Collective Intelligence Under a Volatile Task Environment: a Behavioral Experiment Using Social Networks and Computer Simulations

Collective intelligence in our highly-connected world is a topic of interdisciplinary interest. Previous research has demonstrated that social network structures can affect collective intelligence, but the potential network impact is unknown when the task environment is volatile (i.e., optimal behavioral options can change over time), a common situation in modern societies. Here, we report a laboratory experiment in which a total of 250 participants performed a “restless” two-armed bandit task either alone, or collectively in a centralized or decentralized network. Although both network conditions outperformed the solo condition, no sizable performance difference was detected between the centralized and decentralized networks. To understand the absence of network effects, we analyzed participants’ behavior parametrically using an individual choice model. We then conducted exhaustive agent-based simulations to examine how different choice strategies may underlie collective performance in centralized or decentralized networks under volatile or stationary task environments. We found that, compared to the stationary environment, the difference in network structure had a much weaker impact on collective performance under the volatile environment across broad parametric variations. These results suggest that structural impacts of networks on collective intelligence may be constrained by the degree of environmental volatility.

The effects of task environment and organizational agility on perceived managerial discretion and strategy implementation in a pharmaceutical company

Purpose This study aims to investigate the relationship between task environment, organizational agility, perceived managerial discretion and strategy implementation on unit performance. Design/methodology/approach Based on the literature review, a structural model was developed. A 74-item questionnaire was circulated among middle managers in sales and marketing. The data collection method used purposive sampling. A total of 228 valid responses were obtained. This study was conducted in a leading pharmaceutical company in Indonesia. The data were analyzed using structural equation modeling. Findings Based on the data analysis, this study shows that task environment and organizational agility act as antecedents of perceived managerial discretion, which drives strategy implementation resulting in unit performance. Originality/value Different from previous studies that examined the linkage of inertial forces and discretion, this research scrutinized the effects of organizational agility on perceived managerial discretion and the direct role of perceived managerial discretion on internal strategy implementation.

Development and Validation of a Multi-Dimensional Measure of Activity-Based Working Behaviors

Most work on activity-based working centers on the physical environment and digital technologies enabling flexible working. While important, we believe the key components for implementing activity-based working are employee and manager behaviors. To measure the degree of enactment of activity-based work, based on workshops with experienced practitioners as well as previous literature, we have developed and validated a behavior-focused measure of activity-based working behaviors. In our initial sample (Sample 1, N = 234), three subscales were identified: task – environment crafting, workday planning, and social needs prioritization. In the replication sample (Sample 2, N = 434), this model also showed adequate fit. Moreover, task – environment crafting was related to general health and lower stress in sample 1 (multi-organization sample), but not in the single-organization sample (sample 2). Workday planning was associated with higher concentration in both samples and in the second sample with general health and work engagement; the latter was also related to social needs prioritization.

Inhibitory Cognitive Control Allows Automated Advice to Improve Accuracy While Minimizing Misuse

Humans increasingly use automated decision aids. However, environmental uncertainty means that automated advice can be incorrect, creating the potential for humans to act on incorrect advice or to disregard correct advice. We present a quantitative model of the cognitive process by which humans use automation when deciding whether aircraft would violate requirements for minimum separation. The model closely fitted the performance of 24 participants, who each made 2,400 conflict-detection decisions (conflict vs. nonconflict), either manually (with no assistance) or with the assistance of 90% reliable automation. When the decision aid was correct, conflict-detection accuracy improved, but when the decision aid was incorrect, accuracy and response time were impaired. The model indicated that participants integrated advice into their decision process by inhibiting evidence accumulation toward the task response that was incongruent with that advice, thereby ensuring that decisions could not be made solely on automated advice without first sampling information from the task environment.

A Remote Synthetic Testbed For Human-Robot Teaming: An Iterative Design Process

Urban Search and Rescue (USAR) missions often involve a need to complete tasks in hazardous environments. In such situations, human-robot teams (HRT) may be essential tools for future USAR missions. Transparency and explanation are two information exchange processes where transparency is real-time information exchange and explanation is not. For effective HRTs, certain levels of transparency and explanation must be met, but how can these modes of team communication be operationalized? During the COVID-19 pandemic, our approach to answering this question involved an iterative design process that factored in our research objectives as inputs and pilot studies with remote participants. Our final research testbed design resulted in converting an in-person task environment to a completely remote study and task environment. Changes to the study environment included: utilizing user-friendly video conferencing tools such as Zoom and a custom-built application for research administration tasks and improved modes of HRT communication that helped us avoid confounding our performance measures.

Human Factors Considerations for Teaming between Construction Workersand Voice-based Intelligent Virtual Agent (VIVA)

Construction sites experience low productivity due to particular characteristics such as unique designs in each project, sporadic arrival of projects, and complexity of the process. Another contributing factor to low productivity is poor communication among workers, supervisors, and a site’s centralized knowledge hub. Research shows that introducing advanced artificial intelligence (AI) technology in construction can tackle these problems. In this paper, we analyzed human factors considerations–user, task, environment, and technology and identified their characteristics and challenges to design an interactive AI system to facilitate communication between workers and other stakeholders. Based on the analysis, we propose a voice-based intelligent virtual agent (VIVA) as a multi-purpose AI system on construction sites with a further research agenda. We hope that this effort can guide the design of construction-specific AI systems and that this worker-AI teaming can improve overall work processes, enhance productivity, and promote safety in construction.

Designing for Bi-Directional Transparency in Human-AI-Robot-Teaming

This paper takes a practitioner’s perspective on advancing bi-directional transparency in human-AI-robot teams (HARTs). Bi-directional transparency is important for HARTs because the better that people and artificially intelligent agents can understand one another’s capabilities, limits, inputs, outputs and contexts in a given task environment; the better they can work as a team to accomplish shared goals, interdependent tasks, and overall missions. This understanding can be built, augmented, broken and repaired at various stages across the technology life cycle, including the conceptual design; iterative design of software, hardware and interfaces; marketing and sales; system training; operational use; and system updating and adaptation stages. This paper provides an overview of some best practices and challenges in building this bi-directional transparency at different points in the technology life cycle of human-AI-robot systems. The goal is to help advance a wider discussion and sharing of lessons learned from recent work in this area.

SpearSim: Design and Evaluation of Synthetic Task Environment for Studies on Spear Phishing Attacks

Despite significant advancements in security technologies, phishing attacks continue to be rampant and successful because distinguishing phishing emails from real messages remains difficult to most end-users, mainly the targeted kind known as spear-phishing. There is a severe lack of human factor studies on spear-phishing attacks due to lack of methods and datasets. We have designed a novel multi-player synthetic task environment, called SpearSim, for conducting laboratory experiments on spear-phishing attacks. Using SpearSim, we have conducted an experiment to understand how information exploitation in spear-phishing attacks influences end-user decision-making. This paper describes the SpearSim system’s design and discusses the results from the experiment conducted with SpearSim. The experiment results show that people are more vulnerable to spear-phishing attacks when attackers can explore and exploit different kinds of personal information available to them about their targets. We discuss the implications of this research for the design of anti-phishing training solutions and privacy enhancing technologies.

Building a Synthetic Task Environment to Support Artificial Social Intelligence Research

To support research on artificial social intelligence for successful teams (ASIST), an urban search and rescue task (USAR) was simulated within Minecraft to serve as a Synthetic Task Environment (STE). The goal for the development of the present STE was to create an environment that provides ample opportunities to allow ASI agents to demonstrate the theory of mind by making inferences and predictions of humans’ states and actions in the USAR task environment, and in the future to intervene to improve teamwork in real-time. This paper describes the STE design background, design potentials and considerations, rich data collection opportunities, and potential usage for more broad research.