Artificial Intelligence as a Service: Trade-Offs Impacting Service Design and Selection

2023 | International Conference on Information Systems | Citations: 0

Authors: Brecker, Kathrin; Lins, Sebastian; Trenz, Manuel; Sunyaev, Ali

Abstract: AI as a Service (AIaaS) is a promising path to leverage AI capabilities from th ... Expand

Abstract: AI as a Service (AIaaS) is a promising path to leverage AI capabilities from the cloud. However, there is no one-size-fits-all service, but heterogenous service options since interdependent AIaaS characteristics require trade-offs. We lack knowledge on these tradeoffs and how they result from conflicting characteristics. Therefore, we interviewed 39 AIaaS providers, customers, and consultants to provide rich descriptions of interdependent characteristics and uncover resulting trade-offs and their consequences. This study contributes to a better understanding of the inner functioning and interplay of AIaaS characteristics and discusses how this complex nature of service offerings impacts providers' design and customers' selection decisions. Collapse

Topics: artificial intelligence cloud computing ease of use data protection economies of scale

Methods: qualitative interview design artifact constant comparative analysis qualitative coding triangulation

Theories: SERVQUAL core competency theory

Bots with Feelings: Should AI Agents Express Positive Emotion in Customer Service?

2023 | Information Systems Research | Citations: 0

Authors: Han, Elizabeth; Yin, Dezhi; Zhang, Han

Abstract: Customer service employees are generally advised to express positive emotion dur ... Expand

Abstract: Customer service employees are generally advised to express positive emotion during their interactions with customers. The rise and maturity of artificial intelligence (AI)–powered conversational agents, also known as chatbots, beg the question: should AI agents be equipped with the ability to express positive emotion during customer service interactions? This research explores how, when, and why an AI agent’s expression of positive emotion affects customers’ service evaluations. We argue that AI-expressed positive emotion can influence customers via dual pathways: an affective pathway of emotional contagion and a cognitive pathway of expectation–disconfirmation. We propose that positive emotion expressed by an AI agent (versus a human employee) is less effective in facilitating service evaluations because of a heightened level of expectation–disconfirmation. We further introduce a novel individual difference variable, customers’ relationship norm orientation, which affects their expectations toward the AI agent and moderates the cognitive pathway. Results from three laboratory experiments substantiate our claims. By revealing a distinctive impact of positive emotion expressed by an AI agent compared with a human employee, these findings deepen our understanding of customers’ reactions to emotional AIs, and they offer valuable insights for the deployment of AIs in customer service. History: Deepa Mani, Senior Editor; Pallab Sanyal, Associate Editor. Supplemental Material: The online appendix is available at https://doi.org/10.1287/isre.2022.1179. Collapse

Topics: artificial intelligence service quality customer service human AI collaboration social media

Methods: analysis of covariance experiment statistical hypothesis test chatbot theory development

Theories: information theory dual-system theory

Implementing Choices in Chatbot-initiated Service Interactions: Helpful or Harmful?

2022 | International Conference on Information Systems | Citations: 0

Authors: Han, Elizabeth; Yin, Dezhi (Denny); Zhang, Han

Abstract: Chatbots are increasingly equipped to provide choices for customers to click an ... Expand

Abstract: Chatbots are increasingly equipped to provide choices for customers to click and choose from when communicating with the chatbots. This research investigates when and why implementing choices enhances or impairs customers' service experience. Based on the concept of fluency, we posit that the implementation of choices is beneficial only after a conversational breakdown occurs because the value of choice provision for facilitating fluency may not be recognizable or realized in the absence of service breakdowns. We further propose that the implementation of choices is counterproductive when the choice set is perceived as incomprehensive because it decreases the perception of fluency. We conducted several experiments to test these hypotheses. By illuminating when and why choice implementation may help or harm customers during a chatbot-initiated service interaction, we augment the current understanding of a chatbot's role in customers' service experience and provide insights for the deployment of choice-equipped chatbots in customer service. Collapse

Topics: task complexity service quality customer service human AI collaboration JavaScript

Methods: chatbot experiment multivariate analysis of variance business process modeling moderation analysis

Integrating human knowledge into artificial intelligence for complex and ill-structured problems: Informed artificial intelligence

2022 | International Journal of Information Management | Citations: 0

Authors: Johnson, Marina; Albizri, Abdullah; Harfouche, Antoine; Fosso-Wamba, Samuel

Abstract: Artificial intelligence (AI) has been gaining significant attention in various f ... Expand

Abstract: Artificial intelligence (AI) has been gaining significant attention in various fields to reduce costs, increase rev­ enue, and improve customer satisfaction. AI can be particularly beneficial in enhancing decision-making pro­ cesses for complex and ill-structured problems that lack transparency and have unclear goals. Most AI algorithms require labeled datasets to learn the problem characteristics, draw decision boundaries, and generalize. However, most datasets collected to solve complex and ill-structured problems do not have labels. Additionally, most AI algorithms are opaque and not easily interpretable, making it hard for decision-makers to obtain model insights for developing effective solution strategies. To this end, we examine existing AI paradigms, mainly symbolic AI (SAI) guided by human domain knowledge and data-driven AI (DAI) guided by data. We propose an approach called informed AI (IAI) by integrating human domain knowledge into AI to develop effective and reliable data labeling and model explainability processes. We demonstrate and validate the use of IAI by applying it to a social media dataset comprised of conversations between customers and customer support agents to construct a so­ lution – IAI defect explorer (I-AIDE). I-AIDE is utilized to identify product defects and extract the voice of cus­ tomers to help managers make decisions to improve quality and enhance customer satisfaction. Collapse

Topics: artificial intelligence social media Twitter decision making shapley additive explanation

Methods: computational algorithm support vector machine design science data transformation machine learning

Interruptions during a service encounter: Dealing with imperfect chatbots

2021 | International Conference on Information Systems | Citations: 2

Authors: Han, Elizabeth; Yin, Dezhi; Zhang, Han

Abstract: ... Expand

Abstract: Collapse

Topics: service quality artificial intelligence customer service human AI collaboration business process innovation

Methods: chatbot parametric test experiment two-way ANOVA theory development

Theories: reactance theory