Charting the ChatGPT landscape: Insights from academic discourse on Twitter

In the wake of the emergence of ChatGPT as a transformative tool, effective policies and regulations for its integration into research and education are vital. This paper addresses challenges in the rapidly evolving AI landscape by identifying key events and discussions. To that end, the current study extracted insights from the academic community that had come during the early stages of ChatGPT's adoption. A dataset of 84,706 sentences sourced from Twitter (X.com) users in the research and academic community and collected over the course of eight months between November 2022 and June 2023 were examined using topic modeling and aspect-based sentiment analysis to explore prevailing reactions and perceptions. Nine key themes such as academic writing, coding, and time-saving tasks were identified. Strong sentiments emerged around policy debates, the issue of data security, and evolving research practices. Further, a causal analysis was performed to identify discussions and events that had triggered shifts in public sentiment. Examples include temporary restrictions on generative AI, new institutional policies, and legislative efforts to ensure responsible AI integration. This paper provides a timeline-linked perspective on how the academic community, thus far, has responded to generative AI. The findings can inform pragmatic, context-sensitive policies that foster innovation while safeguarding academic values.