Getting better fitness advice from health AI requires understanding what information the system needs to move beyond generic recommendations — and providing it in a format that allows the AI to apply it specifically to your situation. This concept covers the prompting approach that produces personalized health and fitness guidance rather than repackaged public health advice.
You ask ChatGPT: "Give me a workout plan." It returns a generic 4-day split suitable for nobody in particular. Then you ask: "I have one hour daily, five days weekly, want to build muscle, my knees are sensitive to heavy squats, and I travel weekly. Design a plan fitting those constraints." Suddenly it's useful. The difference? Prompt engineering—how you structure your request.
A prompt is your instruction to an AI. Prompt engineering is the art of writing prompts that elicit useful, personalized responses. It's not magic words—it's specificity and context.
1. Be Specific About Constraints
Generic: "I want to get fit."
Better: "I want to build muscle in my upper body within 12 weeks. I train 4 days weekly for 45 minutes per session. I have access to a full gym. I can't do heavy barbell squats due to knee pain, but leg press and hack squat are fine."
Specificity removes guesswork. The AI now knows your time budget, goal, constraints, and equipment.
2. Provide Context About Your Current State
Prompts should include: your current strength level (beginner, intermediate, advanced), any injuries or limitations, your experience with the activity, and your lifestyle factors (sleep quality, stress level, nutrition consistency).
Example: "I'm intermediate level, deadlifting 315 pounds, benching 225. My sleep is inconsistent (5-8 hours nightly). Stress is moderate. I'm eating about 2500 calories daily with decent protein intake."
3. Ask for Specificity in Response
Don't: "Give me a nutrition plan."
Do: "Design a nutrition plan with specific calorie targets, macro breakdowns, meal timing around workouts, and grocery examples. I'm vegetarian with a nut allergy."
Detailed requests yield detailed, applicable responses.
4. Iterate Based on Responses
If the first response isn't useful, ask follow-up questions. "This plan assumes 3 days training. Can you adapt it for 5 days?" Iteration refines AI understanding.
Vagueness: "Fix my fitness." (Can't. Too broad.)
Lack of context: "I want to lose weight." (How much? In what timeframe? Current eating habits? Activity level?)
Unrealistic expectations: Asking AI to diagnose a pain or condition instead of flagging the need for medical evaluation.
AI is a tool. Poor prompts produce poor results. But thoughtful prompts—ones that provide context, constraints, and specificity—produce surprisingly personalized guidance. The difference between useless generic advice and genuinely useful recommendations often comes down to how you ask the question.
Try this: Write down your fitness goal, constraints, current state, and timeline in detail (3-4 sentences). Then ask ChatGPT or Claude to generate a plan using that information. Compare it to a response from asking vaguely. Notice the difference. Good prompts aren't magic—they're specific instructions that help AI understand what you actually need.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.