Monday, June 3, 2013

IBM and the Rise of Big Data


IBM CEO Ginni Rometty gave a presentation last week at a conference last week in New York.  Several of my colleagues heard her talk, and recommended that I listen to a replay.

Every two days, according to Ms. Rometty, the world creates as much data as the world created up until 2003. 

This astonishing amount of "big data" is beyond the capability of any human to process. IBM and other tech leaders are creating ways that companies and other institutions can manage this data in the most effective way possible.

The future of technology is one that is exciting and yet vaguely unsettling to many.

Since they were first introduced in the 1930's, computers were largely used to tabulate or calculate large sums of figures at a rapid pace.  Today, however, computers are able to manipulate data, and come up with solutions or predictions without any human intervention.

In the investment world, on a typical day roughly 70% of the trading volume on the stock exchanges are done by computers executing trades in nanoseconds to capture the smallest amounts of pricing inefficiencies. This creates a serious challenge to active management of equity portfolios by humans, since in a sense managers are competing against supercomputers rather than other humans.


The blog Seeking Alpha carried a complete transcript of her 45 minute presentation. Here's an excerpt (I added the emphasis):

In all of time, all of our lives, computing has gone through three different eras. The first machine were things that tabulated and counted, if I put it simply, tabulating. The next era, everything you have known to date is systems that are programmable, the third era, which we are entering and these will go for decades, is going to be an era of cognitive systems, those are systems that learn, and if you just think for a minute, why this is in type, you have no choice in a world of huge big data, no one will be able to program, how to deal with all that data, the system have to learn itself what the data is saying and our first instantiation of that was something called Watson...

And the reason I say that it will distinctly change three things about a company, and I can talk to you about how I change IBM. It will, one, change how you make decisions, they will be predictive analytics. It will change how you create value that will be the idea of a social business being in the production line. It is going to be more important what you share versus what you know. And then third, how you deliver value will change, it will go to the individual, as I said, it will be depth of the average, that will completely change in all that underpin with data. So kind of a very fast view of the industry which takes me to my second big strategy for us is around creating therefore new clients and new markets.

http://seekingalpha.com/article/1468881-international-business-machines-ceo-presents-at-sanford-c-bernstein-s-29th-annual-strategic-decisions-conference-transcript?part=single