Artificial Intelligence (AI) has become increasingly present and integrated in our daily lives, with complex and often unpredictable solutions. This complexity and unpredictability make it difficult for users to understand, trust, and adopt such solutions with success. The course aims at introducing methodologies and techniques to design and build intelligent interactive systems that are usable, useful, and trustworthy, also reflecting on the impact that AI applications have and will have on their users. Building on prior knowledge of Human-Computer Interaction, the course adopts a hands-on, project-based approach: students design, prototype, and evaluate an AI-powered interactive system through a human-centred process, while critically using AI tools as part of their own design and development workflow. The course sits at the intersection of HCI and AI, with attention to both the practical design of usable AI systems and the broader societal and ethical implications of deploying them.

By the end of the course, students will be able to: - Articulate what makes human-AI interaction different from traditional human-computer interaction, recognising the design challenges introduced by non-deterministic and probabilistic systems. - Apply established human-AI interaction guidelines to analyse and evaluate AI-powered interactive systems. - Design and prototype AI-augmented interactive systems through a human-centred process, including identifying whether and where AI is an appropriate fit for a domain. - Evaluate AI-powered systems with attention to AI-specific concerns, including calibrated trust, appropriate reliance, error recovery, and explainability. - Critically reflect on the ethical, societal, and second-order implications of AI-powered systems. - Use AI tools effectively and reflectively as part of their own design and development process.

- Knowledge of Human-Computer Interaction principles and methods as typically covered in an introductory Human-Computer Interaction course. - Basic programming skills (for example, Python and JavaScript). - Attitude towards working in teams.

The course is organized in five modules: 1. Foundations of Human-AI Interaction: what makes it different from Human-Computer Interaction, basic concepts and principles (1 credit). 2. Designing for and with AI: HCI methods reframed under an AI lens and guidelines, prototyping AI, real-world evaluation, AI-specific challenges, critical and ethical aspects (2 credits). 3. LLMs, prompting, and agentic systems: prompt engineering as interaction design, tools and patterns for building interactive systems (1.5 credits). 4. Human-AI Interaction in specific domains, such as education, writing, health, creativity, accessibility (1 credit). 5. Speculative design and futures for Human-AI Interaction: extrapolating consequences of AI-powered systems (0.5 credits).

The course will have lectures and exercises in class (4 credits), and lab hours devoted to the development of an AI-based interactive prototype (2 credits). During the course, students will complete three small practical exercises to go deep into selected topics: (i) reading and debating a selection of papers in class, (ii) performing an evaluation of an existing AI-powered system against established human-AI interaction guidelines, and (iii) prepare a critical synthesis of a speculated AI future. The project to be developed in the lab hours, instead, will be realised in group and will consist of the creation of a prototypical human-centred AI system. The project will be a) created using AI in the process and b) constrained by a specific interaction pattern in a chosen domain. The project idea will be student-led, with teachers approving it, and will follow a human-centred process that includes an explicit AI-fit analysis (i.e., justifying whether and where AI is appropriate for the identified need). Students will document and critically reflect on their own use of AI tools throughout the project. During lab hours, teachers will provide continuous feedback on the project to keep the work smooth. During the course, communications and project development will adopt contemporary solutions and tools (e.g., GitHub, Telegram, etc.). Lectures will be video-recorded and made available after each class. In-person attendance at the labs is essential, as projects will be developed and discussed in those hours.

Course slides and related materials (e.g., links, readings, …) will be available in a dedicated webpage on https://elite.polito.it. Additional resources are available on the web and will be pointed out during lectures (https://www.microsoft.com/en-us/haxtoolkit/, https://pair.withgoogle.com/guidebook, etc.).

Slides; Esercizi; Esercizi risolti; Video lezioni dell’anno corrente; Strumenti di collaborazione tra studenti;

Modalita di esame: Prova orale obbligatoria; Elaborato scritto individuale; Elaborato progettuale in gruppo;

Exam: Compulsory oral exam; Individual essay; Group project;

... The exam assesses the theoretical knowledge and practical skills described in the previous fields through a group project, the three practical exercises, and an oral discussion. The group project is developed during the lab hours and consists of a prototypical human-centred AI system, accompanied by its supporting documentation. The project is assessed on: (i) the soundness of the human-centred process followed, including the explicit AI-fit analysis; (ii) the quality of the prototype as an interactive AI-powered system, with attention to the chosen interaction pattern and to the application of human-AI interaction guidelines; (iii) the quality and rigour of the in-context evaluation, with attention to AI-specific concerns such as calibrated trust, appropriate reliance, error recovery, and explainability; and (iv) the documented and reflective use of AI tools throughout the design and development process. Each group delivers a final presentation and a written report, and a single grade is assigned to the group. Individual contributions may be considered when there is clear evidence of unequal participation. The three practical exercises are the one completed in class, during the semester, which will need to be delivered at specific deadlines during the semester. The oral exam lasts approximately 20-30 minutes per group and takes place in a classroom or laboratory. The oral exam takes the form of a group presentation and demonstration of the project, followed by individual questions to each student. The goal is to assess the students understanding of course concepts, the ability to connect them to the work produced during the course, and the ability to reason about current human-AI interaction research and practice. Grading is expressed on a scale from 0 to 30, distributed as follows: up to 15 points for the group project (a single grade is assigned to the group, with individual adjustments where there is clear evidence of unequal participation); up to 9 points for the three practical exercises, completed and graded individually (3 points each); and up to 6 points for the oral exam, which takes the form of a group presentation and demonstration of the project followed by individual questions, with each student receiving their own oral grade. Up to 2 additional points may be awarded for particular quality and richness across the three components, allowing a grade of 30 cum laude.

Gli studenti e le studentesse con disabilita o con Disturbi Specifici di Apprendimento (DSA), oltre alla segnalazione tramite procedura informatizzata, sono invitati a comunicare anche direttamente al/la docente titolare dell'insegnamento, con un preavviso non inferiore ad una settimana dall'avvio della sessione d'esame, gli strumenti compensativi concordati con l'Unita Special Needs, al fine di permettere al/la docente la declinazione piu idonea in riferimento alla specifica tipologia di esame.