Type of Publication: Article in Collected Edition

Exploring Generative Artificial Intelligence: A Taxonomy and Types

Author(s):

Strobel, Gero; Banh, Leonardo; Möller, Frederik; Schoormann, Thorsten

Title of Anthology:

Proceedings of the 57th Hawaii International Conference on System Sciences (HICSS)

Location(s):

Hawaii, USA

Publication Date:

2024

Language:

Englisch

Keywords:

Generative Artificial Intelligence, Machine Learning, LLM, Taxonomy, Typology

Link to complete version:

https://scholarspace.manoa.hawaii.edu/items/2323c064-ba02-4fc5-826b-2ff3364bdfd9

Citation:

Download BibTeX

Abstract

Generative Artificial Intelligence (GAI) is a prevalent topic in recent research and business, seemingly taking the position of a disruptive technology that has the potential to significantly transform industries ranging from productivity (e.g., ChatGPT-4) to creativity (e.g., DALL-E). While the emerging scientific discussion on GAI covers a variety of fields and issues, such as privacy, accuracy, and application scenarios, this paper sheds light on the business side of GAI by investigating the morphologic nature of start-ups and incumbents leveraging GAI. Based on the structured analysis of 100 real-world instances, we report on a taxonomy of GAI applications and services that advances our practical understanding, strengthens the distinguishability, as well as adds clarity to the discourse of GAI potentials. We provide an initial framework and five types of GAI, namely Generator, Reimaginator, Synthesizer, Assistant, and Enabler, that are informed by the core characteristics of the technology paradigm.