Harnessing Interdisciplinarity to Promote the Ethical Design of AI Systems
Bibliographic record
Abstract
There is a growing global awareness that increasingly powerful AI technologies are being developed which have the potential to reshape societies and institutions. ICT researchers and practitioners are under pressure to consider and reflect on the motivations, purposes and possible consequences of their innovations. Whilst it has long been recognised that technological innovations have social and ethical impacts, a gap remains in practice between ethics and social science research on the one hand, and computer science and engineering on the other. Few opportunities exist to incorporate ethical or social reflection into system development in order to design more responsible technologies. We argue that interdisciplinarity is fundamental to identifying pathways to best practice in the design and development of AI innovations - including their deployment in, and impact on, society. In this paper, we detail our experience of conducting an `ethical hackathon' as a tool for the facilitation of the ethical design of AI systems. This non-conventional hackathon model draws on Responsible Innovation (RI) and places primacy on the the integration of ethics by bringing together a range of disciplines as a necessary part of addressing a design task. In an ethical hackathon, computer scientists and engineers collaborate closely with specialists from other fields in order to learn how to work together effectively to design more responsible technologies. Teams which include computer scientists, engineers, ethicists, social scientists and business students, complete a task that requires them to anticipate and reflect on the social and ethical issues that may emerge from an innovation, and also consider how to address these in their technical designs. Through a qualitative analysis we highlight the significant potential of the model to facilitate the ethical design and development of AI systems. However, we also identify several barriers to the success of the approach and conclude that in order to conduct a successful ethical hackathon, and engender a truly interdisciplinary consideration of the ethics of AI, careful design and management of participants' expectations is required. To this end, we conclude the paper by providing design implications<br/>which build on our experiences.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.003 | 0.003 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".