Friday, January 16, 2015

Impromptu #6: Robotics

Yesterday, i found myself reading an old collection of stories by Asimov after a gap of nearly two years and what caught me by surprise was how close we are to the ones mentioned in the stories. We haven't yet achieved, now probably ever will achieve the capability to program all the actions needed for a robot to understand and react to a world as chaotic as ours but in the serene environment of laboratories, we have developed robots capable of most basic human functions from running faster to understanding pictures better.



1416_20_music-playing-toyota-robot.jpg (500×351)Several laboratories across the world compete on a regular basis to develop the best algorithm to understand pictures and group massive numbers of them based on inputs from humans. What is most intriguing though, is that these codes aren't written in the Start -A-B-C-Finish fashion that we assume code to be. These codes learn from their interaction with the user and improve on their next iteration. This means that the longer such programs are used, the better they get and fewer user inputs are necessary.



Such codes arent limited to image search algorithms. Similar codes are applied to social networking sites like facebook to show you links to sites you would be most interested in and this is the scaries t bit. Eli Pariser does a brilliant job explaining why.

http://www.ted.com/talks/eli_pariser_beware_online_filter_bubbles?language=en

The crux being that you are no longer exposed to a balanced view on the subject as you would expect from a 'free' internet, instead algorithms such as these will present skewed results that match your opinion creating a blanket of un-debated opinions on issues which really matter.



We have reached a point where we are incredibly dependent on Robots, or rather AI and for those of you who disagree, think of the last time you used a dumb device for some productive work.

The world Asimov dreamt of is here and  scarier than the worst he thought of. Robots no longer need to harm humans or human society to rule the world. All that is needed for it to be the big brother. If any of you have heard the news, countries across the world are reacting to terrorist strikes by increasing surveillance and how do you monitor a few million cameras? AI. The amount of digital data being generated is several TB/s or more and the information in there is capable of generating unique, near perfect images of people across the world, even people who don't use the internet much. What is worse is that organizations across the world understand this. Sony hack anyone?

All we need to spice things up and bring on judgement day is for a government to introduce a new AI that can make sense of all the big data out there to protect all of us from terror and watch over us. 
df5d6714fe78ce09cabef24acab89064_large.jpeg (1280×1024)

As much as i want to end this here, i would like to add this. The only way to prevent total surveillance in the name of security is if we can stand up for ourselves. To demand the freedom that we have the right to. To be able to express all that we want when we want to without the threat of getting killed or getting jailed. AI is incredibly powerful tool, both for the common man and for those who want to control the common man. Don't let its use for the second reason overshadow the first.


Iron Sky by Paolo Nutini on Grooveshark

Language and the internet

Language evolves. The pace of the evolution is dependent on the frequency of its use. For much of its history, the written word has followed...