MODUL University Vienna

As AI models tackle increasingly complex problems, ensuring reliable human oversight becomes more challenging due to the difficulty of verifying solutions. Approaches to scaling AI supervision include debate, in which two agents engage in structured dialogue to help a judge evaluate claims; critique, in which models identify potential flaws in proposed solutions; and prover-verifier games, in which a capable 'prover' model generates solutions that must be verifiable by a less capable 'verifier'. Evaluations of the scalability of these and similar approaches to difficult problems benefit from datasets that include (1) long-form expert-verified correct solutions and (2) long-form flawed solutions with annotations highlighting specific errors, but few are available. To address this gap, we present FindTheFlaws, a group of five diverse datasets spanning medicine, mathematics, science, coding, and the Lojban language. Each dataset contains questions and long-form solutions with expert annotations validating their correctness or identifying specific error(s) in the reasoning. We evaluate frontier models' critiquing capabilities and observe a range of performance that can be leveraged for scalable oversight experiments: models performing more poorly on particular datasets can serve as judges/verifiers for more capable models.

ABSTRACT Nonverbal communication represents a driving force behind successful selling. Drawing on the Emotions as Social Information (EASI) model, this research investigates the impact of salespeople's nonverbal expressiveness (manifested through dynamic facial, bodily, and vocal cues) and customer responses in a selling context. Innovative techniques for dynamic and holistic measurement of nonverbal signals are employed in a sales context, including facial recognition software, body action and posture coding, and voice analysis. A program analyzer traces customers' real‐time responses during exposure to videotaped sales presentations. Findings from large‐scale experimental studies are reported. Study 1 manipulates nonverbal expressiveness (IV: dynamic vs. restrained) and shows that expressive presentations lead to more favorable customer evaluations (DVs). Positive facial expressions—such as happiness—and direct head orientation significantly improve customer responses. Moreover, in line with findings in negotiation contexts, this study shows that under certain conditions, subtle displays of negative emotions—such as anger—can elicit positive responses in personal selling contexts. Study 2 manipulates product type (IV: hedonic vs. utilitarian) and finds that nonverbal expressiveness is more influential for hedonic offerings. This research provides novel insights into the complex dynamics of nonverbal communication and offers managerial guidance on effectively managing such behaviors in sales interactions.

Abstract The growing importance of big data and data analytics has made these skills indispensable for the tourism industry, necessitating an urgent revision of tourism education curricula. Previous research, however, lacks both ontological clarity and empirical findings regarding the specific skills and competencies required. To address this knowledge gap, this study employs a grounded theory approach through the analysis of job advertisements. The results first confirm a significant industry demand for data science and analytics (DSA) roles, answering why this topic is critical. Second, the study establishes an ontology of DSA by conceptualizing and categorizing the required skills and competencies, thus defining what is needed. Finally, by identifying specific teams and job focus areas, the research clarifies where and how these skills can be applied in practice and integrated into educational syllabi. The study concludes by discussing key theoretical, methodological, and practical implications that underscore its contribution to the field.