We are pleased to share this report on the activities and impact of the Centre for Technomoral Futures in its first two years. Under the leadership of our Director, Shannon Vallor, the Centre has achieved an incredible amount in just a short time, and this represents our first effort to share that work publicly in a summarized form.

Letter from the Director

Professor Shannon Vallor, Baillie Gifford Chair in the Ethics of Data and AI (and Director of the Centre for Technomoral Futures)

Successful recruitment of a diverse and exceptionally talented group of PhD students taking on timely and novel interdisciplinary research projects in AI and data ethics;
The creation of a core staff team with both the administrative and academic excellence to help deliver on the Centre’s promise;
The creation of innovative EFI undergraduate and postgraduate curricula in AI and data ethics, as well as courses for philosophy and executive education programmes;
Securing more than £4 million of competitive external grant funding for impactful applied interdisciplinary research in AI and data ethics;
Building strong connections in the Edinburgh city-region, Scotland, the UK and Europe to foster the Centre’s future collaborations and partnerships;
Establishing the Centre’s visibility as both a regional and international pioneer in interdisciplinary AI and data ethics research, education and public engagement, positioning the Centre to become a world-leading attractor of exceptional students and researchers who will define the future of this urgently needed area of expertise.

While our public engagement plans for lecture series, workshops and other gatherings have been constrained by COVID-19 restrictions in some ways, they have been enabled in others, allowing large audiences to join us economically via online platforms. For example, in 2021, the Centre co-organised the Realising a Compassionate Planet event during COP26, registering 600+ guests from 25+ countries. In a year-long collaboration with Edinburgh’s Global Compassion Initiative and Stanford University, we co-hosted an unprecedented global conversation about the future of the planet.

In staff recruitment, our prestigious Chancellor’s Fellow post resulted in the selection of Dr. Atoosa Kasirzadeh, one of the most talented junior scholars in the world who holds dual PhDs in Philosophy (University of Toronto) and Applied Mathematics/Machine Learning (Ecole Polytechnique Montreal), and has already completed prestigious research posts in AI Ethics at Australian National University and DeepMind. Furthermore, the Centre recruited a new Programme Manager, Dr Gina Helfrich, who took up the post in July 2022 and leads on operational management, strategy and governance infrastructure.

Our PhD and postdoctoral research cohort, now spanning twelve different areas of disciplinary study related to AI and data ethics, is bridging the harmful social divide between technical and moral knowledge. Our students are building the shared vocabulary and interdisciplinary research tools that this new field urgently needs for the future, and from which societies and institutions globally will benefit. This was the promise of the Baillie Gifford gift that brought me to Edinburgh, and I am profoundly grateful to see it becoming a reality.”

Funding in action

Technomoral Conversations event series

When Artificial Intelligence Speaks: A Conversation on the Ethics of Large Language Models (7th Dec 2021)

A digital screen capture Dr Atoosa Kasirzadeh and Dr Iason Gabriel in conversation on Microsoft Teams. Atoosa's image is above, and Iason's below. Both are smiling. — *When Artificial Intelligence Speaks* event. Pictured: Dr Atoosa Kasirzadeh (Chancellor’s Fellow) and Iason Gabriel (DeepMind).

On 7th December 2021, the Centre launched a new public event series: Technomoral Conversations. When Artificial Intelligence Speaks: A Conversation on the Ethics of Large Language Models was the inaugural event, which attracted over 200 registrations and sold out all available tickets. This first event featured DeepMind's Iason Gabriel in conversation with our Chancellor's Fellow, Dr Atoosa Kasirzadeh, about the ethical issues emerging with large language AI models such as GPT-3. The audience included viewers from not only the United Kingdom and the United States, but also Poland, Portugal, Ireland, Mexico, Romania, Jamaica, France, Canada, Denmark, Bahrain, Spain, Chile, Belgium, Taiwan, Netherlands, Austria and Uruguay.

Building Responsible AI Cultures and Alliances (29th Mar 2022)

Speakers Albert King and Virginia Dignum, they are sat next to each other and Virginia is speaking. — Speakers at the *Building Responsible AI Cultures and Alliances* event; Albert King and Virginia Dignum

A conversation between two leading voices in European and Scottish AI strategy and policy: Virginia Dignum (Umea University), author and advisor on EU AI policy and governance, and Albert King, Chief Data Officer in the Scottish Government. The conversation was framed as follows: “The European Union and the Scottish Government have each expressed commitments to fostering responsible and ethical AI development. What opportunities and obstacles lie ahead on that path? What new alliances and governance cultures are needed to bring the ideal of ‘responsible AI’ into existence?”

This hybrid event was free and open for all, with an intimate in-person audience, the speakers present in the room, and the conversation streamed live to Zoom participants.

Sustainability and AI, 21st November 2022

Speakers at our Technomoral Conversations: Sustainability and AI event. The speakers are sitting on chairs in a semi-circle formation with a projection of the event poster behind them, facing the audience. — *Sustainability and AI* event. L-R: Director Shannon Vallor, Dr Mirella Lapata, Dr Ramit Debnath, Dr Ronita Bardhan, Jonathan Bean.

The Centre’s most recent event was a panel discussion on sustainability and AI, featuring four experts in the field who are tackling different facets of this challenge, chaired by our Director Shannon Vallor.

The conversation wasf ramed as follows: “What is the impact of AI on the planet? There are reasons to hope that data-driven efficiencies and insights can help produce innovative solutions to the climate emergency, touching areas such as energy, biodiversity monitoring and conservation, transportation, water conservation, and agriculture, among others. On the other hand, there are also real concerns about the climate cost of producing AI models in the first place. Researchers have found that training a single AI model can emit as much carbon as five cars in their lifetimes. Will AI turn out to be a net negative or a net positive for the sustainability of life on the planet?” This hybrid event was free and open for all.

This event was followed by a lecture the following day by two of the panel members, Dr Ronita Bardhan and Dr Ramit Debnath of Cambridge University, who discussed their own research on sustainable engineering and policy.

Other events

Realising a Compassionate Planet, 5th November 2021

Realising a Compassionate Planet event poster

This event was co-organised with Stanford University and the Edinburgh Global Compassion Initiative, and directed by Liz Grant. The Centre for Technomoral Futures had a dedicated exhibitor’s online booth, with Director Shannon Vallor moderating a session on Human values, inhuman futures: Tensions and paradoxes in realising a compassionate planet with technology. The event attracted a notable social media presence and sparked meaningful conversations and collaborations.

The Limits of Fairness by Abeba Birhane, 13th Jun 2022

Speaker Abeba Birhane stands in front of a projected powerpoint presentation, there is a quote that is only partly visible. — Keynote speaker Abeba Birhane

Our annual flagship lecture this year was held in June, with Abeba Birhane (University College Dublin) as the keynote speaker. The lecture was organised in collaboration with Edinburgh’s student-led AI Ethics & Society group, as well as UNICEF’s Data for Children Collaborative. The event drew a full-capacity crowd, who enjoyed the talk and the vibrant Q&A session and reception that followed.

The following day, Birhane joined us for a day-long hybrid workshop on relational AI ethics with DeepMind Research Scientist Shakir Mohamed, Oxford researcher Dr Marie Therese-Png, and remote participants Professor Aimee van Wynsberghe (Director of Bonn’s Sustainable AI Institute), Dr Aristea Fotopolou (Reader in Digital Communication, Culture and Society at Brighton), Dr Stephanie Russo Carroll (leader of University of Arizona’s Collaborative on Indigenous Data Governance) and Carolina Villegas-Galaviz (Notre Dame’s Center for Technology Ethics), as well as a range of Edinburgh scholars working on related topics.

Curricular Innovation

The Ethics of Politics and Data (2021-22)

Edinburgh Castle in front of a pink sky.

In 2021-22, the Centre developed and delivered in-person a new undergraduate course under EFI. The course drew 40 first year University of Edinburgh students from a range of disciplinary backgrounds who learned critical data skills from Professor Vallor, Dr Kasirzadeh and guest lecturers in data visualisation and the history of data science and mathematics. The innovative course design included tutorials that allowed students to work hands-on with data artefacts and, as confirmed in the feedback from students, helps to “consolidate the content discussed in the lectures” and “allows us to see how the theoretical side comes into practice.” In the words of a student, “It uncovers a lot of things I hadn’t really thought of before which I think are really important to know for anyone going into computer science or any data-based field.” Another student stated that they valued “the development of critical thinking skills and building a useful mindset to observe and question data technologies,” another report from the course feedback states “this course should be mandatory for all students.”

MSc in Data & AI Ethics (starting 2023-24)

Opening soon for applications for both online and in-person study, this programme meets the urgent demand for interdisciplinary skills and knowledge in the ethical design, use and governance of artificial intelligence and other data-intensive technologies.

This interdisciplinary degree is designed with the University’s world-leading academic expertise in the area of Data & AI Ethics, as well as philosophy, law, informatics and science and technology innovation studies (STIS). It leverages the research power of the Centre for Technomoral Futures in promoting sustainable, just and ethical outcomes for AI and data-driven technology.

Working with the Scottish and UK Governments

Since the Centre’s inception, Director Shannon Vallor has…

The Scottish Parliament is in the foreground, with Calton Hill in the background. The sky is clear and blue. — Scottish Parliament

Been chair of the Scottish Government’s Data Delivery Group since March 2020; the Data Delivery Group reviews the progress of various streams of work under the Scottish Government’s data vision as part of its digital strategy.

Led the drafting (with the University’s AI and Ethics Advisory Board) of input to the UoE/DDI (Data-Driven Innovation) response to the UK National Data Strategy. Our input was a central piece of the University and DDI’s response and identified key ethical opportunities and risks associated with the government’s data strategy. We also offered strategic advice on how the government can manage those risks through broader investment in ethical data standards to secure the strategy’s promised benefits.

Interior of the Bayes Centre. An open plan space with several floors, and a suspended staircase which criss-crosses between floors. — Bayes Centre, University of Edinburgh

Given evidence to UK All-Party Parliamentary Group on ethics of AI applications for COVID-19 in April 2020. Evidence given focused on the importance of transparency, humility, honest discussion of risks, and open, reciprocal engagement with diverse citizen communities to ensure and sustain fragile public trust in both AI and public health interventions.

Formed research partnerships with the UK’s NHS AI Lab and Scottish Government’s Digital Directorate as part of a UKRI-funded research programme on responsibility in trustworthy autonomous systems.
Given evidence to Scottish Parliament in June 2022 on AI and Accountability in the Public Sector as part of Scotland’s Futures Forum.
Been advisor on ethics to Scotland’s QCovid Project in 2021-22.

Professor Vallor also served on Scotland’s Digital Ethics Expert Group, the advisory boards of Research Data Scotland, Scotland’s QCovid Project and the Scottish Biometrics Commissioner, and the UK National Statistician’s Inclusive Data Advisory Committee.

Research highlights

A stock image of a person typing on a laptop. The image is overlayed with geometric lines and shapes.

Led by Professor Vallor as principal investigator, a University of Edinburgh team was awarded £673,000 by UKRI EPSRC for research on responsibility in Trustworthy Autonomous Systems. The proposed 30-month project, Making Systems Answer: Dialogical Design as a Bridge for Responsibility Gaps in Trustworthy Autonomous Systems, includes co-investigators in Philosophy, Informatics, and Law/Medicine. The project launched in January 2022. External research partners included SAS, the NHS AI Lab (NHSX), and the Scottish government’s Digital Directorate.
The Centre recruited a Research Associate, co-funded with UNICEF’s Data for Children Collaborative, to develop standards for ethics assessment in data-driven humanitarian research. The researcher, SJ Bennett, began her 18-month term on 1 October 2021.
CTMF supported the UKRI EPSRC grant awarded in November 2020 to Informatics (£3.2 million) for Governance of Trustworthy Autonomous Systems; Baillie Gifford PhD student Bhargavi Ganesh, co-supervised by grant co-investigator Prof Vallor, is linked to this project and started her research on September 2021; scholarship funding was included as an in-kind contribution to the project from EFI.
With ECA/EFI’s Professor Ewa Luger, co-Director of the Institute for Design Informatics, in November 2022 the Centre’s Director Prof Vallor won a £3.5 million bid to co-direct the UKRI AHRC ‘Building a Responsible AI Ecosystem’ programme.

Education highlights

A busy open plan office with people dressed casually and working at communal desks.

In 2022, the Centre delivered the new undergraduate EFI course, Ethics and Politics of Data. Director Vallor further taught a postgraduate course in Philosophy, Ethics of Artificial Intelligence.
Approval in spring 2021 of a new core PG course required for all students in EFI Masters programmes, Ethical Data Futures, planned for first delivery in academic year 2022-2023.
In November 2022, a new MSc in Data and AI Ethics was approved by the EFI Curriculum Board for launch in 2023.
Plans to revise and bring EFI’s highly successful Data Ethics, AI and Responsible Innovation MOOC into the Centre.
Successful recruitment of new Baillie Gifford Chancellor’s Fellow Dr. Atoosa Kasirzadeh, who has recently held research appointments at Australian National University and DeepMind, and holds dual PhD degrees in Philosophy (University of Toronto) and Applied Mathematics (Ecole Polytechnique de Montreal). Dr. Kasirzadeh joined the Centre for Technomoral Futures in December 2021.

Plans for the coming year

A key priority for the Centre in the coming year is expanding its public engagement programme and research visibility through a number of paths:

Approx. 11 people sit round a conference-style table. They are all dressed smartly, and most have a glass of water and/or a laptop and notes in front of them

Workshops eliciting New Perspectives on AI Futures will be held in early 2023 with support from the Alan Turing Institute, and the outputs will feed into an event at the Scottish AI Summit in March 2023.
In Spring 2023, we will host and contribute to several public events as part of the Edinburgh Futures Institute’s events series Love Machines.
In September 2023, the Centre will host its 2023 annual flagship lecture featuring philosopher C. Thi Nguyen (winner of the 2021 APA Book Prize for Games: Agency as Art) as a hybrid in-person and online event.
In late 2023, we will host the Centre’s inaugural lecture from Director Shannon Vallor in the new EFI building.

Artist's rendering of the new EFI building. The building was the old Edinburgh Royal Infirmary, and is built in the baronial style. In the centre of the image is the building's spire/clocktower, and there are trees in the courtyard. — Artwork of the new EFI building, Quartermile, Edinburgh

A second priority for the Centre will be rolling out the new interdisciplinary MSc in Data & AI Ethics, which received approval from the EFI Curriculum Board in November 2022. The programme, which will be launched in 2023, is directed by Prof Shannon Vallor, and co-developed by Drs. Atoosa Kasirzadh, John Zerilli, Morgan Currie, Rob Smith and James Garforth. It meets the urgent demand for interdisciplinary skills and knowledge in the ethical design, use and governance of artificial intelligence and other data-intensive technologies.

The Centre will continue to strengthen its highly successful mentorship programme for the Baillie Gifford PhD cohort and Centre-associated postdoctoral researchers, in anticipation of the 2023 opening of the new EFI building that will house the Centre.

The Centre will also drive the impact of ongoing research in two UKRI EPSRC grants on Trustworthy Autonomous Systems for which Professor Vallor is an investigator (on Responsibility as principal investigator, and on Governance and Regulation as a co-investigator). The Centre’s new MSc in Data and AI Ethics will begin teaching in the 2023-24 academic year.

Finally, from 2023-2025, Professor Vallor will co-direct with Prof Ewa Luger (ECA) a £3.5 million AHRC-funded programme, ‘Enabling a Responsible AI Ecosystem,’ in collaboration with the Ada Lovelace Institute.