API
< Back to all clusters
[TECHNOLOGY] · 7 sources

started · updated

Artificial Intelligence challenges in model selection and academic citation

The rapid adoption of Large Language Models (LLMs) has brought significant technical and ethical challenges to the forefront. Technically, there is a growing emphasis on selecting appropriate AI model types to optimize return on investment and reduce total cost of ownership. While LLMs are widely used, they represent only one subset of artificial intelligence; other categories such as machine learning and deep learning offer specialized capabilities that may be more efficient for specific tasks depending on the required GPU power and energy consumption.

Ethically, the use of LLMs in academic settings presents difficulties regarding proper citation. Because many chatbots function as a ‘plagiaristic black-box’—imitating training data without clearly identifying original authors—it is difficult for students to give proper credit. Furthermore, while some advanced models can point to specific training data, these are often too expensive for universal access, creating potential inequities in educational environments. Additionally, the inability to reliably detect AI-generated text complicates the evaluation of student work.

Entities

VMware