TALKING TRUST
Trust, Responsibility, AI, Agency & Autonomy
An Arts-Based approach to the TAS Usecase Library
This collection of videos represents part of the UKRI TAS Hub funded project. Each video is a discussion with an artist who has worked on the TAS project. They discuss their work in the context of the Use-Case Library as a way to engage, provoke and encourage others to both think about the nature of Trustworthy Autonomous Systems (TAS), but also as a way to perhaps show the different ways that interdisciplinary researchers might want to add their own content & use the Use-Case Library.
Each artist had a short time (a week) to develop their design, each had a different approach, with some expanding and evolving upon their work theoretically, with others taking a more practice-based approach to highlight Trust in the process. Additionally the artists were also asked to add an entry into the Use-Case Library relating to a design.
The interviews were carried out by Dr Alan Chamberlain (University of Nottingham). These videos are an output of the project targeting the TAS Symposium, the use-case entries may be updated until the end of the project. Some artists may provide more than one example, while other artists have opted to attend the TAS All Hands event, to discuss their work in person. Link to Alan's Introduction
We would like to acknowledge the following TAS Hub Agile projects who all played a part in this: TAS Benchmarks Library and Critical Review, Co-Design of Context-Aware Trustworthy Audio Capture & TAS RRI. Without the insights from these other projects we would not have been able to articulate a response that covered the range of themes and topics that are covered in these conversations.
The artists involved in this project were - Rachel Jacobs, Ali Hossaini, Richard Ramchurn (Albino Mosquito), Roma Patel & Rachel Ramchurn (Makers of Imaginary Worlds). Click on an artist to link to the overview of their work.
Each video has been captioned and comes with its own transcript, this means the this collection can be searched, using words and phrases, we have timestamped the transcriptions. These are currently being used in our research in a publication in progress.
You can access our recent publication here A Practical Taxonomy of TAS-related Usecase Scenarios
This work was commissioned by the University of Nottingham as part to the TAS Hub, please contact Alan Chamberlain and the Artists in Residence prior to re-use. To enable re-use we have used a Creative Commons agreement.
Rachel Jacobs
FUTURE MACHINE: Technology for When the Future Comes
Future Machine is a large sculptural artwork, an octagon on a rectangular base on wheels, made of ash, oak, steel, brass, rubber and copper, with a hand crank, lever and dials. It collects messages for the future, plays the messages left by people in the past, prints out stories for people to take away with them, and plays sounds that respond to live data, from weather and pollution sensors on a pole at the back of the machine, capturing and interpreting climate and weather data. A webcam captures images of the horizon, and audio is captured in the form of the 'messages for the future' recorded by the people that gather around when it appears in five places across England - in Nottingham, London, Oxfordshire, Cumbria and Somerset at the same time every year for 30 years.
Creating a sustainable, net zero interactive device appears near impossible, particularly one that will run for 30 years. This speculative design project proposes to implement a set of design challenges developed by the creative team who created the Future Machine, to explore how they might turn Future Machine into a trustworthy autonomous system, whilst maintaining its entangled, slow and sustainable nature. Asking the question can we design autonomous systems to support human-nature-technology entanglements.
Link to Rachel's entry into the Usecase Library (pdf)
Beyond GROUPTHINK: At the intersection of Trust, Agency & Autonomy
GROUPTHINK enacts a liminal moment in the development of engineering. Technology has always groped at biology, but now it aims for convergence. There are several vectors for the merger of technology and biology, and within most lies a complex merger of previously separate domains within and outside engineering. AI is the enabling factor, as it is possible to conceive, but not realise, advances in bioengineering which enable physiological interfaces with networks, individuals, collectives and computing resources, notably AI itself. GROUPTHINK dissects this historically liminal moment by posing a near future scenario in a plausible artistic context. Media are a product, but they invoke an active process - mediation - which severs the value associated with live experience which itself is a marker of authenticity. Uniqueness, dialogue and participation are from the framework of live interaction. They build trust, or more accurately, they develop the dynamics where trust resides. GROUPTHINK Publication & GROUPTHINK Video. We also think the people might be interested in this film, The First directed by Ali Hossaini and Written by Luca Viganò.
Link to Ali's entry into the Usecase Library (pdf)
Richard Ramchurn (Albino Mosquito)
Synthetic Reflections: Decoding Performative Boundaries
This installation delves into the role of identity and trust within AI avatars. It poses the question to viewers: where does the human factor come into play in virtual performances? The installation comprises a series of thought-provoking on-camera interactions that have been collaboratively created with AI systems and virtual production techniques, exploring different configurations of authorship, performance, and voice. The content of the installation aims to prompt audiences to question the presence of the human within the performance and discern the involvement of AI. It delves into themes of identity and ethnicity, both in the content itself and in the process of its creation. The practical application of AI tools in the creation process has brought to light biases within datasets, resulting in particular stereotypes in images, as well as technological limitations that do not adequately accommodate diverse skin colours. Central to the installation is the exploration of trust within AI. As viewers engage with the artwork, they are confronted with the nuanced complexities of placing trust in AI systems. The installation urges individuals to critically examine the boundaries of trust, highlighting the potential vulnerabilities and ethical implications that arise in the realm of AI-driven performances. The audience, as users, engages with the artwork by attempting to comprehend what they see and hear. The intentional design of the installation challenges conventional expectations of human interaction and performance, inviting viewers to explore new dimensions of engagement.
Link to Richard's entry into the Use-Case Library (pdf)
(in progress as the piece develops)
Roma Patel & Rachel Ramchurn
The Close Encounter
Our installation with 'NED' - the Never-Ending Dancer, is often a children and families first close encounter with a robot that is not a toy. The aesthetic dimension of the environment and costuming the robot is central to facilitating an open engagement that enhances curiosity and trust that helps to promote communication, and the child’s willingness to interact.
The Close Encounter moves away from the mirrored encounters with the audience to embedding AI as a central feature to combine both controlled and improvised movements. The robot will learn from choreographed movements of professional dancers and audience interactions, to create a generative conversional model to enhance the playful, fun relationship between the participant and the robot. The work will explore Gordon Pask's conversational theory and poetic system to help enhance trust between the robot and participants and experiment with potential of robot audition and computational auditory scene analysis, soundtracking & localisation of sound in the installation.
Link to the artist's entry into the Usecase Library (pdf)
Video Link to content with transcription
For more related TAS resources and impact activities see the links below:
Responsible Research and Innovation
TAS Creative Impact
Other Related Publications at First International Symposium on Trustworthy Autonomous Systems (TAS ’23)
Anna-Maria Piskopani, Alan Chamberlain, Carolyn Ten-Holter. 2023. Responsible AI and the Arts: The Ethical and Legal Implications of AI in the Arts and Creative Industries. In First International Symposium on Trustworthy Autonomous Systems (TAS ’23), July 11–12, 2023, Edinburgh, United Kingdom. ACM, New York, NY, USA, 5 pages. https://doi.org/10.1145/3597512.3597528 - PDF
Eike Schneiders, Alan Chamberlain, Joel E. Fischer, Steve Benford, Simon Castle-Green, Victor Ngo, Ayse Kucukyilmaz, Pepita Barnard, Ju Row Farr, Matt Adams, Nick Tandavanitj, Kate Devlin, Clara Mancini, and Daniel S. Mills. 2023. TAS for Cats: An Artist-led Exploration of Trustworthy Autonomous Systems for Companion Animals. In First International Symposium on Trustworthy Autonomous Systems (TAS ’23), July 11–12, 2023, Edinburgh, United Kingdom. ACM, New York, NY, USA, 5 pages. https://dl.acm.org/doi/10.1145/3597512.3597517
Steve Benford, Adrian Hazzard, Craig Vear, Helena Webb, Alan Chamberlain, Chris Greenhalgh, Richard Ramchurn, and Joe Marshall. 2023. Five Provocations for a More Creative TAS. In First International Symposium on Trustworthy Autonomous Systems (TAS ’23), July 11–12, 2023, Edinburgh, United Kingdom. ACM, New York, NY, USA, 10 pages. https://dl.acm.org/doi/10.1145/3597512.3599709
Past Workshops - TAS related
Explainable AI for the Arts: XAIxArts - Nick Bryan-Kinns, Corey Ford, Alan Chamberlain, Steven David Benford, Helen Kennedy, Zijin Li, Wu Qiong, Gus G. Xia, and Jeba Rezwana. 2023. Explainable AI for the Arts: XAIxArts. In Proceedings of the 15th Conference on Creativity and Cognition (C&C '23). Association for Computing Machinery, New York, NY, USA, 1–7. https://doi.org/10.1145/3591196.3593517 - Open Access - PDF
Designing for Trust - Alan Chamberlain, Steve Benford, Joel Fischer, Pepita Barnard, Chris Greenhalgh, Ju Row Farr, Nick Tandavanitj, and Matt Adams. 2023. Designing for Trust: Autonomous Animal - Centric Robotic & AI Systems. In Proceedings of the Ninth International Conference on Animal-Computer Interaction (ACI '22). Association for Computing Machinery, New York, NY, USA, Article 26, 1–4. https://doi.org/10.1145/3565995.3566046 - PDF hosted by institution
Responsibility Reimagined: Responsible Innovation & Sustainability in the Digital Arts in an Age of AI - Chairs/Organisers: Alan Chamberlin and Dave De Roure, University of Nottingham, 15th May 2023. A joint event between TAS and PETRAS. Presenters' slides available on the webpage
Music and the Internet of Things: Physical Computing, Creativity, and Sensory Heritage - Chairs/Organisers: Alan Chamberlain, Dave De Roure and Emanuela Via. Oxford e-Research Centre, University of Oxford, 1st June 2023. A joint event between TAS and PETRAS. Presenters' Slides available on the webpage.
Acknowledgement
We would like to acknowledge the following TAS Hub Agile projects who all played a part in this: TAS Benchmarks Library and Critical Review Co-Design of Context-Aware Trustworthy Audio Capture, TAS RRI I. Without the insights from these other projects we would not have been able to articulate a response that covered the range of themes and topics that are covered in these conversations.