About     Art/Projects     Talks     Writing

Imagining Feminist Interfaces

The multi-year project Imagining Feminist Interfaces (2019-), led by our collective tendernet, is a zine series and a workshop series that explores what participatory methods for doing research might look like in the development of AI/ML.

You can read more about the project in our Adjacent article “Imagining Feminist Voice Technologies.”

Imagining Feminist Interfaces: Voice interactions for agency, empowerment, and justice

Historically, technology has created or shaped the conditions for violence and oppression. Many technologies further entrench – rather than question– existing systems of power. In our work, we’ve explored how an intersectional feminist lens might help us reimagine the design of voice technologies.

Instead of nudging us to buy more toilet paper on Amazon, could a voice assistant initiate difficult conversations? Could it attune us to our natural surroundings and ecology? Could it draw our attention to technological and human infrastructure, rather than erase the labor and resources required to maintain it? Could it be bilingual? Could a small language community design it for their own needs?

In a series of workshops, we’ve worked together to imagine and prototype feminist alternatives for voice technologies. Through a series of speculative design exercises, we discussed what voice technologies and software could look like if we designed them in line with the central commitments of intersectional feminism: participation, agency, embodiment, equity, empowerment, plurality, and justice.

The default “voice” of many of our technologies has typically been female: From early telephone operators connecting people through wires, to office secretaries and personal assistants, to voice assistants like Alexa and Siri. Often these disembodied female voices contribute to gendered fantasies of women occupying subservient “caregiver” or “secretary” roles.

What’s more, in a 2017 survey of popular voice assistants, journalists found that when confronted with abusive, violent, or sexually explicit language, most voice assistants responded by evading the comment (“Let’s change the topic”) or defusing the situation with humor (“I’d blush if I could”). Such interactions suggest that Siri or Alexa were not designed to confront our biases, model difficult conversation, or educate us.

In our collective’s zine, we describe some of the principles that are central to designing feminist interactions and think through strategies for integrating those principles into voice technologies.

Say hi ︎ baricks (at) protonmail.com