Finola Finn

(4/4)
- Iqra Aslam, Rahul Nagshi, and Donal Khosrowi’s critical examination of Machine Unlearning

- Bonam Mingole et al.’s work on the value of democratising bias detection and critique of LLMs

A lot to take away! All full papers available here: www.aies-conference.com/2025/proceed...

(3/4)
- Maria Eriksson et al.’s thorough and policy-focused review of current issues in AI benchmarking

- Hyo Jin Do and Werner Geyer’s recommendations for how best to present factuality estimates to users in AI-generated answers

(2/4)
Among the many thought-provoking contributions, I was particularly struck by:

- Shreya Chappidi et al.’s “accountability capture”, a framework for understanding the under-considered effects of implementing algorithmic record-keeping on sociotechnical systems

The AAAI/ACM Conference on AI, Ethics, and Society (AIES25) was an enriching few days, providing a great forum for engaging with the latest research on the ethical challenges of AI. (1/4)

The whiteboards were full with opinions and experiences by the session’s end. Many thanks to the participants and my co-panellists Stefan Esselborn, @cgoetter.bsky.social, and @nescioquid.bsky.social

We explored such questions as: What new teaching and research methods are becoming possible? What skills should we teach students in the age of AI? What can History of Technology contribute to the discussion?

Very much enjoyed the lively discussions at our interactive roundtable ‘Artificial Intelligence in Teaching History of Technology’ at the SHOT 2025 Annual Meeting!

Un livre écrit par deux historien.nes, @ccarolinemuller.bsky.social et @artificialmemory.bsky.social , qui parlent d'une manière intelligente du numérique parce qu'ils le pratiquent depuis longtemps et qu'ils l'approchent comme un 'bête' sujet de sciences humaines
#paniquesmoralessabstenir

The submission deadline for #CHR2025 is fast approaching! Authors, you have until Friday, 18 July, to submit your short and long papers, lightning talk abstracts, and workshop ideas.

⚠️ Please note: there will be NO extension of this deadline.

🔗 2025.computational-humanities-research.org/cfp/

Promotional image. Text reads: The people's dispensary digitised archive

Check out #ThePeoplesDispensary – our free new online resource about charitable medicine in Edinburgh. It includes a fully transcribed and digitised archive of over 10,000 pages of Georgian patient case notes from the Edinburgh Public Dispensary

👉 www.rcpe.ac.uk/peoplesdispe...

Alongside the fairground rides, an additional highlight was presenting my research on uses of the soul-body relationship in 17thC accounts of melancholy in this gorgeous pub! A delightful juxtaposition. Many thanks to everyone who squeezed into the lounge room to hear our panel

What a brilliant conference! Feeling very lucky to have heard fascinating papers and met so many new faces within the atmospheric surrounds of the Black Country Living Museum! Inspired more than ever by the breadth and depth of social histories #SHSConf2025 @socialhistsoc.bsky.social

Such a great evening!

Highlights for me included @myra.bsky.social et al’s work on measuring and understanding public perceptions of AI through crowdsourced metaphors, and Prakhar Ganesh et al’s tutorial on the implications of multiplicity for fairness and explainability in algorithmic decision-making (plus much more!)

So glad to be at #FAccT25! Fantastic first day soaking up the latest cross-disciplinary research on fairness, accountability, and transparency in socio-technical systems

Public history for contested and conflicting past(s) International Public History online seminar. For student - by students.

This Thursday, seminar online organised by the students of the Master in Digital and Public History at @uni.lu with @c2dh.uni.lu
Presentations by students from all around the world
www.uni.lu/c2dh-en/even... #PublicHistory #student #history

Silicon Islands and War Exhibition In the framework of the European Month of Photography (EMoP), this exhibition discusses how in an era of unprecedented AI development, the nature of

The video itself also seeks to challenge “the vastness of digital memory and its overwhelming flood of images” through the use of the oral traditions of poetry and song. More about these thought-provoking works here: www.uni.lu/life-en/even...

… reminding us that “the production chain of images is becoming an increasingly complex web of resource supply, energy consumption, and cutting-edge chip manufacturing” - shaped and driven by geopolitics, especially between China, the U.S, Japan, South Korea and Taiwan.

Such an engaging guided tour by the artist Daphné Le Sergent this afternoon of her exhibition ‘Silicon Islands and War’, hosted by @uni.lu as part of EMoP. Her video essay “invites us to explore an alternative history of photography in relation to the semiconductor industry” ->

Engaging the many-hands problem of generative-AI outputs: a framework for attributing credit - AI and Ethics The recent wave of generative AI (GenAI) systems like Stable Diffusion or ChatGPT that can produce images, text and code from human prompts raises controversial issues about creatorship, originality, ...

Our full article on the CCC framework was published in AI & Ethics last year, and is available here:
doi.org/10.1007/s436...

We were excited to see many other teams also working to address the moral, legal and regulatory uncertainties that surround GenAI use – and it was fantastic to discuss these issues at such an interdisciplinary and vibrant event!

Our presentation outlined the Collective-Centered Creation framework (developed by Donal Khosrowi, Elinor Bell-Clark, and me), which provides an approach for tracking and assessing the significance of different contributions, by humans and machines, to the creation of GenAI images.

It was great to share our work on attributing credit and responsibility for generative AI outputs at the ‘Ethics and Aesthetics of Artificial Images’ conference at Venice International University last month ->

Less than a week till our @cemskcl.bsky.social colloquium on early modern war narratives! Sign up here to attend in person or online! 👉 www.tickettailor.com/events/centr...

Please repost! ⚔️

The conference was full of thought-provoking and motivating reflections on what British history can do for the present - and I’m really glad to have been a part of it!

I argued that, in the media and PR, such parallels are often used to reductively normalise the adoption of AI by suggesting historical precedent – but that historians can (and should) contribute more nuanced, insightful comparisons to the discussions.

It was such a pleasure to be part of the British History Today conference at Queen Mary University London earlier this month, presenting my work on the use of historical analogies in public discourse on AI.

Image generated by dall-e-3. Prompt: “A water colour drawing of language and historical bias in large language models (LLMs)”.

New blog post! In their article “LLM Biases: Expected and Unexpected Model Design Effects in Historical Newspaper Article Extraction on the Messina Earthquake”, Johanna Mauermann and Sarah Oberbichler (@SarahOberbichler) look at the fact that LLMs are a powerful […]

[Original post on fedihum.org]

Do you work on #earlymodern English women and religion (c.1500-1800)? Would you like to join a community of scholars asking questions, sharing research, and disseminating news of upcoming events related to the field? Please subscribe to our new mailing list: www.jiscmail.ac.uk/EMEW-R 🗃️

Book cover reveal!!! So excited to see this 🌟🎉@reaktionbooks.bsky.social