Guest Post: Why is Decision Intelligence a new field?

One of the first questions that occurred to me during my summer internship this year was “what is Decision Intelligence”?  And is it really so different than AI?  As part of my blog series, I interviewed Quantellia’s Chief Scientist and Co-founder, Lorien Pratt.  Here’s what I learned.

Decision Intelligence (DI) as a discipline is starting to come into focus.  Since she comes from an AI background, Lorien tends to see DI through the lens of AI, in fact she has been known to call it “multi-link AI”.

DI is much more than technology

However, DI is much broader than this limited definition.  You are a DI practitioner, says Lorien, if your work involves understanding or helping how actions lead to outcomes, and/or the thought process that you go through before taking an action, to help it lead to the outcome you want (and to avoid outcomes you don’t want). So this means that the DI umbrella includes economists, social scientists, neuropsychologists, teachers, leaders, and many more.  DI is about:

  1. The integration of these previously separate disciplines.
  2. The focus of these disciplines on how they support decisions, which many have realized is the right focal point for working together between humans, scientific fields, and of course, technology to solve important and hard problems.

Lorien and other DI pioneers including Cassie Kozyrkov at Google, Chuck Davis at Element Data, and Vishal Chatrath at Prowler.io are taking on the very hard problem of crystallizing this discipline.  They are driven by the recognition that we can do a much better job of working together if we focus on understanding and improving decision-making.

DI from an AI perspective

From the point of view of AI experts, DI can be seen as a way of combining multiple AI systems together and analyzing causal structures between multiple factors — both tangible and intangible — in order to identify the best actions to producing a certain outcome.

Lorien explains that, from this point of view, DI binds multiple AI systems together to generate a more holistic approach to decision-making.

“People in traditional AI, they don’t realize that there are these ten thousand use cases that are being ignored, that DI answers the question, if I make this decision which leads to this action, what will the outcome be from it,” Lorien said.

Traditional AI has largely been designed for direct single-link systems. In the realm of science, the norm is publishing a paper, or gaining new insight to accumulate knowledge. Historically, the focus of science has been on discovering new things about the world, which is inherently different from dissecting the causal structures of the world: chains of events that can then be combined to enhance our understanding of the outcome of actions that we might take.

DI deviates from traditional AI because the methodologies and underlying goal of DI stems from a different desire to understand the long-term effects of a decision and places more value on human reasoning. DI looks to social science as it seeks to better understand relationships in an increasingly globalized society. The emphasis is shifted toward using visual maps, talking through a decision, and brainstorming the outcomes and effects of events.

So I have come to understand that DI warrants a new field because it spans considerably beyond technology to incorporate academic and other disciplines as well, and because it bridges the gap between technology and the natural way that humans think about the decisions that they make.

Emily Zhao

Emily Zhao is a summer intern at Quantellia. She is a rising third year undergraduate at the University of California, San Diego, studying Applied Math, Cognitive Science, and Computer Science. She enjoys learning and writing about the powerful potential of AI/DI.

You may also like...