Minds and Machines, Год журнала: 2024, Номер 34(4)
Опубликована: Окт. 9, 2024
Язык: Английский
Minds and Machines, Год журнала: 2024, Номер 34(4)
Опубликована: Сен. 17, 2024
Язык: Английский
Процитировано
13
Journal of Radiological Protection, Год журнала: 2024, Номер 44(1), С. 011510 - 011510
Опубликована: Фев. 7, 2024
Abstract In recent times, the field of artificial intelligence (AI) has been transformed by introduction large language models (LLMs). These models, popularized OpenAI’s GPT-3, have demonstrated emergent capabilities AI in comprehending and producing text resembling human language, which helped them transform several industries. But its role yet to be explored nuclear industry, specifically managing radiation emergencies. The present work explores LLMs’ contextual awareness, natural interaction, their capacity comprehend diverse queries a emergency response setting. this study we identify different user types specific LLM use-cases Their possible interactions with ChatGPT, popular LLM, also simulated preliminary results are presented. Drawing on insights gained from exercise address concerns reliability misinformation, advocates for expert guided domain-specific LLMs trained safety protocols historical data. This aims guide management practitioners decision-makers effectively incorporating into decision support framework.
Язык: Английский
Процитировано
6
SSRN Electronic Journal, Год журнала: 2025, Номер unknown
Опубликована: Янв. 1, 2025
Язык: Английский
Процитировано
0
2022 IEEE 12th Annual Computing and Communication Workshop and Conference (CCWC), Год журнала: 2025, Номер unknown, С. 00633 - 00639
Опубликована: Янв. 6, 2025
Язык: Английский
Процитировано
0
Опубликована: Апрель 24, 2025
Язык: Английский
Процитировано
0
Analecta política, Год журнала: 2024, Номер 14(26)
Опубликована: Янв. 1, 2024
Este artículo realiza una revisión sistemática de treinta y seis taxonomías riesgos asociados a la Inteligencia Artificial (IA) que se han realizado desde el 2010 hasta fecha, utilizando como metodología protocolo Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). El estudio basa en importancia estas para estructurar investigación los distinguir definir amenazas. Ello permite identificar las cuestiones generan mayor preocupación y, por lo tanto, requieren mejor gobernanza. La extraer tres conclusiones. En primer lugar, observa mayoría estudios centran amenazas privacidad desinformación, posiblemente debido su concreción evidencia empírica existente. Por contrario, ciberataques desarrollo tecnologías estratégicas son menos citadas, pesar creciente relevancia. segundo encontramos artículos enfocados origen del riesgo tienden considerar más frecuentemente extremos comparación con trabajos abordan consecuencias. Esto sugiere literatura ha sabido potenciales causas catástrofe, pero no formas concretas esta puede materializar práctica. Finalmente, existe cierta división entre aquellos tratan daños tangibles presentes cubren futuros. No obstante, varias todo espectro indicando existen puntos unión clústeres.
Процитировано
3
Proceedings of the AAAI Symposium Series, Год журнала: 2024, Номер 3(1), С. 143 - 147
Опубликована: Май 20, 2024
This study delves into the application of large language models (LLMs), such as ChatGPT-4, for automated evaluation student essays, with a focus on case conducted at Swiss Institute Business Administration. It explores effectiveness LLMs in assessing German-language transfer assignments, and contrasts their performance traditional evaluations by human lecturers. The primary findings highlight challenges faced terms accurately grading complex texts according to predefined categories providing detailed feedback. research illuminates gap between capabilities nuanced requirements essay evaluation. conclusion emphasizes necessity ongoing development area LLM technology improve accuracy, reliability, consistency assessments educational contexts.
Язык: Английский
Процитировано
2
Minds and Machines, Год журнала: 2024, Номер 34(4)
Опубликована: Окт. 9, 2024
Язык: Английский
Процитировано
0