Submit a Manuscript to the Journal

Applied Artificial Intelligence

For an Article Collection on

Primary Theoretical and Practical Applications and Limitations of Generative Artificial Intelligence

Manuscript deadline
19 January 2025

Cover image - Applied Artificial Intelligence

Article collection guest advisor(s)

Dr. Silvia García Méndez, University of Vigo
[email protected]

Dr. Francisco de Arriba-Pérez, University of Vigo
[email protected]

Submit an ArticleVisit JournalArticles

Primary Theoretical and Practical Applications and Limitations of Generative Artificial Intelligence

In recent years, generative artificial intelligence models have emerged with promising capabilities. For instance, models like Google Bard, ChatGPT, DALL-E, and Midjourney allow the creation of virtual assistants, summarizing texts, detecting entities, and automatically creating images.

However, generative artificial intelligence is not free from limitations and challenges that have yet to be explored. These models lack memory, are too heavy for use on conventional computers and servers, and exhibit privacy risks. Furthermore, their predictions need to be traced, which limits their ability to explain and interpret and reduces users' confidence in them. Moreover, the data sets used for training may also be subject to bias while at the same time lacking enough global context knowledge, that is, information for particularisation and personalisation.

Although the potential of generative artificial intelligence models is evident, their limitations still need to be addressed. Finding solutions that include short- and long-term memory allows us to personalize and contextualize the system responses. Reducing the hidden layers would allow us to make on-demand models executable on machines with limited resources and increase privacy by limiting access by third-party entities. Furthermore, achieving explainable models would allow end users to understand the nature of the decisions made by these systems. Image-generating systems are a representative example. Tagging the sources contributing to a new picture can significantly streamline assigning specific copyrights. This paves the way for effective regulation and monetization of human content creators, who may perceive these systems as a threat rather than an opportunity.

This call aims to collect original articles focusing on the primary limitations of generative artificial intelligence models, not only practical (e.g., context and memory management, layer pruning, etc.) but also theoretical (e.g., data privacy through encrypting methodologies), towards ethical and cognitive, self-awareness AI.

These works must analyse the limitations and feasible solutions. Thus, both practical and theoretical works that explore the hidden layers, provide alternatives for traceability, and develop applications in relevant and multidisciplinary use cases will be considered.

The main topics are:

• Lack of memory in Large Language Models (LLM)
• On-demand generative artificial intelligence models
• Pruned models
• LLM encryption
• Privacy in generative artificial intelligence
• Explain and interpret in generative artificial intelligence
• Personalized generative artificial intelligence models
• Engineering applications of LLMs
• Generative artificial intelligence in multidisciplinary use cases

Benefits of publishing open access within Taylor & Francis

Global marketing and publicity, ensuring your research reaches the people you want it to.

Article Collections bring together the latest research on hot topics from influential researchers across the globe.

Rigorous peer review for every open access article.

Rapid online publication allowing you to share your work quickly.

All manuscripts submitted to this Article Collection will undergo desk assessment and peer-review as part of our standard editorial process. Guest Advisors for this collection will not be involved in peer-reviewing manuscripts unless they are an existing member of the Editorial Board. Please review the journal Aims and Scope and author submission instructions prior to submitting a manuscript.