what was possible. You’ll see in the video that Jules uses his device in at least three different ways. He clearly is using both of the voice cloning technologies that I have identified above, and he uses message banking, a system of storing previously recorded messages in an easily accessible playback mechanism. He has clearly banked some very funny things as well, which I love.
Well, really the inspiration came from Jules and Maria
Beyond the Conan O’Brien episode, i encourage you to go listen to Jules and Maria’s podcast, The Couple Shift where you’ll learn a lot more about their lives together and how they’ve grown as a couple living with ALS. They truly are beautiful souls from whom we all can learn so much about living life, in general. It was listening to Jules and Maria that inspired a lot of what informed the concepts that went into Talk to Me, Goose!
As I watched and listened and did more research, it became clear to me that what was missing, at least as far as I could tell, was a simple app that brought together a high-quality, emotionally rich AI-voice synthesis capability with a mobile text-to-speech device that could also bank frequently used messages and would enable rapid rendering of speech in a natural and realistic way. Thus was born Talk to Me, Goose!
But, what about the name?
Now, where did the name come from? Well, when I was in high school, I certainly was a fan of the original Top Gun movie. I mean, who my age wasn’t, right? And, I recall that whenever Maverick found himself in a moment of uncertainty, either before (and most certainly after) the accident he relied on his closest friend, his Wingman, for support and guidance. He relied on Goose. When he needed a little help, he uttered those words, “Talk to me, Goose!” And, right now, I know that I’m going to need a little bit of help, and I’m going to need to rely on a Wingman to help, so “Talk to me, Goose!”
So, go get it, help me out
The last few months have been quite the journey as I went from not knowing anything about how to build an iOS application to building and launching what I think is a pretty good version 1.0 of an application that is now live in the App Store. It’s not perfect. Just in the writing of this today, in fact, I found a bug that I need to go fix. But, it’s out there in the world, and I decided not to let perfect be the enemy of good. Lord knows it’s not perfect, and I have a backlog of ideas as to how to make it better. I want to get it in the hands of people who can help me make it better by giving me ideas as to what it needs.
I learned along the way of building it that it has multiple use cases, some that are completely unrelated to the one that inspired its creation which could help fund this, its most important. I’m working on getting it out into the world to generate feedback, and if I can get it into the hands of people who can benefit from it, I will feel that I have contributed something. In the meantime, I know that I have a Wingman to help me when I need him. I hope that I have given some others the chance to have a Wingman too, a chance to use their own voice, or at least a clone of it, in a natural way when their real one might be failing them.