A Primer on Cognitive Computing
These three components of cognitive computing mechanisms can be broken down into endless combinations of technologies and algorithms. Cognitive computing systems have additional characteristics have been left out of this model for simplicity – however, I would suggest that most of these other characteristics fall into one of these broad classes of functionality. These are broad and encompassing –akin to describing buildings as things that contain walls doors and a roof. That definition applies to the Mall of America and to a shack on an island. The range of cognitive computing systems is as diverse as this analogy implies. Yet they form a good starting point for understanding cognitive computing, because essentially every cognitive computing system will need to include these components and each of these components require answering key questions about the organization’s customer strategy, business processes, and knowledge and information systems.
There are two main take-aways from this discussion:
1. Cognitive computing will increasingly be part of our world and will be subsumed into every system and process just as smart phones and the Internet are part of our world today.
2. Organizations need to put certain foundational elements in place in order to remain competitive as transformational technologies upend and disrupt the marketplace.
To prepare for cognitive computing, organizations should do the following:
1. Assess areas of opportunity in client-facing processes (customer support, customer service, marketing automation and ecommerce).
2. Continue to manage and curate knowledge and data (foundational governance and data onboarding will be key capabilities moving forward).
3. Understand and build on your organization’s maturity in data science and analytics (this does not necessarily mean hiring a team of data scientists, but means being intentional about enabling critical functions with analytic capabilities).
4. Investigate and experiment with technologies in key competitive areas that will differentiate your products and services in the evolving marketplace (use envisioning sessions to get a shared understanding of the future state of the industry and organizational capabilities).
5. Invest in educating the organization in foundational technologies and processes (knowledge management is not going away or being superseded in the immediate future – these technologies will build on core knowledge capabilities and processes).
Beware of vendor claims such as “our system develops all the algorithms,” “you don’t need to organize any content or worry about data quality”, “our software emulates the human brain,” “it’s based on our proprietary algorithm – you don’t need to tune it,” or “we develop and test all of the hypotheses – you don’t need any special expertise to use it”, etc. I have heard each of these claims and they are only reasonable in very narrow use cases. There is no magic here – cognitive computing requires that we design systems with the customers’ needs and tasks in mind, and support them with upstream internal processes. Cognitive computing is a tool that will allow for amazing new capabilities. Getting there will still require the blocking and tackling of data, content, and knowledge processes - though with new tools and improved outcomes.
Every major technology enterprise is investing in this area in one way or another, and many are already gaining advantages and improving the ways in which they conduct business. Cognitive computing will change the business landscape. With the speed of adoption and technology evolution, it will likely happen faster than many might expect. Which is all the more reason to get your knowledge house in order.