Weizenbaum Library (Weizenbaum Institute)
Not a member yet
864 research outputs found
Sort by
Measures to reduce corporate GHG emissions: A review-based taxonomy and survey-based cluster analysis of their application and perceived effectiveness
Get PDFCompanies contribute to a large extent to greenhouse gas emission. To mitigate this, measures for reducing these emissions can be applied. There is, however, neither a systematized general overview of existing measures nor an estimation of their application and their effectiveness to reduce greenhouse gas emissions. This study strives to close this gap by reviewing research on the reduction of corporate greenhouse gas emissions and synthesizing emission reduction measures in a taxonomy. Furthermore, the application of these measures and their perceived effectiveness is empirically assessed using a survey among companies that are involved in emission reduction activities. On this basis, a cluster analysis is conducted to identify measure types and to unveil application patterns. 27 different measures and 65 respective implementation examples are identified and structured within nine categories: energy, product, process, technology, 6R and waste management, office and mobility, management, reporting and disclosure, and compensation measures. The empirical analysis shows that there exist measures with a high efficiency to reduce emission, which are rarely applied in companies. On the other side, a large share of applied measures is not perceived as highly effective. Companies can use these results to structure their emission reduction activities and identify best practices
Why Does the AI Say That I Am Too Far Away from the Job Market?
Get PDFAs artificial intelligence (AI) is increasingly being deployed in various domains such as healthcare (Qayyum et al., 2021), finance (Dastile, Celik & Potsane, 2020) and public welfare (Saxena et al., 2020; Carney, 2020), there is a growing need for understanding how stakeholders are affected by AI (Vaassen, 2022) and how to design and present explanations of AI-based decisions in ways that humans can understand and use (Miller, 2019). This paper contributes to these efforts by examining an AI-based decision-support system (DSS) launched by the Swedish Public Employment Service (PES) in 2020. Specifically, the study investigates to what extent the studied system enables affected jobseekers to understand the basis of AI-assisted decisions, to negotiate or contest dispreferred decisions, and to use the AI as a tool for increasing their job chances.This work has been funded by the Federal Ministry of Education and Research of Germany (BMBF) (grant no.: 16DII121, 16DII122, 16DII123, 16DII124, 16DII125, 16DII126, 16DII127, 16DII128 – “Deutsches Internet-Institut”)
Digitally Aided Sovereignty: A Suitable Guide for the E-Government Transformation?
Get PDFWe advocate for the adoption of an integrated strategy aimed at achieving increased participation via effective digital public administration services. We argue that it is urgent to understand the integration of participatory approaches from the field of e-democracy in digitalized public administration, as trendsetting e-government implementations are already underway. We base our arguments on the observation that the approaches in e-democracy and e-government seem to be locked into extremes: In e-democracy, (experimental) platforms have failed to create a participative political culture. E-government, in turn, narrowly perceives citizens as customers. Additionally, efforts to increase digital sovereignty have mostly been educational ones that support citizens' self-determined use of the digital but do not address sovereignty via the digital. As a result, digitalized public administration is not achieving its potential to create opportunities for participation during encounters with the administration. Hence, we argue for the adoption of a digitally aided sovereignty as a normative guide for an e-government transformation that strives to create opportunities for participation via the digital.This work has been funded by the Federal Ministry of Education and Research of Germany (BMBF) (grant no.: 16DII111, 16DII112, 16DII113, 16DII114, 16DII115, 16DII116, 16DII117 – „Deutsches Internet-Institut“
New methodologies for the digital age? How methods (re-)organize research using social media data
Get PDFAs “big and broad” social media data continues to expand and become a more prevalent source for research, much remains to be understood about its epistemological and methodological implications. Drawing on an original data set of 12,732 research articles using social media data, we employ a novel dictionary-based approach to map the use of methods. Specifically, our approach draws on a combination of manual coding and embedding-enhanced query expansion. We cluster journals in groups of densely connected research communities to investigate how heterogeneous these groups are in terms of the methods used. First, our results indicate that research in this domain is largely organized by methods. Some communities tend to have a monomethod culture, and others combine methods in novel ways. Comparing practices across communities, we observe that computational methods have penetrated many research areas but not the research space surrounding ethnography. Second, we identify two core axes of variation—social sciences vs. computer science and methodological individualism vs. relationalism—that organize the domain as a whole, suggesting new methodological divisions and debates.The Lab has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement no. 759681). The paper has been written with the support of a Fellowship grant at the Weizenbaum Institute, Berlin
The Problems of the Automation Bias in the Public Sector: A Legal Perspective
Get PDFThe automation bias describes the phenomenon, proven in behavioural psychology, that people place excessive trust in the decision suggestions of machines. The law currently sees a dichotomy—and covers only fully automated decisions, and not those involving human decision makers at any stage of the process. However, the widespread use of such systems, for example to inform decisions in education or benefits administration, creates a leverage effect and increases the number of people affected. Particularly in environments where people routinely have to make a large number of similar decisions, the risk of automation bias increases. As an example, automated decisions providing suggestions for job placements illustrate the particular challenges of decision support systems in the public sector. So far, the risks have not been sufficiently addressed in egislation, as the analysis of the GDPR and the draft Artificial Intelligence Act show. I argue for the need for regulation and present initial approaches.This work has been funded by the Federal Ministry of Education and Research of Germany (BMBF) (grant no.: 16DII121, 16DII122, 16DII123, 16DII124, 16DII125, 16DII126, 16DII127, 16DII128 – “Deutsches Internet-Institut”)
Theorizing Digital Capitalism
Get PDFThere is little disagreement that digital technologies are transforming contemporary economies and societies. However, scholars have only begun to systematically think about how digitalization – the process whereby more and more of what we say, think, and do becomes mediated by digital technologies – is both driven by and transformative of capitalism. This paper argues that when one speaks about digitalization, one cannot be silent about capitalism. It reconstructs commodification and disruption as key features of capitalist development. It then shows how three digital revolutions – the platform, (big) data, and artificial intelligence revolutions – have ushered in a new wave of commodification and disruption, giving rise to digital capitalism. Finally, it discusses the challenges commodification and disruption pose in the form of redistribution of resources, rebalancing of power, rule adaption, and market re-embedding. The paper brings together a wide range of scholarship to offer a historically and theoretically grounded framework for how to think about and study the rise of digital capitalism
Peace Journalism in the Digital Age: Exploring Opportunities, Impact, and Challenges
Get PDFThe advent of modern means of communication opens up a wide range of possibilities for individual users, organizations, and governments to connect. This paper argues that the concept of peace journalism can leverage the potential of digital developments to maintain relevance in current times. Five areas of peace journalism’s possible synchronization with media digitalization are deduced and elaborated from a pragmatic perspective to facilitate conceptual advancement: (1) digital distribution, (2) utility of the potential of two-way communication, (3) exploration of new forms of digital storytelling, (4) curation of various digital sources of conflict actors and fact-checking, and (5) incorporation of virtual training and digital skills into journalism curricula. By addressing these aspects of media digitalization, peace journalism outlets can receive acclaim within modern journalistic circles while also attracting wider audience support.The Weizenbaum Institute is funded by the German Federal Ministry of Education and Research (BMBF
Algorithmen der Alterität - Alterität der Algorithmen. Überlegungen zu einem komplexen Verhältnis
Get PDFHow Should We Regulate AI?
Get PDFIn the last decade, artificial intelligence (AI) – which describes the mimicking of human intelligence using technology – has made significant progress. Driven by algorithmic design, computing power and large amounts of training data, machine learning has transformed information technology, which can now augment and replace human intelligence, something that was thought impossible just a decade ago. In 2018, the European Commission labelled AI a transformative technology with the potential to raise new ethical and legal questions. Now, with the advent of generative AI, which can create content that could previously only be created by human beings, this potential has become visible to the wider public. At the same time, the European Commission’s proposal for an Artificial Intelligence Act (AIA) (which is now entering the final legislative stage) indicates its intentions to regulate AI. This comment wishes to highlight some key points regarding the regulation of artificial intelligence and, in doing so, comment on the current proposal.This publication has been funded by the Federal Ministry of Education and Research of Germany (BMBF) (grant no.: 16DII121, 16DII122, 16DII123, 16DII124, 16DII125, 16DII126, 16DII127, 16DII128 – “Deutsches Internet-Institut”)
How Far Can It Go? On Intrinsic Gender Bias Mitigation for Text Classification
Get PDFTo mitigate gender bias in contextualized language models, different intrinsic mitigation strategies have been proposed, alongside many bias metrics. Considering that the end use of these language models is for downstream tasks like text classification, it is important to understand how these intrinsic bias mitigation strategies actually translate to fairness in downstream tasks and the extent of this. In this work, we design a probe to investigate the effects that some of the major intrinsic gender bias mitigation strategies have on downstream text classification tasks. We discover that instead of resolving gender bias, intrinsic mitigation techniques and metrics are able to hide it in such a way that significant gender information is retained in the embeddings. Furthermore, we show that each mitigation technique is able to hide the bias from some of the intrinsic bias measures but not all, and each intrinsic bias measure can be fooled by some mitigation techniques, but not all. We confirm experimentally, that none of the intrinsic mitigation techniques used without any other fairness intervention is able to consistently impact extrinsic bias. We recommend that intrinsic bias mitigation techniques should be combined with other fairness interventions for downstream tasks