Monday November 14, 2005 The desktop war is over. Who won? Mobile devices! - perhaps a bit exaggerated statement and perhaps a bit ahead of time, but I really believe the mobile devices (mostly in a form of mobile phones) will be the most widely used computers by the general public in the future. If you look at the stuff the non-technical people do (like my wife for example
), they usually send/receive emails, look for some information on the web, book their vacations, perhaps do some shopping on the web and use some type of instant messenger. These applications cover well over 95% of their needs and to be using those you don't need a desktop computer even today. Also the fact the device is connected to the network in most of the places is a great thing - a week ago I missed a bus when getting home in the late night and withing a couple of seconds I got a list of alternative lines I could use - exactly right at the place and the time I needed this information (this wouldn't be possible with my desktop).
A very important thing is also complexity (or rather simplicity
) - mobile phones are usually quite easy to use (ok, not every manufacturer is able to produce user friendly interface, but this is changing rapidly), and usually don't need any kind of setup - most of the things are usually preset from the cellular network operator or can be set even remotely using a special kind of SMS. This is one of the thing I really enjoy about this technology - being a computer guy usually means solving many problems with desktop computers across the whole family, but I never had to help anyone with a mobile phone
.
The only thing the mobile devices are currently missing compared to desktops is a good input device and to lesser degree a display with high resolution. Both can be fixed though - a long time ago I was reading about a projections keyboard, which works by projecting a red outline of a keyboard on a desk by a laser beam and reading back the keys you would like to press by noticing reflection changes on the projected keys. I definitely wouldn't mind to have such thing built-in in my mobile phone.
Display resolution is also increasing rapidly - PocketPCs have resolution of 480x640 for some time and the new Nokia N80 has resolution of 352x416 - not a bad resolution for a mobile phone. Anyway in the future I would definitely like to see a 3D holographic projection from a mobile device.
There is also a lot of future in voice recognition when dealing with mobile devices, but I still keep wondering why does it take so much time to make voice recognition a reliable technology - as far as I remember predictions 20 years ago, we should have been using voice recognition for some time now. On the other hand, speech synthesis is working quite well - I looked at FreeTTS project some time ago and I was amazed by the speech quality.
So what do you think? Do you see any future in the desktops? I don't, so that's the reason the desktop war is simply over.
P.S. I'm writing this on my laptop with an external keyboard attached :-o - I still don't have a decent device with a projection keyboard and a 3D holographic projector.
Agree.
However, part of the problem is the formidable difficulty in eliminating ambiguity at several levels: the phoneme isolation level, wordsplitting, validating morpological constructs, grammar validity, semantic-level checking. Not to mention things like context. And it is estimated that even things like rhethorics or prosody can be used in this disambiguation process. Overall, a very complicated (and challenging) problem.
AFAIK, right now, speech recognition techniques have a low rate of success - aroud 80-90%. And this is not enough, given that users are easily frustrated with such high failure rates.
Speaking takes time (at least, much more time thanmoving a mouse or typing), and nobody likes to speak to somebody that fails to understand 10%-20% of what is he saying...
Posted by Adi Oltean on November 15, 2005 at 07:46 AM CET #
Posted by Martin Brehovsky on November 15, 2005 at 10:26 AM CET #
Posted by Lukas on January 05, 2006 at 12:03 AM CET #