AI & Data Strategy
We offer companies new market opportunities through the use of cutting-edge technologies and AI-based data analysis processes.
In this article, I address my concerns regarding the implications of uncertainty in decision support systems, with a particular focus on its effects on AI-based decision support systems (AI-DSS). The genesis of this article stems from the reflections shared in a previous piece, where I explore the concept of uncertainty beyond the realm of data quality (here is the link).
Decision Support Systems Overview - A decision support system (DSS) is a software system designed to collect, organize, and analyze company data to provide valuable information for decision-making or to support management and planning activities. Essentially, a DSS aids individuals tasked with making strategic decisions in complex contexts where determining the optimal choice or decision-making strategy is challenging. Their reliability hinges on the algorithms used to process data and the quality of the dataset utilized, constituting a simplified model of reality derived from the data available to the decision-maker.
AI-based Decision Support Systems - In recent years, DSS software systems have evolved with the integration of artificial intelligence (AI) to enhance the reliability of the representation model of reality upon which calculations are based. AI autonomously constructs the representation model of the analyzed universe (AI model) solely from the dataset provided by analysts.
The Importance of Dataset Selection - An AI develops its representation model based solely on the dataset designed by analysts. However, since the real world surpasses our ability to synthesize, analysts strive to strike a balance between capturing enough dimensions to represent macro dynamics while avoiding complexity that hampers result verification.
Emerging Doubts - Despite meticulous dataset design, it remains one of many possible representations of the real-world environment. As one form of uncertainty is linked to environmental complexity, doubts arise regarding whether the dataset itself is immune from environmental uncertainty. This concern transcends bias and addresses a potentially impactful yet less tangible aspect.
Consideration of Dimensions - Doubts arise concerning the selection of dimensions within the dataset and the uncertainty surrounding their influence on the AI model and processing outcomes. Unexpected interactions or interdependencies among dimensions could affect processing results, even those deemed marginal or independent.
Artificial Intelligence Development Process - AI systems operate through neural networks trained for specific tasks, utilizing deep learning. These networks employ layered structures where each layer contributes to final processing, with the ability to learn and solve complex problems autonomously. However, the nonlinear data processing within neural networks renders their processing opaque, resembling black boxes.
Certainty of Results - The primary limitation of AI today lies not in computing power or data availability but in the structure of AI models. Without a comprehensive understanding of the context, caution is warranted when entrusting AI to identify solutions, as it may generate optimal solutions in contexts diverging from reality.
Article source: Linkedin Article by Vincenzo Gioia
“These are times of great uncertainty”
The times I have said and heard this phrase cannot be counted. I usually hear it when a decision needs to be made, and we find ourselves in a complex context. But what is uncertainty? How does it affect our decisions? How significant is the impact of uncertainty on AI-based decision-making systems?
In this first article, I share some reflections on the concept of uncertainty, hoping to stimulate a useful discussion that helps me delve deeper into aspects I might not currently be considering.
I've noticed that we don't always share a common understanding of what uncertainty is. According to the Treccani dictionary, uncertainty is the condition one finds oneself in when the information or data relating to a fact are uncertain or contradictory, offering insufficient or not entirely well-founded knowledge. Usually, this condition of uncertainty is determined by complex contexts in which it is not easy to determine the level of risk associated with the decision we have to make.
The conjunction of risk and complexity often leads us to believe that uncertainty is determined by the complexity of a context and/or its riskiness. However, while risk can be assessed through statistical methods, uncertainty involves the inherent inability to recognize influential decision variables and their functional relationships. Regarding complexity, I don't think a complex system is necessarily uncertain since it is characterized by numerous parts that interact in a complex but not necessarily uncertain way.
Having established that uncertainty is not linked solely to the complexity of a context or the risk associated with decisions, I consider it useful to expand its nature beyond the initial definition, which attributes it solely to data incompleteness. In my opinion, uncertainty can also stem from an inadequate understanding of the context and the undifferentiated alternatives among which we make our choices. To put it more elegantly and in line with academic discourse, the origin of uncertainty can be divided into three dimensions: informational, environmental, and intentional.
Informational uncertainty derives from incomplete information and is the most frequent source of uncertainty. It manifests when there is a lack of data acquisition or when one possesses an incomplete, low-quality, or insufficiently diverse data set.
Environmental uncertainty is linked to the natural complexity of the context (the real world), where it is crucial to distinguish between cause-effect relationships and other phenomena that are not causally linked. If we cannot understand the context and its determining relationships, we risk being confused by the contradictory meanings conveyed by some phenomena.
Intentional uncertainty arises from the fact that, in many cases, the decision-maker relies on variable criteria even when the generated choice alternatives are concrete and objective. In other words, humans make final decisions based on subjective perspectives or preferences, even in the presence of a standard decision-making process. This context with a strong subjective imprint is further accentuated by a decision-making model in which the available options are characterized by what we can define as an "undifferentiated alternative," where the choices do not present a clear distinctive factor.
A decision-making process affected by uncertainty can be characterized by three different sources of uncertainty: informational, environmental, and intentional. Environmental uncertainty arises from the unpredictability of the environment, while intentional uncertainty originates from individuals' specific preferences and needs. Informational uncertainty, although considered manageable to the extent that one can understand and describe the causal relationships between variables and acquire the most complete information possible, remains difficult to manage due to the unpredictability of some real-world events, whose nature is often determined by components linked to intentional uncertainty.
Article source: Linkedin Article by Vincenzo Gioia
