πŸ³οΈβ€πŸŒˆ Ashwin (they/them)

Hello! I am an activist-researcher based in Barcelona, Catalunya. My work derives from algorithmic fairness and theories of participation to make socio-technical systems equitable.

My research has won the Best Paper Award πŸ† at FAccT, and been published in {A*, A} conferences including AIES, Hypertext, ICTD, INTERACT + more.

I am currently pursuing my MSc. at Universitat Pompeu Fabra, and building the next-gen recommender systems for travel experiences at Headout.

I am also a part of the Web Science and Social Computing Research Group, working under the supervision of Dr. Carlos Castillo towards my MSc. thesis, in collaboration with Departament de JustΓ­cia i Qualitat DemocrΓ tica, Generalitat de Catalunya.

Email  /  CV  /  Google Scholar  /  Twitter  /  Github

profile photo
πŸ“° Updates

πŸ”¬ Research
Queer In AI: A Case Study in Community-Led Participatory AI
Ashwin S, Organizers of QueerInAI

ACM FAccT, 2023 | πŸ“ Chicago, USA
Paper / Slides

πŸ† Best Paper Award

Critically examining how participatory design and intersectional tenets started and shaped QueerInAI's programs during the last five years, this work outlines lessons, opportunities and challenges for practitioners and theorists of participatory methods.

Bound by the Bounty: Collaboratively Shaping Evaluation Processes for Queer AI Harms
N Dennler, A Ovalle, Ashwin S, L Soldaini, A Subramonian, H Tu, W Agnew,
A Ghosh, K Yee, IF Peradejordi, Z Talat, M Russo, JDJDP Pinhal

ACM/AAAI AIES, 2023 | πŸ“ Montreal, Canada
Paper / Coverage in the AI Ethics Brief

Emerging from a participatory workshop (CRAFT Session) at FAccT 2022, this work covers positions and desires of the queer community on bias bounties, specifically questioning their ownership, incentives, accessibility and efficacy.

Erasing Labor with Labor: Dark Patterns and Lockstep Behvaiors on Google Play
Ashwin S, A Arun, P Malhotra, P Desur, A Jain, D H Chau, P Kumaraguru

ACM Hypertext, 2022 | πŸ“ Barcelona, Spain
Paper / Data / Poster / Slides / Coverage in the AI Ethics Brief

We uncover how install-incentivizing fraud on Google Play is facilitated via dark patterns and lockstep behaviors, possible due to the exploited labor of users and crowdworkers.

QueerInAI Demographic Survey 2022-23
Ashwin S, Organizers of QueerInAI

Led participatory design with QueerInAI's community to develop a new demographic survey for queer AI researchers and practitioners worldwide.

Through an intersectional lens, the survey facilitates understanding of issues and barriers covering several themes, including outness, social and legal climates, harassment, discrimination, and academic conferences.

Diagnosing Data from ICTs to Provide Focused Assistance in Agricultural Adoptions
Ashwin S, M Subramanian, A Agarwal, P Priyadarshi, S Gupta, K Garimella,
P Kumaraguru, S Kumar, R Kumar, L Garg, E Arya

ICTD, 2022 | πŸ“ Seattle, USA
Paper

Our work in collaboration with Digital Green diagnoses their ICT for disparities in adoption rates of farming practices, identifies potential sources of inequalities and provides implications for alleviating adoption rates of nearly a million farmers in India.

Hanging Out Online: Social Life During the Pandemic
Ashwin S, Grace Eden

IFIP INTERACT, 2021 | πŸ“ Bari, Italy
Paper / Slides

Dean's Special Award for significant achievements in BTech thesis (<2%)

Understanding how the shift from in-person to online computer-mediated social interaction presented opportunities and challenges for college students in India, generating implications for facilitating social serendipity, negotiating intimacy, mutual understanding and more.

βš–οΈ Advocacy

I belong to a marginalized caste-group (Bahujan) in India and identify as queer. Thus, my advocacy is focused around the regulation of AI systems that harm individuals at the intersection of caste, gender and sexuality.

  • AAAI '22: Hosted Sai Bourothu at QueerInAI's workshop to talk about the lack of consideration for queer and trans identities in the digitization of criminal records in India and its implications for AI systems built on top of that data.
  • NeurIPS '21: Moderated a Panel on Caste in Institutions and Tech at QueerInAI's workshop that involved scholars and activists from backgrounds spanning Anthropology, STS, Policy, Law and Economics. [Video] [Twitter Thread]
  • Helped draft QueerInAI's approach to AI Risk Management, where I elaborated on various historical, social and political considerations that are key to preserving the inclusive and democratic nature of participatory design.

Affiliations & Collaborators

                        


Page Design by Jon Barron