A study asked ChatGPT-4 to create weekly training plans for elite swimmers and sprinters, then asked coaches and athletes to evaluate them.
Coaches considered many of the suggestions broadly usable. Athletes were noticeably more critical.
Problems with complex load management
[bild id="XX"]
Weaknesses appeared especially in moderate- and high-intensity training zones and in complex periodisation.
AI can therefore be a useful tool for ideas, structuring and simple planning tasks. At present, it does not replace the informed judgement of an experienced coach.
Where AI can already help today
Artificial intelligence is well suited to tasks in which information needs to be structured or variants created. Coaches can, for example, collect ideas for drills, summarise training documentation or generate first drafts of a weekly plan.
It becomes problematic when an individual training programme is derived automatically from such a draft. Without additional data, an AI system knows neither the athlete's actual fatigue nor technical problems, injuries or the dynamics of a training group.
Coaching expertise remains the key filter
The study therefore points to a sensible middle ground. AI can accelerate workflows, but it has to be checked professionally.
Especially for intensity control and periodisation, it is not enough simply to reproduce general training principles correctly. The load has to fit the individual, the current training state and the long-term plan.
For coaches, this therefore creates a new tool rather than a replacement. Those with sufficient expertise can use AI to develop variants faster or prepare data. The final decision, however, remains a professional coaching task.
SCIENTIFIC SOURCE
https://pubmed.ncbi.nlm.nih.gov/41668944/