Is it just me or is the smart-phone market actually going down the similar path as the PC market did a decade ago – chasing more cores and more Mega-Hertz speeds. Even while the smart-phone market is dominated by ARM-only platforms, it is still folly to be making hand-sets that are faster and furiouser than the rest.
I would think that these companies would have realised that it is far more important to improve the experience and functionality of the smart-phone than to increase the speed of the phone. Increasing the speed of a phone is easy – just wait a few months and Moore’s Law will ensure that it happens eventually. There is no magic in that.
I would think that the smart-phone is a booming market, that has still got so many pain-points. Although I must confess that I do not own a smart-phone, I am fairly familiar with their architectures and have a good idea of how to actually build one. The reason that I do not own one is because I do not see any useful phones out in the market.
If you were to just sit down and observe a smart-phone user, you will find that there are lots of problems with using the phone. For one, it’s terribly small and it requires a person to stare at the screen to use it. While I think that the touch interface is a pretty cool invention, I do not understand why it is that for a device that comes with a microphone and a speaker, I need to use my fingers and eyes to interact with it. To me, this is just silly.
If I was one of those smart-phone hand-set makers, I would put some good money into developing a new user experience that does not take my eyes off the road and fingers off the wheel. The first should be easy enough to accomplish. With gobs of memory and processing power, the phone should be able to speak to me instead of presenting a visual response. You don’t need to build some fancy text-to-speech engine to do this – just store MP3 samples of audio output.
As for getting user input, again, the system does not need to have some super speech-recognition engine built in. Just being able to discern a “yes” from a “no” can already build us a tremendously complex menu system. Then, put the money into the smarts of predicting what it is that the user wants to do based on context. If the user is driving, the user may be interested in accessing the GPS and *not* the camera application.
With that, we can probably come up with a better smart-phone – one that actually has some ‘smarts’ and is not dependent on my smarts.