Human experts and Generative AI interaction: From real-world and real-life language learning diary-research app to a diary-learning app

Abstract

Generative AI (GenAI) could be beneficial for instructional designers, expediting the design of educational mobile apps. One needed app, currently absent, is a digital diary for second and foreign-language learners to record their outside-classroom activities. Given the research and learning merits of keeping such diaries, a diary-research app (Langy) was previously developed to measure learners’ real-life (RL) and real-world (RW) language learning. This study aims to provide design guidelines to transform this app into a learning app based on the interaction with ChatGPT during this design process and examine the human-machine interaction. The prompt submitted to ChatGPT4 described the diary-research app and asked for suggestions on how to transform it into a diary-learning app. ChatGPT produced 14 suggestions, upon which the researchers deliberated. The mutually created design suggestions emerging from this reciprocal interaction provide a blueprint for the new diary-learning app, adding a new purpose of promoting RL and RW learning (in addition to measuring) and relevant features such as goal settings, gaming rewards, and reflective prompts. Three processes were revealed during the complementary human-machine interaction: 1. ChatGPT’s offered generic suggestions, which the researchers adapted to more specific app requirements. 2. The researchers linked between the ChatGPT’s suggestions and added theoretical grounds. 3. The researchers, in some cases, altered or completely negated the ChatGPT’s suggestions. This study contributes to the research field of language-learning diaries. Moreover, it contributes to our understanding of how GenAI can be used as an instructional designer to assist in the design of educational mobile apps.