Against the moral Turing test: accountable design and the moral reasoning of autonomous systems

This paper argues against the moral Turing test (MTT) as a framework for evaluating the moral performance of autonomous systems. Though the term has been carefully introduced, considered, and cautioned about in previous discussions (Allen et al. in J Exp Theor Artif Intell 12(3):251–261, 2000; Allen and Wallach 2009), it has lingered on as a touchstone for developing computational approaches to moral reasoning (Gerdes and Øhrstrøm in J Inf Commun Ethics Soc 13(2):98–109, 2015). While these efforts have not led to the detailed development of an MTT, they nonetheless retain the idea to discuss what kinds of action and reasoning should be demanded of autonomous systems. We explore the flawed basis of an MTT in imitation, even one based on scenarios of morally accountable actions. MTT-based evaluations are vulnerable to deception, inadequate reasoning, and inferior moral performance vis a vis a system’s capabilities. We propose verification—which demands the design of transparent, accountable processes of reasoning that reliably prefigure the performance of autonomous systems—serves as a superior framework for both designer and system alike. As autonomous social robots in particular take on an increasing range of critical roles within society, we conclude that verification offers an essential, albeit challenging, moral measure of their design and performance.     

Teleology as a tacit dimension of teaching and learning evolution: A sociological approach to classroom interaction in science education

Teleology has been described as an intuitive cognitive bias and as a major type of student conception. There is controversy regarding whether teleological explanations are a central obstacle to, are legitimate in, or are even supportive of science learning. However, interaction in science classrooms has not yet been investigated with regard to teleology. Consequently, this study addresses the question of how teleological explanations emerge in science classroom interactions about evolution and how teachers and students address emerging teleology. In this article, we introduce a theoretical and methodological framework drawing from the sociology of knowledge and systems theory, suggesting that this framework may enrich the understanding of knowledge construction and of social practices in the science classroom because it enables distinguishing between explicit and tacit knowledge. We investigated seven secondary school units about evolution and present data from four grade-12 classes in Germany, a country with very few creationists, to contrast two ways in which teleology is addressed. In the first type, the teachers combine intentional and need-based teleological explanations with aspects of scientific theories in an ambiguous way. Contrastingly, in the second type, the teachers construct a duality between correct mechanistic and incorrect teleological explanations by discrediting preceding scientific theories. In the discussion, we argue that the presented sociological approach can also be valuable in other science education contexts, such as creationism, the nature of science and socio-scientific issues, because classroom interaction involves tacit communication, such as a tacit epistemology, which are essential grounds for the students' knowledge construction.     

Performanzfortschritt in der Lehrerausbildung durch die Arbeit an eigenen video- und textbasierten Unterrichtsfällen? Eine Interventionsstudie zur Verbesserung des Feedbacks bei angehenden Sportlehrkräften

This intervention study examines the different effects that work on video- or text-based teaching cases has on improving feedback-related performance (dependent variable) for trainee PE teachers. Within a teaching-learning arrangement, one group (n?=?24) worked with video-based teaching cases while another group (n?=?9) worked with text-based ones. In addition, both groups completed a school practicum. Additionally, a control group (n?=?23) completed the practicum, without any teaching case intervention. The quality of the feedback-related performance of the trainee PE teachers was recorded using video vignettes both before and after the respective interventions and then evaluated on the basis of an external assessment (expert rating). The results show that the work on video- and text-based teaching cases was effective, but that the videobased procedure had more of an impact, with a high level of practical significance. The control group was unable to improve on the basis of the placement within a school. The results demonstrate their usefulness for future training and competence diagnostics.     

Situating teachers’ developmental engineering experiences in an inquiry-based,laboratory learning environment

Many secondary math and science teachers don’t understand the nature and application of engineering adequately to transfer that understanding to their students. Research is needed that investigates and illuminates the process and characteristics of development that addresses this gap. This mixed-method study examines the developmental experiences of 16 public math and science teachers in a yearlong engineering experience. Teachers partnered with university engineering professor-mentors for six weeks on campus followed by ongoing support for transfer to their secondary classes (grades 9–12). Teachers gained engineering knowledge and skills, and an understanding of engineering’s diversity and interdisciplinary nature, plus integrative applied strategies for math and science instruction. They developed productive perceptions (of value, utility, benefits, feasibility and fit) of engineering for their grade 9–12 students, along with efficacy and intent to use what they had learned. Teachers identified positive, profound shifts in their perceptions of engineering and engineers. All teachers integrated engineering principles into their classrooms, but the process was more challenging for math teachers than for science teachers. This developmental experience may help bridge gaps of school-to-life and disciplinary boundaries that separate formal educational environments from authentic life experience.     

Tools of the Trade: UK Research Intermediaries and the Politics of Impacts

In recent years questions concerning the impact of public research funding have become the preeminent site at which struggles over the meanings and value of science are played out. In this paper we explore the ‘politics of impact’ in contemporary UK science and research policy and, in particular, detail the ways in which UK research councils have responded to and reframed recent calls for the quantitative measurement of research impacts. Operating as ‘boundary organisations’ research councils are embroiled in what might be characterised as the ‘politics of demarcation’ in which competing understandings of the cultural values of science are traded, exchanged and contested. In this paper we focus on the way the UK’s ‘Engineering and Physical Sciences Research Council’ (EPSRC) has responded to contemporary policy discourses concerning the impacts of public research expenditure. We argue that, in response to the shifting terms of contemporary science policy, the EPSRC has adopted three distinct strategies. Firstly, in collaboration with other research councils the EPSRC have emphasised the intellectual and metrological challenge presented by attempts to quantify the economic impact of public research expenditure, emphasising instead the cumulative impacts of a broad portfolio of ‘basic science’. Secondly, the EPSRC has sought to widen the discursive meaning of research impacts – specifically to include societal and policy impacts in addition to economic ones. Thirdly, the EPSRC has introduced a new framing into the ‘impact agenda’, preferring to talk about ‘pathways to impact’ rather than research impacts per se. In responding to government priority setting, we argue that the EPSRC has sought to exploit both the technical fragility of auditing techniques and the discursive ambiguity of notions of impact.     

Evolution?rer Test von Realzeitsystemen

Zusammenfassung.   Die Entwicklung von Realzeitsystemen ist eine wesentliche Aufgabe im industriellen Umfeld. Der dynamische Test ist die wichtigste analytische Qualit?tssicherungsmethode für Realzeitsysteme. Der dynamische Test ist das einzige Verfahren, mit dem das Laufzeitverhalten der Systeme in der tats?chlichen Einsatzumgebung überprüft werden kann. Bestehende Testmethoden konzentrieren sich weitgehend auf den Test des funktionalen Verhaltens von Systemen. Sie bieten keine besondere Unterstützung für den Test des zeitlichen Verhaltens. Für die Korrektheit von Realzeitsystemen sind jedoch beide Aspekte gleicherma?en wichtig. Folglich müssen die bestehenden Methoden um neue Ans?tze erg?nzt werden, die auf die Prüfung des zeitlichen Verhaltens spezialisiert sind. Ein Fehler im Zeitverhalten von Realzeit-Systemen liegt dann vor, wenn es zul?ssige Eingabesituationen gibt, für die das System nicht in der Lage ist, die spezifizierten zeitlichen Anforderungen zu erfüllen. In den meisten F?llen ist dies darauf zurückzuführen, da?das System eine Ausgabe zu früh macht oder da?die Berechnung einer Ausgabe zu lange dauert und diese daher zu sp?t erfolgt. Aufgabe des Testers ist es folglich, die Eingabesituationen mit den l?ngsten und kürzesten Ausführungszeiten zu finden, um zu überprüfen, ob die für das System spezifizierten Zeitschranken verletzt werden. Wird die Suche nach entsprechenden Eingabesituationen als Optimierungsproblem interpretiert, dann k?nnen evolution?re Algorithmen verwendet werden, um die Eingaben mit den l?ngsten und kürzesten Ausführungszeiten automatisch zu suchen. Die zu optimierende Zielfunktion entspricht hierbei den gemessenen Ausführungszeiten, die generierten Individuen repr?sentieren Testdaten. Die Verwendung evolution?rer Algorithmen für den Test nennen wir evolution?ren Test. Zum Vergleich des evolution?ren Tests mit anderen Testmethoden wurden in dieser Arbeit mehrere Experimente durchgeführt. In allen Experimenten erreicht der evolution?re Test l?ngere bzw. kürzere Ausführungszeiten als durch Zufallstests oder systematische Tests gefunden werden. Der evolution?re Test erscheint daher als gut geeignet für die Prüfung der zeitlichen Korrektheit von Realzeitsystemen. Die Kombination des evolution?ren Tests mit systematischen Tests kann die Testqualit?t weiter verbessern und wird in diesem Papier als Teststrategie für Realzeitsysteme empfohlen. Eingegangen am 14. Dezember 1999 / Angenommen am 30. Mai 2000     

INI-GraphicsNet

Ohne Zusammenfassung     

Überzeugungen vom Englischlernen

Der Beitrag fragt, ob das Konzept der überzeugungen für die sprachen-bzw. englischdidaktische Forschung von Nutzen sein kann. Dazu wird zun?chst kurz die Diskussion der epistemologischen überzeugungen von Schülerinnen und Schülern in der dritten TIMS-Studie vorgestellt. Danach werden einige Ergebnisse aus der angloamerikanischen und deutschsprachigen Forschungsliteratur zu überzeugungen vom Sprachenlernen referiert. Am Beispiel eines empirischen Forschungsprojektes werden ansatzweise überzeugungen zweier Englischlehrender zum Grammatiklernen analysiert. Im Ergebnis vertritt der Autor die These, dass die Erforschung der überzeugungen von Lehrenden und Lernenden im Kontext von Schule ein notwendiges und deshalb zu intensivierendes, wenngleich nicht hinreichendes Element der Entwicklung von fremdsprachlichem Unterricht ist.