Suchergebnisse

This paper argues that the hype around 'ethics' as panacea for remedying algorithmic discrimination is a smokescreen for carrying on with business as usual. First, it analyses how the current discourses around digital innovation and algorithmic technologies (including artificial intelligence or AI), newly emerging technology policy and governmental funding patterns as well as global industry developments are currently re-configured around 'ethical' considerations. Here, the paper shows how this phenomenon can be broken down into policy approaches and technological approaches. Second, it sets out to provide three pillars for a sociological framework that can help reconceptualize the algorithmic harm and discrimination as an issue of social inequality, rather than ethics. Here, it builds on works on data classification, human agency in design and intersectional inequality. To conclude, the paper suggests three pragmatic steps that should be taken in order to center social justice in technology policy and computer science education.

Social inequality in housing is a pressing issue that takes on a material dimension. To start tackling it, we need to rethink how we talk about housing and design, writes Mona Sloane. She explains how under current processes, attempts to improve the building stock often end up perpetuating wider social inequality

AbstractThe introduction of a new generation of AI systems has kicked off another wave of AI hype. Now that AI systems have added the ability to produce new content to their predictive capabilities, extreme excitement about their alleged capabilities and opportunities is matched only by long held fears about job loss and machine control.We typically understand the dynamics of AI hype to be something that happens to us, but in this commentary, we propose to flip the script. We suggest that AI hype is not a social fact, but a widely shared practice. We outline some negative implications of this practice and suggest how these can be mitigated, especially with regards to shifting ways of knowing and learning about AI, in the classroom and beyond. Even though pedagogical efforts (broadly understood) have benefited from AI hyping (there is now more varied AI training than ever), such efforts can also help minimize the impacts of hyping on the public's credulity toward extravagant claims made about AI's potential benefits and dangers.Below, we consider steps that can be taken to address this issue and illustrate pathways for more holistic AI educational approaches that participate to a lesser degree in the practice of AI hyping. We contend that designing better AI futures will require that AI hyping be blunted to enable grounded debates about the ways that AI systems impact people's lives both now and in the near future.

This paper provides a field scan of scholarly work on AI and hiring. It addresses the issue that there still is no comprehensive understanding of how technical, social science, and managerial scholarships around AI bias, recruiting, and inequality in the labor market intersect, particularly vis-à-vis the STEM field. It reports on a semi-systematic literature review and identifies three overlapping meta themes: productivity, gender, and AI bias. It critically discusses these themes and makes recommendations for future work.

What are the limits of design in addressing the political and/or when has design not been enough? A collection of thought pieces written by Theatrum Mundi's Designing Politics Working Group following a workshop at the Villa Vassilieff in Paris on 25th May 2016. This working group is supported by the Global Cities Chair at the Fondation Maison des Sciences de l'Homme in Paris.