The dictionary was the outcome of my Master’s Thesis project at Aalto University. The abstract, and full thesis document can be found below.
<aside> 📅 - 2021
<aside> ⌛ - 6 months
<aside> 🔖 - Design Research
Artificial Intelligence (AI) is omnipresent, with its diverse functionality used in fields ranging from healthcare to law to social media. Its versatility has allowed it to be used as a tool to scale technological solutions and achieve widespread change. However, AI and the algorithms behind it also hold significant risks, especially to those who are already vulnerable. Within the contexts of Necolonialism and Western-centric models of technological development, AI can propagate bias and oppression under the guise of development and progress. This thesis aims to map the vocabulary of AI through a pluriversal lens. It examines normative visions of AI and technological development, who they are built for; and redesign our vocabulary accordingly . The research questions if the current state of AI allows for pluriversality, and how pluriversality could be articulated in the development of AI. This thesis employs a critical approach towards AI, looking to the pluriverse as a more inclusive concept to generate alternative paths for technological development. The format of the thesis follows that of a dictionary, where keywords from critical AI discourse are defined and explored. The dictionary is intended as a primer for questioning the normative understanding of AI development and deployment, as well as articulating how AI solutions could be flawed, universal, exclusionary and discriminatory. Through a mapping and thematic analysis of critical AI literature, five themes emerged, forming the first set of entries in the dictionary: participation, ethics, accountability, explainability as well as the term AI itself. These themes offer insights and examples in relation to AI, with their alternate meanings and interpretations bringing forth pluriversal and critical perspectives. The second set of entries aim to give a provocative, alternate view of AI futures. Generated by an open- source AI based on the GPT-3 language model, these words describe a disruptive vision for AI. Using design fiction co-authored with the AI, storytelling as a method is employed to inspire the reader to imagine unconventional, decolonial and pluriversal forms of AI development. This thesis aims to highlight the inequalities of oppressive algorithms, and the importance of inclusive and pluriversal language and terminology in articulating future technological imaginaries.
The full thesis document can be found (and downloaded) below.