PROJECT SUMMARY

<aside> 🤳🏽 2024

Towards Feminist UX of AI: Provotyping New Design Patterns for Algorithmic Systems

Feminist UX of AI is a design research project using speculative design to imagine a UX of AI beyond commercial design conventions and informed by feminist HCI principles. Through participatory workshops, we co-created a collection of 20+ provotypes to serve as catalysts for critical reflection and dialogue. A project by Nadia Piet as part of the RAAIT program and supported by AIxDESIGN, you can access the library and learn more about the project here.

URL: Website

TIME: 11/2023 - 06/2024

TEAM: Nadia Piet (Research Lead), Christoph Bohne + Arda Awais (Design), Malak Sadek, Udita, Aleksandra (Research Support)

PARTNER: Responsible Applied AI Team @ Hogeschool Rotterdam under the guidance of Maaike Habers

UX of AI Library.png

</aside>

Preface

CONTEXT

Positioning the AI interface as a space of intervention

AI is increasingly shaping our lives. “*Just as we live in the city or the countryside, so too do we live in these systems to some degree or another.”* There’s a general sentiment of people growing critical of their lack of agency and control in algorithmic interactions and seeking improved transparency, empowerment, and ethical design in the digital landscape.

While AI as computationally and socially complex systems can feel elusive, the way people interact with said models is through designed, visible (often visual) user interfaces. In the words of Drucker in her paper on Interface and Interpretation, “digital interfaces are located and used as instruments, consciously or not, of institutionalized relations of power”.

Its real-life impacts, harms & joys all play out / are mediated through / unfold along the interface’s affordances and elements. The UX/UI of algorithmic systems are a bridge between intention and impact; the touchpoint between abstract design principles and people’s lived experiences.

Leaning into the interface as the playground / battleground, can we come to see AI interfaces as spaces for (design) intervention? How might user experience (UX) and user interface (UI) design support user agency, repair harms, safeguard data rights, and allow for user dissent?

Blue sky thinking beyond big tech conventions

Most thinking about how to design for AI originates within the commercial context of big tech companies - at times operating under 'attention economy' business models and mostly prioritizing a specific set of values (e.g. efficiency, profitability, growth, and scale) that may not align with users’ values and best interests.

“Design reproduces dominant relationships unless there is a deliberate effort to change them” —Dan Zollman

Even the UX discipline within these spaces is constrained to focus more on what’s ‘good for the business’ rather than what’s ‘good for the user’.

This is the breeding ground that UX best practices, conventions, and common frameworks flow out of, and It has come to dominate our thinking models around AI and technology at large.

While critique is broadly aligned on what we don’t want (deceptive patterns), there is less work available that identifies and imagines alternatives & new perspectives outside of these conventions; that which we do want.