I have mentioned a wild long-term project I’d like to start working on. I call it, Project Galahad. It will eventually be open-source.
Galahad, pending a name change upon completion, will be my version of Jarvis. Voice control has become very popular on phones. However, the usefulness is up for debate. The other problem is that nobody likes talking to their phone in public. In this case, we can scrap the worry of public embarrassment because I’ll be in the safety of my own lair. Usability is the most important challenge to overcome.
I don’t like to reinvent the wheel so my plan is to simply leverage Cortana and/or Google Now to do the real heavy lifting for me. There is already Cortana API for the Windows phone so I anticipate the same for the Desktop version in Windows 10. Alas, I have a Google powered phone and habit. So, I’m not a huge fan of Cortana being a Bing powered helper. I’d like to be able to leverage both engines and overlay my own choice of voice(s) over the top.
There’s an awesome project underway called, Viv. It is by far the most impressive AI I’ve ever seen. However, it isn’t out yet. I also have no idea if it will ever actually be released to the public or reasonably priced. Thus, I’m not waiting for Viv. I also have a drive to learn about this stuff and make my own.
Welcome to the start of my voyage. Below is a technical beginning to my work.
Where do I start? Here!
I’m going to start familiarizing myself with general speech recognition and writing custom commands. Essentially, I’m going to buy some time until Windows 10 officially comes out and I can leverage Cortana. Until then, I’ll also start working on utilizing Google Now or whatever I can.
I anticipate that it’ll be difficult to make the AI understand varying ways to say the same thing. That’s why I’d like to leverage already created AI’s to do custom things. Hopefully that won’t be a bigger problem than I anticipate because I don’t know that I have the gusto to engineer understanding.