AI for Fitness

Yes, but the quality depends entirely on how much context you give it. A plan generated from "write me a workout plan" is generic to the point of uselessness. A plan generated from your current fitness level, available equipment, days per week, specific goals (fat loss versus strength versus athletic performance), any injuries or limitations, and how much time you have per session is substantially more useful. Give Claude all of that in one prompt and ask it to explain the logic behind each training decision so you can evaluate whether it makes sense for you.

No. A personal trainer watches your movement, corrects form in real time, adjusts based on how you look and feel that day, and provides accountability through the relationship itself. AI can produce a well-structured plan but cannot see you move. Use AI for planning, programming variety, and answering conceptual questions about training. Use a trainer for form correction, injury prevention, and building the physical habits that AI cannot observe.

Yes, particularly for meal planning that fits your actual life rather than an idealised version of it. Give Claude your calorie or macro targets, your dietary restrictions, how many days you want to plan for, how much time you have to cook, your cooking skill level, and which ingredients you already have. Ask for a weekly plan with a single consolidated shopping list grouped by store section. The consolidation step is what makes it actually useful — most meal plans give you recipes but leave the logistics of the week to you.

AI is not a motivation source — it is a friction reducer. Most people do not fail fitness goals because of motivation. They fail because the plan is too complex, the environment is not set up, or the habit structure requires too many decisions. Use Claude to help you simplify your plan to its minimum viable version, identify the specific friction points in your current routine (too long, too early, needs equipment you do not have out), and design a version of the habit that removes those points. A smaller habit you do consistently beats a perfect plan you abandon.

Claude is useful for explaining the science behind sleep and recovery in relation to your training load, and for helping you audit your sleep environment and pre-sleep routine. Give it your typical schedule, training timing, and any sleep problems you are experiencing, and ask for specific protocol changes rather than generic sleep hygiene advice. It cannot access your data, so tools like Whoop, Oura, or Apple Health give you the data to bring to that conversation.

AI does not track automatically — you need to bring it the data. The most useful approach is a weekly check-in prompt: paste your training log, key metrics (weight, measurements, performance benchmarks), and how you felt that week, then ask Claude to identify patterns, flag potential over-training signals, and suggest adjustments to the coming week. This is more useful than tracking apps alone because you are getting interpretation of your data, not just visualisation of it.

AI can provide general information about injury rehabilitation principles and help you formulate questions for your physiotherapist or doctor, but it cannot diagnose injuries or replace clinical assessment. Use it to understand what you are dealing with conceptually, research what recovery typically looks like for your injury type, and prepare a list of specific questions for your next appointment. Never use AI to replace professional medical or physiotherapy advice for an injury.