THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE

Detalhes bibliográficos
Autor(a) principal: Cardoso, Fabio
Data de Publicação: 2024
Tipo de documento: Artigo
Idioma: por
Título da fonte: Revista GEMInIS
Texto Completo: https://www.revistageminis.ufscar.br/index.php/geminis/article/view/848
Resumo: The evolution of artificial intelligence text generation tools opens up the possibility of creatively using technology for the creation of audiovisual scripts. Language models like ChatGPT are capable of generating complex artistic content with user input. This enables the emergence of a new form of audiovisual content: continuous script generation. However, inherent technical difficulties in the current model of creating such software limit the effectiveness and scope of this form of creation. This article aims to investigate the most significant problem of this limitation, which is the lack of memory in LLM algorithms.
id UFSCAR-2_e61e1182ea934710adb3ca394d6565ae
oai_identifier_str oai:ojs.pkp.sfu.ca:article/848
network_acronym_str UFSCAR-2
network_name_str Revista GEMInIS
repository_id_str
spelling THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCEEL PROBLEMA DE (FALTA DE) MEMORIA EN NARRATIVAS GENERADAS POR INTELIGENCIA ARTIFICIALO PROBLEMA DA (FALTA DE) MEMÓRIA EM NARRATIVAS GERADAS POR INTELIGÊNCIA ARTIFICIALLarge Language ModelsMemóriaRoteiros AudiovisuaisLarge Language ModelsMemoriaGuiones AudiovisualesLarge Language ModelsMemoryAudiovisual ScreenplayThe evolution of artificial intelligence text generation tools opens up the possibility of creatively using technology for the creation of audiovisual scripts. Language models like ChatGPT are capable of generating complex artistic content with user input. This enables the emergence of a new form of audiovisual content: continuous script generation. However, inherent technical difficulties in the current model of creating such software limit the effectiveness and scope of this form of creation. This article aims to investigate the most significant problem of this limitation, which is the lack of memory in LLM algorithms.La evolución de las herramientas de generación de texto por inteligencia artificial abre la posibilidad del uso creativo de la tecnología para la creación de guiones audiovisuales. Modelos de lenguaje como ChatGPT son capaces de generar contenido artístico complejo con la entrada del usuario. Esto permite el surgimiento de una nueva modalidad de contenido audiovisual: la generación continua de guiones. Sin embargo, las dificultades técnicas inherentes al modelo actual de creación de este tipo de software limitan la eficacia y alcance de esta forma de creación. Este artículo tiene como objetivo investigar el problema más significativo de esta limitación, que es la falta de memoria en los algoritmos de LLM.A evolução das ferramentas de geração de texto por inteligência artificial abre a possibilidade do uso criativo da tecnologia para criação de roteiros audiovisuais. Modelos de linguagem como o ChatGPT são capazes de, com uma entrada do usuário, gerar conteúdo artístico complexo. Isso viabiliza o surgimento de uma nova modalidade de conteúdo audiovisual: a geração contínua de roteiros. No entanto, dificuldades técnicas inerente do modelo atual de criação desses softwares limitam a eficácia e abrangência dessa forma de criação. Este artigo busca investigar o problema mais significativo desta limitação, que é a falta de memória dos algoritmos de LLM.UFSCar2024-05-06info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttps://www.revistageminis.ufscar.br/index.php/geminis/article/view/84810.14244/2179-1465.RG.2023v14i3p44-53Revista GEMInIS; v. 15 n. 1 (2024): Narrativas de IA: tendências da produção audiovisual - Parte 2; 44-53Revista GEMInIS; Vol. 15 No. 1 (2024): AI narratives: trends in audiovisual production - Part 2; 44-53Revista GEMInIS; Vol. 15 Núm. 1 (2024): Narrativas de IA: tendencias en la producción audiovisual - Parte 2; 44-532179-1465reponame:Revista GEMInISinstname:Universidade Federal de São Carlos (UFSCAR)instacron:UFSCARporhttps://www.revistageminis.ufscar.br/index.php/geminis/article/view/848/536Copyright (c) 2024 Fabio Cardosohttp://creativecommons.org/licenses/by-sa/4.0info:eu-repo/semantics/openAccessCardoso, Fabio 2024-05-06T01:35:12Zoai:ojs.pkp.sfu.ca:article/848Revistahttps://www.revistageminis.ufscar.br/index.php/geminisPUBhttps://www.revistageminis.ufscar.br/index.php/geminis/oai||revista.geminisufscar@gmail.com2179-14652179-1465opendoar:2024-05-06T01:35:12Revista GEMInIS - Universidade Federal de São Carlos (UFSCAR)false
dc.title.none.fl_str_mv THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
EL PROBLEMA DE (FALTA DE) MEMORIA EN NARRATIVAS GENERADAS POR INTELIGENCIA ARTIFICIAL
O PROBLEMA DA (FALTA DE) MEMÓRIA EM NARRATIVAS GERADAS POR INTELIGÊNCIA ARTIFICIAL
title THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
spellingShingle THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
Cardoso, Fabio
Large Language Models
Memória
Roteiros Audiovisuais
Large Language Models
Memoria
Guiones Audiovisuales
Large Language Models
Memory
Audiovisual Screenplay
title_short THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
title_full THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
title_fullStr THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
title_full_unstemmed THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
title_sort THE PROBLEM OF (LACK OF) MEMORY IN NARRATIVES GENERATED BY ARTIFICIAL INTELLIGENCE
author Cardoso, Fabio
author_facet Cardoso, Fabio
author_role author
dc.contributor.author.fl_str_mv Cardoso, Fabio
dc.subject.por.fl_str_mv Large Language Models
Memória
Roteiros Audiovisuais
Large Language Models
Memoria
Guiones Audiovisuales
Large Language Models
Memory
Audiovisual Screenplay
topic Large Language Models
Memória
Roteiros Audiovisuais
Large Language Models
Memoria
Guiones Audiovisuales
Large Language Models
Memory
Audiovisual Screenplay
description The evolution of artificial intelligence text generation tools opens up the possibility of creatively using technology for the creation of audiovisual scripts. Language models like ChatGPT are capable of generating complex artistic content with user input. This enables the emergence of a new form of audiovisual content: continuous script generation. However, inherent technical difficulties in the current model of creating such software limit the effectiveness and scope of this form of creation. This article aims to investigate the most significant problem of this limitation, which is the lack of memory in LLM algorithms.
publishDate 2024
dc.date.none.fl_str_mv 2024-05-06
dc.type.driver.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.uri.fl_str_mv https://www.revistageminis.ufscar.br/index.php/geminis/article/view/848
10.14244/2179-1465.RG.2023v14i3p44-53
url https://www.revistageminis.ufscar.br/index.php/geminis/article/view/848
identifier_str_mv 10.14244/2179-1465.RG.2023v14i3p44-53
dc.language.iso.fl_str_mv por
language por
dc.relation.none.fl_str_mv https://www.revistageminis.ufscar.br/index.php/geminis/article/view/848/536
dc.rights.driver.fl_str_mv Copyright (c) 2024 Fabio Cardoso
http://creativecommons.org/licenses/by-sa/4.0
info:eu-repo/semantics/openAccess
rights_invalid_str_mv Copyright (c) 2024 Fabio Cardoso
http://creativecommons.org/licenses/by-sa/4.0
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv UFSCar
publisher.none.fl_str_mv UFSCar
dc.source.none.fl_str_mv Revista GEMInIS; v. 15 n. 1 (2024): Narrativas de IA: tendências da produção audiovisual - Parte 2; 44-53
Revista GEMInIS; Vol. 15 No. 1 (2024): AI narratives: trends in audiovisual production - Part 2; 44-53
Revista GEMInIS; Vol. 15 Núm. 1 (2024): Narrativas de IA: tendencias en la producción audiovisual - Parte 2; 44-53
2179-1465
reponame:Revista GEMInIS
instname:Universidade Federal de São Carlos (UFSCAR)
instacron:UFSCAR
instname_str Universidade Federal de São Carlos (UFSCAR)
instacron_str UFSCAR
institution UFSCAR
reponame_str Revista GEMInIS
collection Revista GEMInIS
repository.name.fl_str_mv Revista GEMInIS - Universidade Federal de São Carlos (UFSCAR)
repository.mail.fl_str_mv ||revista.geminisufscar@gmail.com
_version_ 1814253129773875200

Registros relacionados