Inventions throughout history have redefined evolution and the perception of what it means to be human. Today, Artificial Intelligence (AI) is the latest destabilising force, rapidly transforming the processes of design and communication. This thesis explores the intersections among AI, architecture, and urban design through a feminist theoretical-critical perspective, based on the assumption that AI is not a neutral tool but a reflection of the power structures from which it learns. By investigating how these technologies expose and amplify long-standing architectural biases, such as the centrality of the masculine as a universal norm and the invisibility of diverse bodies, this research asks how feminist methodologies can influence AI to promote a more equitable and just spatial future. The core of this work is the application of feminist spatial practices, which define architecture as a cultural artefact that reflects social hierarchies, rather than as a purely technical act. However, the research also highlights the significant difficulties in discussing these themes within the architectural field, from the specific indignation over ‘selling the soul of architecture’ to AI, to the general resistance to the term ‘feminism’ itself. The thesis is structured into three distinct parts that place these themes in a critical dialogue. The first part, the Scope, establishes the theoretical framework by examining gender studies and intersectionality, utilising a comparative analysis of Milan, Bologna, and Vienna to evaluate how urban policies translate feminist theory into spatial practice. The second part, the Tool, investigates the role of AI in the AEC sector, deconstructing the ‘Bias Loop’ involving data, algorithms, outputs, and humans to explore how we can ‘teach the machine to Care.’ Finally, the work arrives at the definition of a Manifesto - both written and digital form -, a call for a new collaboration built on a three-way partnership: those who create space, AI, and the community. The heart of the proposal lies in the role of users as ‘tipping points’ and critical mediators: human sensitivity, empathy, and care are introduced as necessary infrastructures for navigating the opacity of the ML pipeline. In support of this vision, a narrative and technical device is introduced: an online AI-based chatbot that critically subverts the technical, non-emotional personality of AI chatbots.
Le invenzioni nel corso della storia hanno ridefinito la percezione di cosa significhi essere umani. Oggi, l’Intelligenza Artificiale (IA) si presenta come l’ultima forza destabilizzante, trasformando rapidamente i processi di progettazione e comunicazione nell'ambiente costruito. Questa tesi esplora le intersezioni tra IA, architettura e progettazione urbana attraverso una prospettiva teorico-critica femminista, basata sulla premessa che l’IA non sia uno strumento neutrale, ma un riflesso delle strutture di potere da cui apprende. Indagando come queste tecnologie espongano e amplifichino pregiudizi di lunga data - come la centralità del maschile come norma universale e l'invisibilità dei corpi diversi - questa ricerca si domanda come le metodologie femministe possano influenzare l'IA per promuovere un futuro spaziale più equo e giusto. Il fulcro del lavoro è l’applicazione delle pratiche spaziali femministe, che definiscono l'architettura come un artefatto culturale riflettente delle gerarchie sociali, piuttosto che un atto puramente tecnico. Tuttavia, la ricerca evidenzia anche le significative difficoltà nel discutere questi temi all'interno del campo architettonico, dall’indignazione specifica del ‘vendere l’anima dell’architettura’ all’IA, alla più generale resistenza del termine stesso ‘femminismo’. La tesi è strutturata in tre parti distinte che pongono questi temi in un dialogo critico. La prima parte, il Campo di Applicazione, stabilisce il quadro teorico esaminando i gender studies e l’intersezionalità, utilizzando un’analisi comparativa tra Milano, Bologna e Vienna per valutare come le politiche urbane traducano la teoria femminista in pratica spaziale. La seconda parte, lo Strumento, indaga il ruolo dell’IA nel settore architettonico, decostruendo il ‘Loop del Bias’ che coinvolge dati, algoritmi ed esseri umani per esplorare come ‘insegnare alla macchina a prendersi cura’ (Teaching the Machine to Care). Infine, il lavoro giunge alla definizione di un Manifesto (in forma scritta e digitale), un invito a una nuova collaborazione costruito su una partnership a tre vie: chi crea lo spazio, l’IA e chi vive entrambi, ovvero la comunità. Il cuore della proposta risiede nel ruolo degli utenti come ‘punti di svolta’ e mediatori critici: la sensibilità umana, l’empatia e la cura sono introdotte come infrastrutture necessarie per navigare nell'opacità degli algoritmi. A sostegno di questa visione, viene introdotto un dispositivo tecnico: un chatbot basato sull'intelligenza artificiale online che sovverte la comune percezione dei chatbot.
Blind spots. Feminist perspectives on AI in architecture and urban design
Novotny, Hannah;Pannella, Maria
2024/2025
Abstract
Inventions throughout history have redefined evolution and the perception of what it means to be human. Today, Artificial Intelligence (AI) is the latest destabilising force, rapidly transforming the processes of design and communication. This thesis explores the intersections among AI, architecture, and urban design through a feminist theoretical-critical perspective, based on the assumption that AI is not a neutral tool but a reflection of the power structures from which it learns. By investigating how these technologies expose and amplify long-standing architectural biases, such as the centrality of the masculine as a universal norm and the invisibility of diverse bodies, this research asks how feminist methodologies can influence AI to promote a more equitable and just spatial future. The core of this work is the application of feminist spatial practices, which define architecture as a cultural artefact that reflects social hierarchies, rather than as a purely technical act. However, the research also highlights the significant difficulties in discussing these themes within the architectural field, from the specific indignation over ‘selling the soul of architecture’ to AI, to the general resistance to the term ‘feminism’ itself. The thesis is structured into three distinct parts that place these themes in a critical dialogue. The first part, the Scope, establishes the theoretical framework by examining gender studies and intersectionality, utilising a comparative analysis of Milan, Bologna, and Vienna to evaluate how urban policies translate feminist theory into spatial practice. The second part, the Tool, investigates the role of AI in the AEC sector, deconstructing the ‘Bias Loop’ involving data, algorithms, outputs, and humans to explore how we can ‘teach the machine to Care.’ Finally, the work arrives at the definition of a Manifesto - both written and digital form -, a call for a new collaboration built on a three-way partnership: those who create space, AI, and the community. The heart of the proposal lies in the role of users as ‘tipping points’ and critical mediators: human sensitivity, empathy, and care are introduced as necessary infrastructures for navigating the opacity of the ML pipeline. In support of this vision, a narrative and technical device is introduced: an online AI-based chatbot that critically subverts the technical, non-emotional personality of AI chatbots.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Pannella_Novotny.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Blind Spots.
Dimensione
57.38 MB
Formato
Adobe PDF
|
57.38 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/10589/251720