Decision Options ® - Artificial Intelligence for Decisions

As the “net neutrality” debate kicks in – effectively forced by a handful of monopolistic companies who almost already control the internet in an attempt to make it even less accessible, it is encouraging to see that the city of San Francisco (SF) is heading in the other (and correct) direction. The recently appointed Chief Data Officer (CDO) of SF is initiating a strategic plan for making data open. Data, the only remaining valuable raw materials in the world, are largely locked up by a handful of companies, generating monopoly profits for them.

It is high time that data are democratized. As the economy shifts into making every decision data driven, delegating the meaning of the phrase “gut feel,” to the pain one gets after a heavy and unpleasant meal, existing and antiquated monopoly laws need to be rewritten. The industrial revolution is over and today’s economy is not driven by the manufacturing of nuts, bolts and automobiles – it is driven by information and intellectual property (IP). Real monopolies of today are those who are hoarding data and those who have unfair advantages in the use of the central nervous system of the economy, the internet, to create and lock up IP. The fact that the regulators are even considering the argument that the internet needs to be subdivided, providing more skewed advantages to those who already operate as monopolies indicate that they are completely out of touch. A congress, filled with octogenarians and a bureaucracy, only slightly younger, are in no position to make laws in the information age.

Net neutrality and democratized data are necessary conditions for unbridled innovation and economic growth. Anything less will be a step back – something regulators are well advised to stay as far away as possible.

Big data and analytics are increasingly fashionable in every field - business, science, politics, policy and others. Four vectors - cloud/computing, data, analytics and context - need to come together to enhance value in companies embarking on using these emerging ideas. However, most providers specialize in only one or few of these necessary categories making implementation of complete solutions complex, costly and often not effective. Just as in many of such previous waves - Enterprise Resource Planning (ERP) and Business Intelligence (BI) to mention just two, the hype surrounding big data analytics is probably more than what it will ever deliver. More importantly, the technology companies involved in this wave, just as previous ones, seem to believe that it is an IT opportunity and the users of the plethora of the available analytical tools both inside and outside companies, don't seem to be equipped sufficiently to add value. 

Management consultants, adept at delivering PowerPoint documents filled with one-off analysis, are becoming obsolete. What companies need are not static documents but dynamic models that provide decision guidance based on changing and growing data. Such models have to take into account all types of data - structured and unstructured, deterministic and probabilistic, quantitative and qualitative - in a systematic way to improve decision quality. This is not a mathematical problem - there has not been any new Math for several decades - but a decision problem, something that requires a clear understanding of the context. Information Technology (IT) - a pure commodity - should be fully scalable in a cloud infrastructure - either public (less expensive) or private (more secure). Analytics need to be complete and systematized and should not be delegated to a few people with expensive statistical tools - leading to biases in methodologies utilized and models built. 

A focus on decisions - not data and analytics - is likely a better way to approach this problem. Frameworks that encapsulate all available techniques, operating from a flexible and infinitely scalable cloud infrastructure, leading to models that are able to learn from growing and dynamic data and provide decision guidance at the right time and place - where, what, who, how, how-much - will dominate this market. Companies should be careful not to over invest in hype, doled out by consultants specializing in one or few of the necessary areas - IT, Analytics, Data and Management. The solution has to encompass all of these - and it should begin at the end objectives and goals.

Data and analytics are just raw materials that aid solutions for better decisions. It is obvious that the design of a house will not start from accumulating concrete, wood and steel. This is equally true in big data and analytics.