Capturing Business Value with Cognitive Computing

For many businesses, data and artificial intelligence pose opportunities but also risk. 

Hadley Reynolds, co-founder of the Cognitive Computing Consortium, presented a keynote looking at the meaning of AI and cognitive computing and showcased an open reference framework jointly developed by the consortium with Babson College’s technology management program to help organizations gain value from the rapidly evolving technologies.

Since we are still in the early days of AI and cognitive computing, there is confusion about terms and how the two come together, said Reynolds, who said that the fundamental difference in the view of the Cognitive Computing Consortium is the extent to which the machine can emulate thought processes, behaviors, and interactions. 

In the consortium’s view, in AI, machines are fully autonomous with the machine literally doing the work of the human brain, and the author of its own actions. 

In cognitive computing, on the other hand, the machine behavior is dependent on humans with the machine as an information tool and an agent of some intention or process of a human being.  

In the view of the consortium almost no system that is being talked about is a fully autonomous system.

However, Reynolds acknowledged, this distinction is not embraced by the industry with these and other terms often being used interchangeably.

Reynolds refers to this era of AI and cognitive computing as “the early chaotic era,” with chaos exacerbated by multiple vendor interpretations, a dearth of credible guidelines and information, a lack of skill sets and cross-silo networks, and uncertainly about ROI. In addition, many applications are highly customized; technology is unsettled with many terms proliferating, such as cognitive, AI, machine intelligence, deep learning, and smart machines; and there are no industry standards across many are areas of the field.

How do you get value out of chaos?

According to a 2018 McKinsey Global Institute report, the current projected global impact of adopting what it calls AI is $3.5 to $5.8 trillion annually globally. McKinsey has identified 400 use cases for AI across 19 industries.

The top 10 industries on the list are:

1. Transport and logistics
2. Retail,
3. Automotive and assembly
4. High tech
5. Oil and gas
6. Chemicals
7. Media and entertainment
8. Basic materials
9. Agriculture
10. Consumer packaged goods

What is Cognitive Computing?

According to the Cognitive Computing Consortium, cognitive computing makes a new class of problems computable.

It addresses complex situations that are characterized by ambiguity and uncertainty so it can handle human types of problems. In addition, cognitive computing systems make context computable and redefines the nature of the relationship between people and their digital environment.

Five pillars of cognitive computing include that it is:

• Learning Adaptive: Reacts and changes based on new information and interactions
• Probabilistic: delivers confidence scored results
• Contextual: filters results depending on who, what, where, when, why, and how
• Conversational: meaning-based, interactive, iterative, stateful
• Highly integrated

The Cognitive Applications Framework was developed by Cognitive Computing Consortium/Babson College to give executives and operating managers a tool to characterize the impact and behaviors of potential AI applications.

Beyond impacts and behaviors, the framework seeks to integrate the profiles of skills and resources required to effectively execute cognitive tasks.

Takeaways of the framework are that:

• Cognitive computing is a set of technologies that support a spectrum of uses. In combining these, designers must make tradeoffs depending on their type of problem, use and users.
• The Cognitive Applications Framework offers a method for decision makers to define their problems and select appropriate cognitive computing investments,
• The framework provides a logical step-through process for understanding and planning cognitive applications projects.
• Understanding context requirements, data, technology, and behavior enables sound judgment about knowledge, skills, and abilities required for successful project teams.