Prepared for what lies ahead in tourism
Stay up-to-date with current, independent statistics and information on tourism in the Alpine region.
Users of AI-based training plans show significantly higher scores for confidence, attitude, usefulness and intention to use
Experts describe AI-generated training plans as easier to understand and more concise than human-generated training plans
Professional trainers are often no longer able to distinguish between AI-generated training plans and those developed by humans
AI-generated training plans received greater trust than human-generated training plans
Abstract of the study
Aims:Large language models are becoming increasingly important tools in everyday life, including the context of training and sport. However, the extent to which recreational athletes actually trust AI-generated training plans and how trust in these technologies differs between users and non-users has not yet been investigated. Furthermore, there is a lack of information regarding the current quality of such AI-generated training plans. The aim of this project was to investigate how users and non-users differ in their trust in these technologies and to assess the quality of AI-generated training plans.
Results:In the sample, 54% of participants trained with a structured training plan, 25% of whom used AI-generated training plans. Users of these AI-based tools showed significantly higher levels of confidence in these technologies compared to non-users . The quality of results from large language models has now reached a level where even professional trainers are often unable to distinguish whether a training plan was AI-generated or created by a human expert. This suggests that AI-generated training plans may meet the standards of plans developed by experienced coaches and thus represent a viable option for athletes seeking guidance for their training.
Key findings of this study
Increased trust among users: Participants who already use AI-based tools show significantly higher levels of trust, attitude, usefulness and intention to use these technologies compared to non-users .
High AI quality and indistinguishability:The quality of the results of large language models (LLMs) has reached a level where professional trainers often cannot tell whether a training plan was created by an AI or a human expert.
Higher confidence in the AI plan: The AI-generated training plan presented in the study received significantly higher overall confidence (3.73 ± 1.1) from participants than the human-designed plan (3.36 ± 1.1).
Better suitability for beginners from an expert perspective: Experts agreed that the AI-generated training plan was better suited for beginners, as it was judged to be "easier to understand" and "more concise" than the human-generated plan.