What if we did AI differently?
Join INVI's new workshop series, co-hosted with Ethos Lab and explore how to 'Do AI otherwise'
The Institute for Wicked Problems (INVI) and ETHOS Lab are excited to announce a new workshop series that critically and playfully explores the role of artificial intelligence and data-driven methods in Danish politics. Bringing together researchers, students, artist, and practitioners, the sessions investigate how we practice AI Otherwise - challenging us to imagine a use of AI beyond established and conventional frames.
With each workshop, participants are invited into the 'machine room' of INVI's work and encouraged to participate in practical experiments related to using AI in Danish policymaking. The series explores tactics for testing bias and fairness (first session), visualization and representation of AI models (second session), and prompts scholarly co-curation of ethical guidelines for how we use of AI in decision-making (third session). Finally, it encourages participants to cultivate a new aesthetic and rhetoric of AI (fourth session) as an alternative to the dominant socio-technical imaginaries that typically construe AI solutions within capitalist, masculine, and glitchy-blue frames. How might concepts such as 'slowness', 'feminism', 'cyberpunk', 'decolonization', and the like help us reconfigure cultural and public imaginaries of AI and help us find AI vernaculars that are humble and build on a collective ethics of care?
Participants will thus be taken on a journey that uses critique and practical experiments as an engine to develop new ideas, toolkits, and perspectives. These will potentially be published at the end of the series as an open resource that can be used by anyone, who works with applications of AI in a Scandinavian political context.
Join us in shaping alternative AI imaginaries through this series of playful research workshops, co-hosted by INVI and ETHOS Lab.
About INVI
INVI is an independent, apolitical think tank based in Copenhagen, developing tools and toolkits to help tackle society's most complex problems. We strive to rethink political systems from critical and innovative perspectives that blend systems thinking, foresight, social innovation, participatory policymaking and computational anthropology. Read more here. R&D Director Sofie Burgos-Thorsen is co-hosting.
About ETHOS Lab
ETHOS Lab is a critical feminist methods laboratory, based at IT University of Copenhagen, dedicated to experimentation at the intersection of digital methods, ethnography, and speculative fabulation. We rework the ways that social and digital worlds collide, exploring the intersections, frictions, seams, and possibilities for a more just (&feminist, &queer, &decolonial, &postcapitalist) techno-mediated reality. Read more here.
Events
The workshop series will consist of four independent events that in turn invite students, researchers, artists and AI practitioners to playful sessions. Participation is open and free, but necessitates signing-up to secure a seat. Seats are assigned on a first come, first served basis.
All events are co-hosted by INVI's Director of R&D, Sofie Burgos-Thorsen, and lab manager at ETHOS Lab, Henriette Friis and take place in INVI's workspace in Snaregade 10B, 1208 København. Already now you can register for the first two events!
10. April, 15.00-18.00
Session #1: Bias in large language models- examining INVI's Model for Wicked Problems from within (Read more and register here)
INVI's Director of R&D, Sofie Burgos-Thorsen, briefly introduces the think tank's work with creating a Model for Wicked Problems and open a set of bias-related questions that this use of large language models brings about. Guest speaker Jonathan Rystrøm from Oxford Internet Institute shares notions of bias and 'algorithmic fairness' to frame the session.
Samples of INVI's datasets are shared with students, who are invited to explore bias-related questions and collaboratively curate a synthetic dataset that will be used at the end of the session to test assumptions about biases in the embedding models. We end with a plenary critical reflection about possibilities and limitations in bias testing LLMs.
- Invited: Students
- Language: Danish
- Guest: Jonathan Rystrom from Oxford Internet Institute.
April 29, 15.00-18.00.
Session #2: Visualizing wickedness - a hands-on workshop on critical and creative information design (Read more and register here)
INVI's Director of R&D, Sofie Burgos-Thorsen, briefly introduces the think tank's work with creating a Model for Wicked Problems and open a set of questions related to the power of visualizing data. Guest speaker Matilde Ficozzi from Techno-Anthropology Lab at AAU shares examples of data visualization and data physicalization practices to frame the session.
Samples of INVI's datasets are shared with students, who are invited in groups to experiment with visualizing data about 'wicked problems' in novel ways, creating low-fi visual prototypes. We end in a plenary, where groups present their work, and we reflect on design choices together.
- Invited: Students, designers, artists
- Language: English
- Guest: Matilde Ficozzi from the Techno-Anthropology Lab.
May 8, 15.00 -18.00.
Session #3: AI ethics in policymaking - from principles to playbook
Bringing the brightest minds of AI journalists, artists, practitioners, and humanities and social science researchers, we critically explore how to turn ethical principles into an actionable playbook for using AI-based solutions in Danish policy. How can we use AI-based solutions in a way that promotes inclusion, fairness, and lifts up more voices-while also facing the many risks and challenges of AI with open eyes? In this third session, we take INVI's Model for Wicked Problems as a jumping-off point and play through a set of hypothetical scenarios, using ChatGPT, to provoke conversation and flesh out a playbook for what ethical principles should guide us when we use AI in Danish policymaking.
- Invited: Researchers, designers, tech journalists and practitioners.
- Language: English
- Guest: Christian Villum.
20. May, 15.00-18.00.
Session #4: Rethinking AI aesthetics: Building alternative vernaculars
How do we create a new aesthetic for AI-both visual and rhetorical? What narratives and expectations around AI are we helping to shape through the 'stories' we contribute to the world? How can we create new references and images that move beyond the ones dominating today?
In this workshop, a mix of students and researchers are invited to use speculative design to develop rhetorical and visual alternatives to today's dominant AI discourse and aesthetic.
- Invited: Researchers, designers, practitioners, students
- Language: English
- Guest: TBD.
Workshop format
- Timeframe: 3 hours in the afternoon from 15.00 to 18.00.
- Location: Institute for Wicked Problems in Snaregade 10b.
- Seats: 15-25 per workshop.
- Drinks and snacks are provided.
- Structure: Lecture to frame the session (30 min), introduction to the workshop (20 min), group work (30 min), break (10 min), group work (60 minutes), plenary reflection (30 min).
Positionality statement
As an apolitical and independent think tank, at INVI we want to open ourselves up to critique and use our work with AI as an opportunity to spark broader, important conversations about the principles we, as a society, believe should guide the use of AI in Denmark. Through our format, Model for Wicked Problems, we are among the first in a Danish context to work directly with the political system on using artificial intelligence to address societal problems, with real stakes and real people involved.
We see that as a major responsibility, and we approach it with great humility. We don't believe we have all the answers ourselves, but we use our position as a think tan, with one foot in research and the other in the political system, to bring together the people and communities in Denmark that we believe are relevant for a collective conversation about AI use in Danish policy development, inviting scholars and students as well as artists, designers, tech journalists and AI practitioners.