Fairness aspects regarding the use of Artificial intelligence in asylum procedures and integration processes for refugees
Policy Community Workshop of the research project “Artificial Intelligence for Assessment” (AI FORA) as part of the “European Workshop on Algorithmic Fairness” (EWAF’24)
2.-3. July 2024, Johannes Gutenberg-University Mainz
Download Summary:
Summary German Policy Workshop
Policy Modeling meets Policy Practice
A workshop during the Annual Modeling and Simulation Conference (ANNSIM)
2024, May 21, American University, Washington D.C., USA
by Prof. Dr. Petra Ahrweiler, Asst.-Prof. Dr. Taylor Anderson, Prof. Dr. Erik W. Johnston, Prof. Dr. Thoams Clemen, Dr. Andreas Tolk
Participants representing policy practice:
Mr. Ryan A. Riccucci, Mr. Duane M Blackburn
Download Flyer:
Flyer policy workshop
Download Summary:
Summary Policy Modeling Workshop ANNSIM
https://media.springernature.com/full/springer-static/cover-hires/book/978-3-031-60401-0?as=webp
Petra Ahrweiler
Angels and Other Cows
A Celestial Adventure into AI Worlds, the Social Good, and Unknown Connections
Springer Nature Switzerland, Juni 2024
DOI: 10.1007/978-3-031-60401-0
https://link.springer.com/book/10.1007/978-3-031-60401-0
Public administrations are increasingly using Artificial Intelligence (AI) algorithms to decide on the provision of public social services such as unemployment benefits, pension entitlements, kindergarten places and social assistance to their citizens, hoping to achieve greater efficiency and objectivity.
Data profiles of citizens are analysed and assessed, and profiles automatically checked and scored to determine whether their owners are eligible to receive support from the state. However, AI-based social assessment systems, because they are based on machine learning from historical data, are accused of perpetuating bias and discrimination, often to the detriment of the most vulnerable groups in society.
Furthermore, who is considered as eligible, needy and deserving to be a beneficiary will always imply decisions that privilege certain groups while discriminating against others. Criteria vary widely around the world. There is no approach to social assessment that would be perceived as fair everywhere. Fairness concepts vary across national welfare systems depending on culture, religious tradition, and belief system.
Investigating the status quo and future options of AI implementations in this area, the goal of the international research project “Artificial Intelligence for Assessment” (AI FORA) is to create “Better AI” via participation of multiple societal stakeholders including vulnerable groups. The project employs a cultural comparison of Germany, Estonia, Spain, India, China, Nigeria, Ukraine, Iran, and USA to analyse how AI can contribute to more social justice.
The question of social assessment, i.e., who gets what from the state, concerns everybody, whether a policymaker hoping for efficiency and objectivity in allocation, a recipient hoping for support and wellbeing, a service provider, a taxpayer, or a member of a vulnerable group. Whether the introduction of AI into social assessment makes things better or worse is of interest to everyone and makes everyone a potential stakeholder in determining the design of social assessment innovations.
However, people are in very different positions to appreciate scientific achievements. How could the proverbial ‘people in the street’ become aware of AI FORA’s research? Research is often only available to experts, shaped by their interests, and framed by their language.
This acts as a threshold, particularly excluding the general public. Hence, science communication requires inclusive formats that are accessible to non-scientific audiences.
One suggestion how to do this was to break out of the silos of academia by switching media. “Angels and other Cows” (Ahrweiler 2024) - the AI FORA novel in literary fiction blending genres such as sci-fi, romance, adventure, mystery and comedy - took the task of inclusive science communication making available research topics, results and consequences of AI use in the public sector to a broad readership for attracting also non-scientists to academic research.
The fictional story of the open access novel introduces, reflects and discusses main epistemological, sociological and technical concepts of the research area. Hidden in a travel story around fictional cases in Spain, Estonia, Germany, India, China and USA, a plot deals with the deep questions of AI FORA:
The readers are invited for feedback. It is still open whether they want to see ‘the science behind’: The novel, which is also made available as graphic novel for visual experience on social media, precedes the two scientific publications of AI FORA and is published in the same series ‘Artificial Intelligence, Simulation and Society’.
The story of “Angels and Other Cows” unfolds both in heaven and on earth. The
real-world part narrates the adventure of Gabriel and Tilda, who work for an inter-
national aid company called B1. Their mission is to recruit case study partners for a
project on AI-based social assessment. Collaborating with partners from diverse
countries and societal backgrounds, they delve into the impact of values and culture
on social assessment, highlighting issues of bias and discrimination in different
systems that categorise humans for the purpose of receiving social welfare. The
novel terms this autoethnographic learning. Gabriel and Tilda grapple with defining
their own relationship, which is fraught with bias, mutual assessments, generational
power struggles and milieu-specific world-view clashes.
Their personal journey is intertwined with a global plea for unity in a world
marked by dystopian conditions, such as limited public resources, economic crises
and disparities in life opportunities across different countries. The world seems
divided into antagonistic societies with unique religious and sociocultural belief sys-
tems that perpetuate social injustice. The reader gains a comprehensive understand-
ing of these complex themes and embarks on a quest with Gabriel and Tilda to
explore whether AI can exacerbate or alleviate these issues. Readers also gain insights into AI and its workings, building a rudimentary understanding of machine learning for AI-based social assessment. Furthermore, they discover the potential benefits and risks, as well as the importance, of involving society in designing technology.
The celestial part of the story is dedicated to reflection. As Gabriel and Tilda
grapple with their intellectual and emotional connection, they are aided by two
guardian angels, GA and TA, tasked with saving the planet through interpersonal
and intercultural integration. These angels mirror their protégés’ struggles and pro-
vide comic relief, reminiscent of characters Statler and Waldorf from The Muppet
Show. The “voice of wisdom” is represented by English Benedictine monk Bede
Griffiths, who enthusiastically supports Tilda and Gabriel in their unlikely love story. On this level, the novel introduces, discusses and negotiates the philosophical and ethical questions.
„Angels and Other Cows” is the introductory novel of a series on AI FORA consisting of four open access books published by Springer Nature. Two academic books about scientific results will follow concluded by another novel which will evaluate the overall project.
AI FORA is funded by German Volkswagen Foundation under its programme „Artificial Intelligence and the Society of the Future“. Additional funding of 36.000 EUR was awarded to coordinating TISSS Lab for the concept of inclusive science communication.
The book „Angels and Other Cows. A Celestial Adventure into AI Worlds, the Social Good, and Unknown Connections” has been published in July 2024 and is available in open access.
Please note the Preface of the novel (section after Prologue / Contents table), which explains where the novel stands in the publication process of the project.
This is the official Springer link:
You can download the open access book for free from this link or order a printed book from the webpage. Of course, the book is also available in all bookshops, and on all platforms such as Amazon etc.
Video recording of the Herrenhäuser Forum of the Volkswagen Foundation "Überwachen – Übernehmen – Überlassen? Der Einsatz Künstlicher Intelligenz in der Pandemiebekämpfung" on the 25th of November 2021 with Prof. Dr. Petra Ahrweiler, Dr. Aljoscha Burchardt, Prof. Dr. Dr. Eric Hilgendorf, Prof. Dr. Judith Simon und Annette Riedel (moderation).
A lecture by Prof. Dr. Petra Ahrweiler on the topic "Artificial intelligence for assessment" in the "Cicle de Seminaris OEIAC: La intel·ligència artificial en la nostra vida quotidiana" organized by the University of Girona.
While AI is an intensively debated topic, the impact of AI on the social and cultural life is less investigated. This will be highlighted in the webinar. The webinar will consist of talks of three experts in the domain of AI and culture and will conclude with a plenary discussion of the speakers.
Talks:
Title: "Communicative Robots and (Post-)human Identities"
Abstract: With the rise of communicative robots like Amazon’s Alexa, language-based interaction with machines is increasingly becoming part of our everyday life. This presentation uses the case of communicative robots to investigate the changing relationship between humans and machines. I suggest that communicative robots challenge established boundaries of the social world but nevertheless reproduce a cultural asymmetry between human and non-human actors.
Person: My research profile connects an academic background in sociology with a transdisciplinary career trajectory in science & technology studies. One of my major research fields are the challenges of digital technologies for society and social theory.
Title: "AI in editing, marketing, and using books. What do we know about potential effects on taste and creation?"
Abstract: In the publishing industry, AI applications are not only used in general management and marketing, but increasingly also at the core of its value chain, in editorial work. After an overview about such applications, the book usage process (buying, reading) and its support and / or surveyllance by AI systems will be focused. Anecdotal evidence (corresponding platforms don´t talk much about what they do ...) and insights from other media systems (film, music) inform a research project that has just started and will research particularly algorithmic recommendations and their effects on cultural tastes and possibly also the creation side.
Person: Christoph Bläsi is a professor for book / publishing studies at JGU´s Gutenberg Institute for World Literature and Written Media. His main areas of interest are digital publishing and book business as well as the application of digital humanities methods to book studies research problems. More recently, he has been starting research on AI applications in the book industry as well as on the effects of recommender engines on cultural tastes.
Title: How does AI shape social change in the domain of public policy?
Abstract: the lab for sociology of technology and innovation of the JGU Mainz will provide an overview of ongoing research projects on the influence of AI on social change. This will be illustrated at two examples: One example is algorithmic assessment for social welfare provision. Research is undertaken in a bidirectional way: It is studied how AI assesses humans and how humans assess AI. A second example is cognition in navigating political landscapes. It is investigated whether AI based complexity reducing devices foster the rise of political populism.
Person: The lab of Professor Petra Ahrweiler is investigating the sociology of technology and innovation. Professor Ahrweiler received her PhD for a study on AI and obtained her habilitation for a study on simulation in Science and Technology Studies. Her lab investigates whether and how new technologies and innovations have the potential to redraw the image of our society in a completely new way. Dr. Frederick Herget will talk about algorithmic assessment for social welfare provision. Dr. Martin Neumann will concentrate on the impact of AI on navigating political landsca
New Book of the Unit published: "Ahrweiler, P., Neumann, M. (2021). Advances in Social Simulation. Proceedings of the 15th Social Simulation Conference: 23–27 September 2019.
As part of the BMBF Cluster4Future program, the TISSS Lab at Johannes Gutenberg University is participating in the SENSORITHM Rhein-Main future cluster together with the partners Johann Wolfgang Goethe University Frankfurt am Main, Darmstadt University of Technology, Fraunhofer Institute for Structural Durability and System Reliability LBF, and the Institute for Animal Ecology and Nature Education. The interdisciplinary consortium combines expertise from physics, biology, computer science, mechanical engineering and social sciences. SENSORITHM will investigate how intelligent sensor technologies help avoid collisions of birds and bats with wind turbines and will develop self-learning sensor systems for monitoring technical components and installations.
In this way, Sensorithm can help resolve a green-green dilemma: On the one hand, the renewable energy generated by wind turbines is intended to halt climate change and thus ultimately safeguard biodiversity; on the other hand, rotor blades endanger rare bird species such as the red kite and various bat species. By introducing smart sensor technologies, the trade-off of climate neutrality vs. energy demand could be resolved with the help of innovation networks.
In the project, TISSS Lab director Prof. Dr. Ahrweiler and her team will be responsible for the social science analysis and design of the multi-dimensionality of this innovation network in the field of tension between technological, ecological, economic, political and social aspects.
For example, it is expected that approval procedures for wind turbines will change and that knowledge about new technological possibilities will find its way into political discourses. The Future Cluster is closely linked to partners from industry, universities, institutes, authorities and civil society (environmentalists, associations, NGOs, school classes) and will cooperate with them. The TISSS Lab will support participatory processes and ensure adequate involvement of relevant stakeholders.
There is a new publication in Quality & Quantity: "Voinea, C., Neumann, M. (2020). Political culture. A theory in search for methodology. An editorial. Quality & Quantity 54(2): 335-360."
An article has been published in Evidence based HRM: "Siebers, Ü., Herath, D., Bardone, E., Farahbakhs, S., Knudes, S., Madsen, J. Mufti, M., Neumann, M., Richards, D., Seri, R., Secchi, D. (2020). On the quest for defining organizational plasticity. Evidence based HRM: a global forum for empirical scholarship (online first)"
