Entry 7 - The Personal Assistant

         Technologies, before developed, are often thought of as the future of the world.  The dreams of one day having portable phones, cars that tell us where to go, or just having a source of light that runs on electricity, rather than gas or candle wax have all been reached, and it’s likely our technological dreams today will one day be achieved as well.  No matter what technological advancement you may be dreaming of, it is most likely followed by the thought that it will truly be the future whenever something so incredible can exist.

…except speech recognition…

       From the beginning of it’s invention, speech recognition has been treated as the type of project that could never be perfected, and is therefore ridiculously stupid.  This is because such a complex task of recognizing human speech and comprehending a string of words takes years upon years to get to even a functioning level, and along the way, any use of that tech would produce poor and often ridiculous results.  Yet installation of these recognition devices has been occurring since the very beginning.  As a way to be on the “cutting edge” of the tech world, companies put sub-par voice control devices in just about anything they could.
    The frustration and laughter that came out of consumers attempting to use these devices has led to them being seen as nothing more than a gimmick.
      However, speech recognition has come a long way, so long in fact that it is now considered one of the “technological backbones” of the “internet of things”(1) which is the idea that all things will soon be connected to a greater “hub” equaling higher inter-connectivity. Programs like Siri on the iPhone claim to be so accurate that it acts as a "personal assistant" you, and can perform hundreds of functions like telling jokes, acting as a calculator, and even locating friends.  Cars that can talk back to you and tell you the closest places to enjoy meals, as well as set navigation for you have also become almost the point of commonplace.

Relation to Computing:
      MIT has recently developed a new chip that reduces the power consumption of voice control by up to 99%.  The continued advancement of speech recognition is led by another “backbone”, programming.  With thousands and thousands of words, and millions of possible sentence combinations, programs must be written to allow the computer to think for itself, deciding which words are most likely to follow others, using complex algorithms and even from drawing on past user input(2).  The computer science behind these devices must be ever-enhancing, working towards the eventual perfection of speech recognition so as not to be considered a joke any more by the whole of consumers.  Institutions such as MIT and companies like Intel are leading this charge and are inventing new technology to assist in the process, including a device that will only begin the voice-recognition process when activated by a "wake word"(1).

References:
1. Hardesty, Larry. "Voice Control Everywhere." MIT News. MIT News Office, 13 Feb. 2017. Web.

2. G, R. L. "How Speech-recognition Software Got so Good." The Economist. The Economist Newspaper, 22 Apr. 2014. Web. 24 Feb. 2017