Skip to main content
Welcome guest. | Register | Login | Post

And I thought "point and click" was user friendly

Get this.

My mom is quite computer illiterate, but the other day she wanted me to show her how to turn the computer on and play some music. Imagine what the most difficult part of the whole process was! It was pointing and clicking with a mouse. She says her arm started hurting of all the trying. Double clicking is particularly hard because as she does it the mouse slightly moves and the double clicking fails.

It turned out that it was easier for her to use the keyboard to start or stop playing music and move between songs than using a mouse. And I thought pointing and clicking was supposed to be new user friendly!

I told her in the future we'll probably, like in Star Trek, control our computers with voice and touch screens and she really welcomed that. It seems that it is much more natural for an average person to touch the stuff on the screen with their fingers (or with a pen) rather than use a mouse to do it, but this seems to escape us computer geeks quite commonly, as we mastered our mouse navigation skills. Smiling

Star Trek was utterly spot on with their vision of computer interfaces, and chances are great that this is exactly how it will be. We already have multi touch screens - they just need to start entering the home users market. Some day soon, noone will need a mouse and everyone will, if they wish to, use voice to control their computers. Keyboard will probably be around for touch typists for fast and quiet (voiceless) input though, but as they already are more and more, they will be slim, compact, wireless and portable. Smiling

One thing left to adapt to this kind of revolution is the operating system software. It must support touch screens and have good voice recognition software. The former is probably not an issue for Free Software, but the latter could be. Windows Vista might, at this point, be slightly ahead of us. There's a hint for whom it may concern. Eye


I've looked at some videos

I've looked at some videos of vista voice recognition just now, it appears to work well enough for replacing the mouse, however writing letters appears to be a problem. Also, I wonder how one would write a document that contains phrases like "close app" or "select all, delete" Laughing out loud .

I imagine a voice-controlled commandline would actually be easier to use than a voice-controlled WIMP interface.

tbuitenh wrote:Also, I

tbuitenh wrote:

Also, I wonder how one would write a document that contains phrases like "close app" or "select all, delete" Laughing out loud Laughing out loud.

Indeed. It just doesn't seem to be good enough for reliably and naturally dictating documents. It's still better, less frustrating, for a user to practice with the keyboard. Smiling

But doing some routine things within the system with voice commands can be quite useful.

But I think voice recognition may be popularized only after everyone already has touch enabled screens. It just seems more of a natural path of evolution considering how "alpha" speech recognition still is.

I'm wondering though, what would it take for it to be completely reliable and failure proof? When I watch people in Star Trek interact with the computer they don't even take much effort in talking to it in any special way, except for saying "computer" when they want to address it. They talk to it as if they are talking to a human.

So my assumption usually is that the computer, that is the speech recognition program it is running, is simply advanced enough to be able to recognize not only words, but tones in which they are spoken, the imperfections in human language, the slang and even the mumbling (to a point) - and perhaps most importantly the context of words (which is why for example, Picard can talk to someone else about a hologram program, giving some comments on it to another person and then address the computer by just saying "change it to that, computer" and it actually knows what to do).

I suppose they need quite a bit of computing power for this? Either that or the algorithm is just that advanced in the way it was designed, and operates on a vast database of speech related variables of all kinds.

I very well think it is possible to have such reliable speech recognition (as in ST), but while the idea is quite old, it's time just isn't coming yet.

Comment viewing options