Causal Artificial Intelligence

CHAPTER 3Elements of Causal AI

In the first two chapters of this book, we provided an overview of the journey to causal AI and an overview of terminology, including graphing, models, and the underlying operations of casual AI. In this chapter, we will delve more deeply into understanding the elements of causal AI with a focus on the correlation and causality. We begin this an overview of the conceptual model designed for causal AI and then move into a discussion of directed acyclic graphs (DAGs) and structural causal models (SCMs) and how the combination of these two elements work together to create models. One of the key benefits of a model-based causal AI approach is that it provides a consistent and understandable model that promotes collaboration with all key constituents. As business professionals can iterate quickly across the various scenarios they are considering, they will learn and know what combinations of programs, offers, and sequences of offers will produce the optimal results. Causal-based AI is one of the most likely paths to user-centered empowerment and daily engagement with powerful and useful AI-enabled models.

Conceptual Models

At a conceptual level, correlation-based AI and causal-based AI approaches have the same theoretical roots; both methodologies can produce actionable models based on leveraging statistical and data analysis. These approaches to modeling have input data that is conditioned and transformed for consumption by the models. Both correlation ...