ARTIFICIAL INTELLIGENCE FOR DECISION-MAKING PROCESSES

The course aims to provide students with the conceptual coordinates and practical tools to navigate the complexity of decision-making processes in the era of Artificial Intelligence. The primary objective is not to train IT technicians, but rather aware managers and decision-makers, capable of designing effective, ethical, and regenerative human-machine systems. Through an interdisciplinary path that combines philosophy, cognitive economics, and technological analysis, the course aims to develop the ability to distinguish tasks that can be delegated to automation (the "mechanical") from those that require human judgment (the "free"). Students will learn to break down the decision-making process into its constituent phases, identifying cognitive biases and understanding how predictive and generative AI can intervene to reduce uncertainty or expand options. Finally, the course aims to cultivate uniquely human higher faculties (primary creativity, ethical judgment, embodied intuition, and metacognition), transforming AI from a potential substitutive threat into a tool of emancipation and augmentation for the manager of the future, understood as an "architect of decision-making systems".

There are no mandatory formal prerequisites. However, for fruitful participation in the course and a better understanding of the contents, a basic knowledge of the principles of Business Management and the fundamentals of Organizational Behavior is considered indispensable. Preliminary familiarity with the basic concepts of descriptive statistics and probability is also strongly recommended, as it is useful for understanding the operating mechanisms of predictive and generative Artificial Intelligence. No prior computer programming (coding) or data science skills are required, as the technological approach will be addressed from a managerial and strategic perspective.

Upon completion of the course, in accordance with the Dublin Descriptors, the student will be able to: 1. Knowledge and understanding: recognize and describe the theoretical foundations of cognitive economics, the limits of human rationality (cognitive biases), and the statistical operating principles of Artificial Intelligence (predictive and generative), clearly distinguishing between "mechanical" tasks and "free" faculties. 2. Applying knowledge and understanding: analyze and break down a real corporate decision-making process into its constituent phases, applying the "Mechanical vs. Free" framework and the "Centaur Matrix" to identify which phases to delegate to automation and which to reserve for human judgment. 3. Making judgements: critically evaluate the ethical, social, and organizational implications of AI adoption, consciously choosing between scenarios of competition (substitution) and scenarios of complementarity (augmentation), and recognizing algorithmic biases. 4. Communication skills: argue and justify their choices in redesigning decision-making processes, effectively communicating the strategic vision of human-machine integration to both technical and managerial stakeholders. 5. Learning skills: develop a metacognitive approach and "Socratic ignorance" that allows for continuous updating of skills in a rapidly evolving technological landscape, maintaining the centrality of the human factor.

The course is structured into three main thematic blocks, designed to guide the student from theory to managerial practice. Block 1 - The Foundations: From Thought to Decision (Lectures 1-5). Explores the philosophical and economic bases of human decision-making. Introduces the Paradox of Freedom (Spinoza), the history of cognitive externalization, the limits of human rationality (cognitive biases according to Kahneman and Tversky), and the architecture of decision-making (Simon, Hidalgo). Defines the fundamental framework distinguishing the "Mechanical" (computable and replicable) from the "Free" (uniquely human), using the Cynefin Framework (Snowden). Block 2 - The New Toolbox: Mastering the "Mechanical" (Lectures 6-11). Analyzes AI tools as levers to automate mechanical tasks. Demystifies Machine Learning and Neural Networks (Mitchell), comparing Predictive AI (uncertainty reduction) with Generative AI (exploration of possibilities through LLMs). Introduces the figure of the manager as an "architect of decision-making systems" and includes a practical laboratory on Prompt Engineering. Block 3 - The Human Factor: Cultivating the "Free" (Lectures 12-18). Focuses on faculties that AI cannot replicate: primary creativity, ethical judgment (qualia), embodied intuition, and metacognition. Analyzes the two future scenarios (Competition vs. Complementarity) and guides students in redesigning real decision-making processes, culminating in the "Centaur Manifesto" for a regenerative economy.

Mandatory texts for all students: 1. Lecture notes, slides, and reasoned guides provided by the professor during the course (available on the University e-learning platform). 2. Mitchell, M. (2019). Artificial Intelligence: A Guide for Thinking Humans. Farrar, Straus and Giroux. 3. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux. Recommended texts for further reading (optional choice): 4. Hidalgo, C. (2015). Why Information Grows: The Evolution of Order, from Atoms to Economies. Basic Books. 5. Acemoglu, D. & Johnson, S. (2023). Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity. PublicAffairs. 6. Kasparov, G. (2017). Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins. PublicAffairs.

Lectures and case studies. The course adopts a blended teaching approach that alternates theoretical lectures with moments of active and laboratory teaching. Lectures are supported by visual presentations (slides) and reasoned guides that deepen the philosophical, economic, and technological concepts covered. Real case studies are used to illustrate the practical application of decision-making frameworks. Two practical laboratories are planned: the first dedicated to Prompt Engineering and direct interaction with Generative AI models (ChatGPT, Gemini, Claude), the second focused on the collaborative mapping and redesign of corporate decision-making processes. Classroom discussion, critical debate, and group activities are constantly encouraged to stimulate students' independent judgment and communication skills.

Oral examination. The assessment of learning is divided into two distinct moments. Final phase: traditional oral exam, focused on verifying the understanding of fundamental theoretical concepts, analytical frameworks (Mechanical vs. Free, Centaur Matrix, Cynefin), and critical discussion of mandatory reference texts. The evaluation will consider the ability to connect the different course themes and to argue with independent judgment. Optional mid-term assessment: evaluation based on the creation and presentation of a group Project Work (1-3 people). The project consists of redesigning a real decision-making process, using an AI tool of choice to automate the "mechanical" parts and redefine the role of the "free" human contribution. The oral exam will consist of the discussion of the project, evaluating: the theoretical coherence of design choices, the quality of human-machine delegation, the critical mastery of the AI tools used, and the originality of the proposed solution.

None in particular. The assignment of the final dissertation (thesis) in this discipline requires a strong interest in the frontier issues between management, technology, and ethics. Willingness to delve into international scientific literature (mostly in English) on the impact of Artificial Intelligence in organizations and decision-making processes is required. It is preferable, but not strictly mandatory, to have passed the exam with an excellent grade and to have demonstrated strong analytical and critical skills during the course, particularly in the drafting of the Project Work. Thesis topics may range from the analysis of corporate AI implementation cases, to the study of algorithmic biases in selection processes, up to the theoretical design of new "centaur" organizational models (human-machine hybrids).