Category: Virtual Agents

Alexa, are you listening... in the stacks?

12/22/2020

I'm happy to share a new article, "Alexa, Are You Listening? An Exploration of Smart Voice Assistant Use and Privacy in Libraries", published with co-author (and SLIS alum) Emma Davis about smart voice assistant use in libraries. Our research explores library use of smart voice assistant technologies in user services and programming, documenting many as-of-yet unresolved privacy issues that these technologies pose for patron communities and also library staff. We urge library workers to consider that smart voice assistants pose harm to many of our patrons as extensions of policing data networks, and argue that the LIS profession has a responsibility to actively engage questions of technological harms and data privacy before advocating adoption of emerging technologies like smart voice assistants in library services.

Read more in our full article, available open access:

Sweeney, M. E., & Davis, E. (2020). Alexa, Are You Listening? An Exploration of Smart Voice Assistant Use and Privacy in Libraries. Information Technology and Libraries, 39(4). https://doi.org/10.6017/ital.v39i4.12363

Interview with Radical AI podcast

6/17/2020

I had the absolute pleasure of talking with Jess and Dylan, the hosts of the Radical AI podcast, about the ethics of chatbots, virtual assistants, and emoji design. They were really gracious and fun to talk with, and I can't say enough good things about the quality of their podcast. I highly recommend subscribing to the Radical AI podcast and soaking up all of the wisdom from their conversations with leading scholars in technology and media studies. (A great podcast to use in class and assign for students as well!)

Subscribe or listen to the episode here!

What is socially responsible AI? Broad Science podcast ep available

4/9/2020

I was interviewed by the amazing folks at Broad Science for their podcast several months ago. Broad Science is "an initiative dedicated to making science inclusive, engaging, and intersectional," and focuses on "telling science stories from the voices and perspectives that often get overlooked and go untold." I love what they are all about, and recommend checking out their full list of podcast eps on soundcloud.

This episode entitled "What is socially responsible AI?" (aired April 3rd, 2020) features interviews with myself, Dr. Timnit Gebru, Surya Matu, and Dr. Kirk Bansak. The interviewers cover a ton of ground in this episode, including racial profiling, machine bias, interface design, and issues of diversity and inclusion in STEM. We discussed issues of gender and labor in virtual assistant design including the impacts of these technologies in society and broader questions about AI ethics. All of these interviews are fabulous; I start around 42:00, but definitely take the time to listen to the whole thing!

Thank you to Broad Science for reaching out to me, and wonderful job on this episode!

Available at Soundcloud: https://soundcloud.com/broad-science/what-is-socially-responsible-ai

Keynote, Library Research Seminar VII

10/22/2019

It was my pleasure to give a keynote at the Library Research Seminar VII event last week in Columbia, South Carolina. This event is sponsored by the American Library Association's Library Research Roundtable (LRRT), and was a really generative and generous space to brainstorm about research with LIS students, practitioners, and scholars of all kinds.
A big thank you to all who planned and executed this event, and much gratitude for inviting me to come share ideas with this warm community!

New article out about USCIS's Latina virtual assistant "Emma"

7/25/2019

My collaborator, Dr. Melissa Villa-Nicholas, and I are proud to announce that our new article,
"Designing the "good citizen" through Latina identity in USCIS's virtual assistant "Emma", is now published in the online issue of Feminist Media Studies!

This paper marks the first of three projects interrogating various cultural, political, and social aspects of Latina AI. Stay tuned for updates on this research!

*You can learn more about Dr. Melissa Villa-Nicholas's pioneering research on Latinx information histories and practices here!

Abstract:
Virtual assistants are increasingly integrated as “user-friendly” interfaces for e-government services. This research investigates the case study of the virtual assistant, “Emma,” that is integrated into the United States Citizenship and Immigration Services (USCIS) website. We conduct an interface analysis of Emma, along with the USCIS website, and related promotional materials, to explore the cultural affordances of Latina identity as a strategic design for this virtual assistant. We argue that the Emma interface makes normative claims about citizenship and inclusion in an attempt to “hail” Latinx users as ideal citizens. We find that the “ideal” citizen is defined through the Emma interface as an assimilated citizen-consumer that engages with digital technologies in ways that produce them as informationally “legible” to the state.

Cite as: Melissa Villa-Nicholas & Miriam E. Sweeney (2019) Designing the “good citizen” through Latina identity in USCIS’s virtual assistant “Emma”, Feminist Media Studies, DOI: 10.1080/14680777.2019.1644657

Harvard talk on Latina AI

4/20/2019

I had a wonderful time presenting with my research partner Melissa Villa-Nicholas on one of our projects about Latina AI at the Harvard Kennedy School on March 25th. Our talk focused on "Emma", the Latina virtual assistant used by the United States Citizenship and Immigration Services (USCIS) as a part of their e-government services. This presentation explores the cultural affordances of Latina identity as a strategic design choice in the Emma interface that extends citizenship and nation-building projects for the state, while masking underlying information and data gathering capabilities.

We were privileged to have Dana Chisnell, co-director of the Center for Civic Design, serve as a moderator for our talk. We felt very welcome, thanks largely to all of the hard work and planning of Vanessa Rhinesmith, the Associate Director of digitalHKS. Thank you to everyone who came and talked with us about the politics and surveillance implications of digital technologies designed to gather information about Latinx communities.

*Watch for our paper, "Designing the ‘good citizen’ through Latina identity in USCIS’s virtual assistant ‘Emma'", in Feminist Media Studies forthcoming later in 2019.

Cultural Affordances of “Emma”, USCIS’s Latina Virtual Assistant from digital HKS on Vimeo.

Google Assistant in the news

5/17/2018

Google's recent I/O developer conference on May 8th, 2018 made waves with the announced updates for the Google Assistant. Among these updates were new features to promote more "natural" conversational applications. Among these updates are six new voice options that include John Legend's musical voicings... a new twist to add some celebrity branding to the palette of customization. Other controversial adds are a "politeness" feature for aimed at children to "encourage polite conversation". Yikes! So much to unpack in all of this. Moneyish's Nicole Pesce reached out to me for some thoughts on these updates and what it all means for the landscape of feminized virtual assistants. Browse her write up of our (very brief) conversation on this here. More thoughts (and far more developed thoughts) to follow in some of my academic pieces under way!

Interview with ABC's Science Friction Podcast

8/11/2017

Several weeks ago ABC's Natasha Mitchell reached out to talk with me about AI, chatbots, gender, and sexism for the Science Friction podcast. You can check out the full podcast episode with my interview starting around minute 8, or read the corresponding article with the interview highlights. Thank you, Science Friction, for thoughtfully covering these issues and featuring my research!

Shout out from post, "When Robots are an Instrument of Male Desire"

7/18/2017

There has been surging interest in sexism in Artificial Intelligence design, largely because there are so many heinous examples to point to right now. This article, "When Robots are an Instrument of Male Desire," by Katherine Cross from The Establishment points to the debacle that was Tay AI, and loops in the standard players like Siri, Cortana, and even Microsoft's Ms. Dewey, with whom I am intimately familiar. I was pleased to come across this article and see some of my research on Ms. Dewey as a part of the conversation! Cheers, Katherine!

Interview, Popaganda: Fembots

1/5/2017

Bitch Media's Sarah Mirk interviews me about Microsoft's Ms. Dewey as a part of the Popaganda podast's episode about Fembots. I recommend the entire podcast, but if you don't have to time, be sure to tune in to my discussion about Microsoft's Ms. Dewey in the third segment, "Siri, are you real?"