When it comes to the coming data explosion, the wholly grail resides not in collecting data, not in storing it but in actually using and harnessing it to extract its value. There are two variables that will command data harnessing: algorithms complexity and processing power.
A recent Twitter connection, and not of the least engaging ones, by the name of Teri Conrad aka @TeriConrad, picked my late night muse with a sentence on Twitter such that it necessitates more than 140 characters. The sentence was:
@yacinebaroudi now I’d be curious how we can start measuring things that actually matter like overall sentiment etc…
— Teri Conrad(@TeriConrad) March 17, 2013
The conversation was initiated around a 1962 write up predicting today’s Big Data phenomenon, and that had recently been released by the CIA.
What I meant by my answer to Teri, was that the insights and sentiment analysis she’s seeking will eventually arrive by way of either being able to better correlate data or by way of being able to processing larger amounts of it.
@yacinebaroudi imminent I suppose but tough to gather.. Data is fine but insights are better
— Teri Conrad(@TeriConrad) March 17, 2013
It is a bit like today’s mobile devices. For several data collection applications, such as maintaining a continuous GPS track and simultaneously checking that data against a locations database, are inadequate for a very simple reason: battery life. We could to a certain extent equate battery life to the Teri’s insights. There are two ways to solve this problem and allow for more sophisticated applications to be developed, thus more sophisticated data to be collected and better insights derived: 1/ optimize the software or 2/ optimize the hardware.
It may sound obvious but despite the cloud, we still live in a physical world and must abide by its laws. The evolution of technology is likely to moving both variables ahead at the same time and it is really together that they will move us closer to, not only collecting data on a frequency (volume) basis that we can only fathom today, but also run that data through (process) more and more complex correlative algorithms that today would make the fastest processors come to an instant crawl.
None other than Qualcomm is already taking a jump start in this race with its latest processor christened “Snapdragon“. Not just for the love of speed but because Qualcomm has a platform, SDK for the geeks out there, called Gimball. Check out the video!
Future! Here we come! And Teri, thanks for the stimulating convo.
About Yacine Baroudi
