Nailing Prediction: Experimental Evidence on the Value of Tools in Predictive Model Development DOI

Daniel Yue,

Paul B. Hamilton, Iavor Bojinov

et al.

SSRN Electronic Journal, Journal Year: 2022, Volume and Issue: unknown

Published: Jan. 1, 2022

Predictive model development is understudied despite its importance to modern businesses. Although prior discussions highlight advances in methods (along the dimensions of data, computing power, and algorithms) as primary driver quality, value tools that implement those has been neglected. In a field experiment leveraging predictive data science contest, we study by restricting access software libraries for machine learning models. By only allowing these our control group, find teams with unrestricted perform 30% better log-loss error — statistically economically significant amount, equivalent 10-fold increase training set size. We further high general data-science skills are less affected intervention, while tool-specific significantly benefit from modeling libraries. Our findings consistent mechanism call 'Tools-as-Skill,' where tooling automates abstracts some but, doing so, creates need new skills.

Language: Английский

Managing with Artificial Intelligence: An Integrative Framework DOI

Luis Hillebrand,

Sebastian Raisch,

Jonathan Schad

et al.

Academy of Management Annals, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 8, 2025

Managing with artificial intelligence (AI) refers to humans' interaction algorithms performing managerial tasks in organizations. Two literatures exploring this interaction—human-AI collaboration (HAIC) and algorithmic management (AM)—have focused on distinct tasks: while HAIC examines executive decision-making, AM focuses control. This article presents a review of both identify opportunities for integration advancement. We observe that HAIC's AM's micro-level emphases different have resulted diverging conceptualizations context, agency, interaction, outcome. Adopting more encompassing systems lens, we unveil previously concealed linkages between AM, suggesting the two analyzed sides same phenomenon: explores how humans use AI manage, describes are managed by AI. develop an integrative framework elevates viewpoint from organizational individual collective local systemic multilevel outcomes. By employing framework, lay foundations perspective managing

Language: Английский

Citations

1
Leveraging AI to improve evidence synthesis in conservation DOI
Oded Berger‐Tal, Bob B. M. Wong, Carrie Ann Adams

et al.

Trends in Ecology & Evolution, Journal Year: 2024, Volume and Issue: 39(6), P. 548 - 557

Published: May 24, 2024

Language: Английский

Citations

8
The emergence of a ‘twin transition’ scientific knowledge base in the European regions DOI
Giacomo Damioli, Stefano Bianchini, Claudia Ghisetti

et al.

Regional Studies, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 17

Published: June 12, 2024

This study reveals both the spatial and combinatory patterns of digital green scientific knowledge bases for creation new in domain so-called 'twin transition'. The recent rapid diffusion this twin European regions has not adhered to clearly defined been characterised by a dynamic process actor reconfiguration. However, with strong science base have greater propensity produce more, better quality more visible knowledge. Among most prevalent fields emerging today are artificial intelligence (AI)- Internet Things (IoT)-powered applications energy storage, distribution consumption; environmental monitoring modelling; urban planning.

Language: Английский

Citations

5
Disciplinary differences in undergraduate students' engagement with generative artificial intelligence DOI Creative Commons

Yao Qu,

Michelle Tan, Jue Wang

et al.

Smart Learning Environments, Journal Year: 2024, Volume and Issue: 11(1)

Published: Nov. 11, 2024

Abstract The rapid development of generative artificial intelligence (GenAI) technologies has sparked widespread discussions about their potential applications in higher education. However, little is known how students from various disciplines engage with GenAI tools. This study explores undergraduate students' knowledge, usage intentions, and task-specific engagement across academic disciplines. Using a disciplinary categorization framework, we examine the hard/soft pure/applied dimensions relate to interactions GenAI. We surveyed 193 undergraduates diverse at university Singapore. questionnaire assessed for cognitive routine tasks against background. results indicate substantial disparities level Compared pure fields, applied fields (both hard soft) consistently exhibit levels knowledge utilization intentions. Furthermore, relatively consistent disciplines; however, there are tasks, exhibiting engagement. These suggest that practical orientation drives adoption settings. emphasizes considering differences better integrate into education calls tailored approaches align each field's unique epistemological methodological traditions balance GenAI's benefits preservation core skills.

Language: Английский

Citations

3
Does work overload of odd-job platform workers lead to turnover intention? An empirical study on platform workers DOI
瑠津子 上山, Guang Xu, Zhong Jie

et al.

Baltic Journal of Management, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 18, 2024

Purpose Against the background of digital economy, odd-job platforms rely on artificial intelligence algorithms to efficiently allocate tasks and monitor platform workers’ performance, putting these workers under enormous pressure. This paper explores relationship between work overload turnover intention factors that lead turnover. Design/methodology/approach Based job demands–resources model (JD-R), we construct a theoretical explain workers. We test burnout as mediator variable perceived algorithmic fairness autonomy moderating variables. conducted study at food delivery ride-hailing in China. Findings The empirical results show that: (1) increases by increasing (2) moderate positive burnout. Originality/value provide basis influence practical recommendations for management

Language: Английский

Citations

2
Interdisciplinary Research in Artificial Intelligence: Lessons from COVID-19 DOI Creative Commons
Diletta Abbonato, Stefano Bianchini, Floriana Gargiulo

et al.

Quantitative Science Studies, Journal Year: 2024, Volume and Issue: 5(4), P. 922 - 935

Published: Jan. 1, 2024

Abstract Artificial intelligence (AI) is widely regarded as one of the most promising technologies for advancing science, fostering innovation, and solving global challenges. Recent years have seen a push teamwork between experts from different fields AI specialists, but outcomes these collaborations yet to be studied. We focus on approximately 15,000 papers at intersection COVID-19—arguably major challenges recent decades—and show that interdisciplinary medical professionals specialists largely resulted in publications with low visibility impact. Our findings suggest impactful research depends less overall author teams more diversity knowledge they actually harness their research. conclude team composition significantly influences successful integration new computational into science obstacles still exist effective realm AI.

Language: Английский

Citations

0
Dali or DALL-E? Popper or …? The Implications of Emerging Generative AI on the Future of Creative Work DOI

Sandra Barbosu,

Pooyan Khashabi

SSRN Electronic Journal, Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

Recent advances in artificial intelligence (AI) have sparked renewed debates about the impact of AI on workforce, industry, and society. In this article, we explore implications novel generative future creative work, distinguishing between two types crea-tivity: scientific artistic. We draw diverse streams literature to highlight distinctions two, argue that while could complement tasks both types, it is more likely substitute humans art rather than science, as creativity relies heavily causal inference, a skill does not possess. Finally, propose moderating factors for different industries, namely indus-try-level importance basic knowledge, exposure AI.

Language: Английский

Citations

1
Mapping the Landscape of Algorithmic Management: Insights from Bibliometrics Using Citespace DOI

Nhan Kim Vo

Published: Jan. 1, 2024

Language: Английский

Citations

0
Nailing Prediction: Experimental Evidence on the Value of Tools in Predictive Model Development DOI

Daniel Yue,

Paul B. Hamilton, Iavor Bojinov

et al.

SSRN Electronic Journal, Journal Year: 2022, Volume and Issue: unknown

Published: Jan. 1, 2022

Predictive model development is understudied despite its importance to modern businesses. Although prior discussions highlight advances in methods (along the dimensions of data, computing power, and algorithms) as primary driver quality, value tools that implement those has been neglected. In a field experiment leveraging predictive data science contest, we study by restricting access software libraries for machine learning models. By only allowing these our control group, find teams with unrestricted perform 30% better log-loss error — statistically economically significant amount, equivalent 10-fold increase training set size. We further high general data-science skills are less affected intervention, while tool-specific significantly benefit from modeling libraries. Our findings consistent mechanism call 'Tools-as-Skill,' where tooling automates abstracts some but, doing so, creates need new skills.

Language: Английский

Citations

0