Preprint
Machine Learning

Generative artificial intelligence

Leonardo Banh(University of Duisburg-Essen), Gero Strobel(University of Duisburg-Essen)
December 1, 2023Electronic Markets486 citations
PDFDOI

486

Citations

4

Influential Citations

Electronic Markets

Venue

2023

Year

Abstract

Abstract Recent developments in the field of artificial intelligence (AI) have enabled new paradigms of machine processing, shifting from data-driven, discriminative AI tasks toward sophisticated, creative tasks through generative AI. Leveraging deep generative models, generative AI is capable of producing novel and realistic content across a broad spectrum (e.g., texts, images, or programming code) for various domains based on basic user prompts. In this article, we offer a comprehensive overview of the fundamentals of generative AI with its underpinning concepts and prospects. We provide a conceptual introduction to relevant terms and techniques, outline the inherent properties that constitute generative AI, and elaborate on the potentials and challenges. We underline the necessity for researchers and practitioners to comprehend the distinctive characteristics of generative artificial intelligence in order to harness its potential while mitigating its risks and to contribute to a principal understanding.

Analysis

Why This Paper Matters

This paper arrives at a critical juncture in AI development, where generative models have transitioned from niche research to mainstream tools impacting industries from content creation to software engineering. By providing a structured overview of generative AI's fundamentals, the authors address a pressing need for clarity amidst rapid technological change. The paper's value lies in its synthesis of disparate concepts into a coherent framework, making it accessible to both newcomers and experienced practitioners seeking a bird's-eye view.

The timing of publication (2023) coincides with the explosion of interest in large language models and diffusion models, making this overview particularly timely. The paper's emphasis on both potentials and challenges reflects a balanced perspective that is crucial for responsible adoption.

Technical Contributions

  • Defines generative AI as a paradigm shift from discriminative to creative machine processing.
  • Introduces deep generative models as the core technology enabling novel content generation across text, images, and code.
  • Clarifies the role of user prompts as the primary interface for controlling generative outputs.
  • Identifies key properties that distinguish generative AI from traditional AI systems.
  • Provides a taxonomy of generative AI applications and their underlying techniques.

Results

As a conceptual paper, no quantitative results or benchmarks are reported. The main output is a comprehensive framework that organizes existing knowledge about generative AI. The paper's impact is reflected in its 486 citations, indicating its adoption as a reference work in the field.

Significance

This paper contributes to the foundational understanding of generative AI, which is essential for both academic research and industrial application. By clearly delineating the capabilities and limitations of generative models, it helps practitioners make informed decisions about deployment. The work also highlights important ethical and societal considerations, encouraging responsible innovation. As generative AI continues to evolve, this overview provides a stable reference point for future developments.

generative-aideep-generative-modelsmachine-learningtext-generationimage-generationcode-generationuser-prompts
Prompting

Related Papers