Using a running training plan generated by ChatGPT without human adjustment is risky, as these models lack progressivity, ignore recovery cycles, and propose dangerous workloads that can lead to injury.

noli's read

The analysis is based on the fundamental principles of sports physiology, specifically training load management and periodization. Research, including meta-analyses on injury prevention in running (e.g., Nielsen et al., British Journal of Sports Medicine), confirms that increasing running volume too quickly is the primary risk factor for injury. Language models like ChatGPT do not possess real-time biometric sensors or a nuanced understanding of individual physiology, which explains why they struggle to structure supercompensation phases or deload weeks, which are essential for adaptation. While AI can provide a theoretical structure, the lack of dynamic personalization is a critical issue. The creator rightly emphasizes that training is not merely a series of sessions, but a balance between stress and recovery. The exaggeration here is not in the observation of danger, but in the excessive trust one might place in a generalist tool for a demanding athletic practice.

the sources

1. 1Gabbett TJ 2016
2. 2Carey DL, Blanch P, Ong KL, Crossley KM, Crow J, Morris ME 2016