I think the predictive analytics and attempts to represent/visualize 1) our (a) unique (human) “Self” in real-time; and 2) fine vs. current course-grain human-connectity (applying such visualizations) in the form of efficacious and value-rich (read: Self-Evolving) prescriptions of “Hu-synergies and Hu-synchrocities”(applied to socnet data clouds) will be one of the best proving grounds for semtech as it adds direct value to AI-ML and sensor efforts.

Big data was created for a reason. It exists, not via the blind and sheer momentum of technology use behavior; but for a reason we have not yet grasped. Human and human-systems “Autopoiesis” will become a valuable data reality for the consuming marketplace within the next 25 years (imho) and begin to be seen as such in this generation. I write to this a bit here: http://imonad.wordpress.com/2009/07/15/weme-metrics/

Best,
Brian

P.S. I am an ex-neuropsych grad stu out of Karl H. Pribram’s lab at Stanford – Mr. “Holographic Hypothesis of Memory and Perception” (… there at the exciting time of – David Bohm + Karl Pribram… and all that)

My main point was the continuing advances already address and overlap what you discuss here in article i.e. advances in sensor networks, cloud computing, and machine learning.

Finally I’ve seen the 2011 Semtech Conf program and it’s quite good !

@Cole – Speaking of semantics… ‘tar sands’ is the popular term favored by the likes of The Economist and even the Wikipedia entry you cite. But point taken.

@David – How important ML is as a competitive advantage depends on how propriety the data is. For the widely available data in the financial markets, ML is the differentiating factor. For other spaces, like traffic prediction from mobile phones, if you can get access to proprietary cell tower data – for example – then, to use a phrase, more data beats better algorithms.

http://en.wikipedia.org/wiki/Athabasca_oil_sands

I see ML algorithms as less of a source of competitive advantage than the exploration of new and/or alternative data sources.

In my view, what ultimately determines the wildcatters success is not building better drills but staking smarter claims.

This trumps what you say in this article i.e. “These are hints of the promise of big data, which will mature in the coming decade, driven by advances in three principal areas: sensor networks, cloud computing, and machine learning”.

GigaOM’s Structure: Big Data conference on March 23 in New York City is good stuff but if you want know more per above then come to 2011 Semtech Conf http://semtech2011.semanticweb.com ( the program will be posted soon )

Cheers….Steve

PS – I’m also an ex geologist ;)