Responsible ethical learning with robotics (in brief  REELER) is notoriously difficult because there is a gap between the lifeworld of roboticists (robot designers, engineers, software developers) and the lifeworld of stakeholders affected by the robots (e.g. health care workers, shop floor workers, farming assistants, school teachers). This may lead to a number of unintended consequences and unexpected reactions that come as a, sometimes, unpleasant surprise to both robot developers and policy makers. The ‘industry 4.0’ may not only be a welcomed and relevant disruption, but also a driver of robo-sabotage, decreased efficiency and new forms of complexity incomprehensible to both robot designers, policy makers and robotic artificial intelligence pervading future work life surroundings.

The highly interdisciplinary H2020-project REELER aims a bridging this gap with a host of new methods including mini-publics. In this talk, I will present a version of mini-publics (inspired by the juridical versions) that can be used by roboticists and policy makers to get new insight into how affected stakeholders may feel about collaborating with robots in the future. Based on our experiences with developing mini-publics as a new tool for research, policy-making and robot design, our work-in-progress research points at mini-publics as a new methodology in a distributed responsibility for ethical and societal issues relating to robotics. 

Mankind is on the brink of the fourth industrial revolution, in which robotics and artificial intelligence are expected to have disruptive effects on types of tasks performed by humans, the level of employment, real wages, and the income distribution.

At present, the popular media and public debate are captivated by fearmongering headlines, and also scientific work paints a gloomy picture of the future with structurally high levels of unemployment, stagnating median wages, and growing income inequality.

In this presentation, I provide a short overview of the ongoing discourse on the impact of diffusion and adoption of robotics and artificial intelligence on (i) shifts in tasks performed, skills required, and jobs done, (ii) employment dynamics and the role of education therein, and (iii) real wages, income distribution, and inequality. In addressing these three topics, I pit the 'sector of application' perspective generally followed in literature against a structural change perspective with different types of sectors, both existing and emerging, all (in)directly affected. Although the findings following the latter perspective are speculative (mostly because they hinge on potential employment and wage growth in sectors not yet existing), the outcome is much less pessimistic about the future of employment. The presentation is concluded with a brief discussion of several policy interventions to overcome adverse effects of robotization and to enhance the dynamic efficiency of structural change.

Recently, there have been numerous attempts to develop ethical guidelines for roboticists and ethical robot design. Depending on the area of robot application, such attempts may take form of institutionalized initiatives, research committees, public debates and so on. This work discusses ethics in robotics through the lens of collaborative learning and on the example of the educational social robot. Both the design process and learning are viewed here as processes and practices developed together by a variety of actors located in specific material and socio-cultural contexts. In order to address ethical concerns, robot designers and robot users collaborate, communicate and act together while sharing specific educational goals. Such a collaborative process, however, often remains tacit, and hence, it requires further deliberate reflection on what meanings ethics has for different actors, what language they use and what learning strategies they apply to develop ethical frameworks for a given type of robots. The example of the educational robot shows that while there has been a common agreement on the need of building robots and using them in a way that robots serve and bring added value to the society, the actual development and application of ethical principles remains a largely informal and intuitive process. The conclusion is that ethics to a large extent remains in the domain of “being” rather than only “doing”, and hence, it often defies collaborative attempts to translate ethics into individual responsibilities or a specific code of conduct.

The presentation clarifies ongoing transformations in the care providers work practice and learning as occasioned by feeding assistive robotics (FAR). Citizens with low or no function in their arms are candidates to use FAR. Although the authorities in Denmark endorse this technology, it proves difficult both to recruit suitable citizens and to ensure implementation and sustained use over an extended period. By using material semiotics (Mol, Moser and Pols, 2010) as an analytic resource, both technological imaginaries (Jasanoff & Kim, 2015) and infrastructural reconfigurations (Slota & Bowker, 2017) implicated by FAR is examined.  Despite the fact that the FAR intends to fit neatly into the work, home and day care environments, it seriously interferes with existing care routines. The food needs to be of a certain kind. It has to be prepared in certain ways, and the care provider has to monitor the meal closely. This is elucidated by way of two cases. A successful use over a long period, and a situation where the user had to stop using the FAR due to neck pain. On this ground, the notions of “tinkering” and “silent work” are discussed. The presentation takes its empirical outset in observations of meals with FAR and 11 interviews with roboticists and affected stakeholders.