Bridging traditional and AI-assisted simultaneous interpreting: empirical insights for curriculum design

As artificial intelligence (AI) technologies increasingly permeate the field of interpreting, it becomes imperative for interpreters to adapt and develop the requisite skills to excel in this evolving landscape. While AI-assisted simultaneous interpreting (SI), specifically through the use of automatic live captioning tools demonstrates potential to enhance rendition accuracy, few empirical studies examine the pedagogical approaches specific to AI-assisted SI training. This study aims to address this gap by engaging 27 interpreter trainees in four SI tasks: SI without text, SI with text, sight translation (ST), and AI-assisted SI. Utilising correlation, regression, and path analyses, the results indicate that performance in traditional SI modes contributes significantly to AI-assisted SI performance. Our analyses reveal that SI without text and SI with text skills transfer distinctively to AI-assisted SI performance, with both modes contributing essential yet complementary competencies. Based on these insights, we advocate for a dual-focused training approach that leverages traditional SI competencies while incorporating specialised AI-specific modules to address unique challenges in AI-assisted interpreting. This study provides empirical evidence to inform curriculum design that bridges traditional interpreter training with emerging technological demands, preparing interpreters to work effectively in technology-augmented settings.