Objective
While artificial intelligence offers immense capabilities to transform industry and society, companies continue to struggle with transitioning from exploratory proof-of-concept pilots to value-generating, reliable systems in real environments. This project has two primary objectives:
First, we aim to comprehensively analyze the factors affecting organizations’ capabilities to adopt AI. This understanding will enable companies to better address adoption challenges and develop effective strategies and techniques to overcome implementation barriers.
Second, a particular emphasis of this project is the organizational adoption of Generative AI (GenAI). Despite rapid technological developments and potential benefits, we need to investigate how to make this technology adoption-ready. Reliability concerns frequently prevent organizations and individuals from implementing GenAI solutions. For example:
- Large Language Models (LLMs), a subset of GenAI, often generate false information or hallucinations that can potentially mislead users
- Models must effectively communicate explanations for their behaviors
- Systems need to express uncertainty appropriately when they lack sufficient information
Methodological Approach
To achieve our objectives, we will employ a diverse spectrum of research methods:
1. Qualitative approaches such as Design Science Research to develop a deep understanding of adoption challenges
2. Technical experiments using both benchmark and real-world data to enhance model reliability
3. Quantitative studies to measure the effectiveness of our proposed methods
Funding and Partner