The digital revolution brought tremendous changes in everyone’s lives, shaping our society. Just think about the Internet, Smartphones and Social Networks, to cite some innovations that have changed our lives and way of working. Artificial Intelligence represents the upcoming technological revolution, being a growing technology evolving so fast, to have the potential to revolutionise society once again and, as such, deeply transform our way of working. The widespread growth of AI is a worldwide phenomenon that is now present in almost every field of humanity and its democratisation was primarily brought by a specific type of AI, Large Language Models (LLMs). This technology, like the well-known ChatGPT, has given access to such powerful technology most easily, using our own language, to achieve results of each kind, and in less than 5 years we have witnessed a shift towards their usage and implementation also in the Design field. This thesis aims to analyse and experiment the implementation of LLMs in the early stages of the Design Process (DP), understanding its impact on the outcomes of these phases and within the collaborative sphere. LLMs have the potential to impact in the early and most strategic phases of the DP, the Discover and Define phases. They can change the way Designers approach a new project, the ways in which we gather information, extract insights, and understand the context of the project and the needs of a user, to create better and more conscious products. The thesis has been structured into two main sections: a theoretical background part, and an experimental section, divided into two distinct activities. The theoretical part is a complete and exhaustive literary review on the subject, understanding what Design Thinking and DP are, the state of the art of AI and LLMs, to contextualise the opportunities and challenges that come from the Design field in the early stages of the DP, and relate them with the current findings of the implementation of AI, focusing on its impact in several key steps of the initial stages of the DP, its potentialities and implications. The experimental research unpacks the concepts analysed in the literary review, exploring and testing in first person the capabilities of LLMs, and their impact in the DP concerning the established tools of the Discover and Define phases. We underline the LLM’s ability to augment the Designer’s notions and capabilities, underlining the identified roles of the AI and its usage criticalities. The assumptions were, then, assessed and expanded in a tailored workshop, in which 86 participants divided into groups, had to redesign a product, using ChatGPT throughout the Discover and Define phases. The obtained results were evaluated and analysed, clarifying the new roles of the Designer and the Artificial Intelligence, the related opportunities and implications that emerge from the collaboration Human-Machine, and, consequently, the outlines of an Integration Framework of Large Language Models within the Design Process.
La rivoluzione digitale ha portato enormi cambiamenti nella vita di tutti, plasmando la nostra società. Basti pensare a Internet, agli Smartphone e ai Social Network, per citare solo alcune delle innovazioni che stanno modificando la nostra vita e il nostro modo di lavorare. L'Intelligenza Artificiale rappresenta la attuale e prossima rivoluzione tecnologica, essendo una tecnologia in continuo sviluppo che si evolve così velocemente da avere il potenziale per rivoluzionare ancora una volta la società e, come tale, trasformare profondamente il nostro modo di lavorare. La crescita esponenziale dell'IA è un fenomeno globale, ormai presente in quasi tutti i campi dell'umanità, e la sua democratizzazione è stata portata avanti particolarmente da un tipo specifico di IA, i Large Language Models (LLM). Questa tecnologia, così come il noto ChatGPT, ci ha permesso di accedere a questa potente tecnologia nel modo più immediato e semplice, utilizzando il nostro stesso linguaggio, in modo da ottenere risultati di ogni tipo. In meno di 5 anni abbiamo quindi assistito ad un crescente interesse verso il loro utilizzo e la loro implementazione, anche nel campo del Design. Questa tesi si propone di analizzare e sperimentare l'implementazione dei LLM nelle prime fasi del Design Process (DP), analizzandone l'impatto sui risultati di queste fasi e nella sfera collaborativa. I LLM hanno il potenziale di impattare particolarmente nelle fasi iniziali e strategiche del DP, le fasi iniziali: Discover e Define. I Large Language Models possono cambiare il modo in cui i designer approcciano un nuovo progetto, i modi in cui raccolgono informazioni, sviluppano la loro creatività e comprendono il contesto del progetto e le esigenze degli utenti, per creare prodotti migliori e più consapevoli. La tesi è stata strutturata in due sezioni principali: una parte teorica di base e una sezione sperimentale, suddivisa a sua volta in due attività distinte. La parte teorica è una rassegna letteraria completa ed esaustiva sull'argomento, per comprendere cosa sono il Design Thinking e il DP, lo stato dell'arte dell'IA e dei LLM, per contestualizzare le opportunità e le sfide che provengono dal campo del Design nelle prime fasi del DP, e metterle in relazione con gli attuali risultati dell'implementazione dell'IA, concentrandosi sul suo impatto in diversi momenti chiave delle fasi iniziali del DP, le sue potenzialità e implicazioni. La ricerca sperimentale dispiega i concetti analizzati nella rassegna letteraria, esplorando e testando in prima persona le capacità dei LLM e il loro impatto nel DP rispetto agli strumenti consolidati delle fasi Discover e Define. Evidenziando così la capacità del LLM di aumentare le nozioni e le prospettive del Designer, evidenziando i ruoli identificati dell'IA e le sue criticità d'uso. Le ipotesi sono state poi valutate e ampliate in un workshop su misura, in cui 86 partecipanti, divisi in gruppi, hanno dovuto riprogettare un prodotto, utilizzando ChatGPT durante le fasi Discover e Define. I risultati ottenuti sono stati valutati e analizzati, chiarendo i nuovi ruoli del Designer e dell'Intelligenza Artificiale, le relative opportunità e implicazioni che emergono dalla collaborazione Uomo-Macchina e, di conseguenza, i contorni di un Frameworks di Integrazione dei Large Language Models all'interno del Design Process.
Design and AI revolution : the impact of large language models in the early stages of the design process
Mastroianni, Andrea
2022/2023
Abstract
The digital revolution brought tremendous changes in everyone’s lives, shaping our society. Just think about the Internet, Smartphones and Social Networks, to cite some innovations that have changed our lives and way of working. Artificial Intelligence represents the upcoming technological revolution, being a growing technology evolving so fast, to have the potential to revolutionise society once again and, as such, deeply transform our way of working. The widespread growth of AI is a worldwide phenomenon that is now present in almost every field of humanity and its democratisation was primarily brought by a specific type of AI, Large Language Models (LLMs). This technology, like the well-known ChatGPT, has given access to such powerful technology most easily, using our own language, to achieve results of each kind, and in less than 5 years we have witnessed a shift towards their usage and implementation also in the Design field. This thesis aims to analyse and experiment the implementation of LLMs in the early stages of the Design Process (DP), understanding its impact on the outcomes of these phases and within the collaborative sphere. LLMs have the potential to impact in the early and most strategic phases of the DP, the Discover and Define phases. They can change the way Designers approach a new project, the ways in which we gather information, extract insights, and understand the context of the project and the needs of a user, to create better and more conscious products. The thesis has been structured into two main sections: a theoretical background part, and an experimental section, divided into two distinct activities. The theoretical part is a complete and exhaustive literary review on the subject, understanding what Design Thinking and DP are, the state of the art of AI and LLMs, to contextualise the opportunities and challenges that come from the Design field in the early stages of the DP, and relate them with the current findings of the implementation of AI, focusing on its impact in several key steps of the initial stages of the DP, its potentialities and implications. The experimental research unpacks the concepts analysed in the literary review, exploring and testing in first person the capabilities of LLMs, and their impact in the DP concerning the established tools of the Discover and Define phases. We underline the LLM’s ability to augment the Designer’s notions and capabilities, underlining the identified roles of the AI and its usage criticalities. The assumptions were, then, assessed and expanded in a tailored workshop, in which 86 participants divided into groups, had to redesign a product, using ChatGPT throughout the Discover and Define phases. The obtained results were evaluated and analysed, clarifying the new roles of the Designer and the Artificial Intelligence, the related opportunities and implications that emerge from the collaboration Human-Machine, and, consequently, the outlines of an Integration Framework of Large Language Models within the Design Process.File | Dimensione | Formato | |
---|---|---|---|
Design & AI Revolution. The Impact of Large Language Models in the Early Stages of the Design Process.pdf
accessibile in internet solo dagli utenti autorizzati
Dimensione
11.91 MB
Formato
Adobe PDF
|
11.91 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/219075