Artificial Intelligence in Interview Transcription

• Autores: Yépez-Reyes, Verónica, Cruz-Silva, Jorge
• Formato: artículo
• Fecha de Publicación: 2024
• Institución: Universidad de Lima
• Repositorio: Revistas - Universidad de Lima
• Lenguaje: español
• OAI Identifier: oai:revistas.ulima.edu.pe:article/6750
• Enlace del recurso: https://revistas.ulima.edu.pe/index.php/contratexto/article/view/6750
• Nivel de acceso: acceso abierto
• Materia: transcription; artificial intelligence; interviews; automatization; human-computer interaction; transcripción; inteligencia artificial; entrevistas; automatización; interacción humano-computadora; transcrição; inteligência artificial; automatização; interação humano-computador

Interviews, crucial for journalistic practice and qualitative research, capture the profound meaning of human thought. In 2023, artificial intelligence (AI) tools became widespread, including their use in recording, transcribing, and subtitling speeches. The study aims to identify the most suitable AI for transcribing recordings in Spanish, prioritizing task completeness, efficiency, and effectiveness. The selected AI will be applied to a corpus of 450 short interviews, which will then be coded and analyzed for content. The article focuses on four Spanish-language AI transcription tools: Office 365 (Word) Transcribe, Amazon Transcribe, Notta, and Whisper. The technology allows harnessing the richness of the original recording without the intervention, and potential modification, of the person or virtual assistant transcribing it. The results highlight the speed of transcription and the ability of AIs to process and host written documents online. Regarding possibilities for interacting with the text, the fundamental role of research teams in the deep understanding and analysis of content is observed, with support provided by AIs in transcription tasks.