The difference between a plan that’s technically correct and one that actually works for you
The Polite Week 1 Problem
We recently analyzed the first weekly plans generated for three new athletes on Afitpilot. All three received technically sound programs. Proper periodization. Appropriate exercise selection. Reasonable RPE targets.
All three started with the same Week 1 theme: “Baseline Establishment.”
One athlete had 3 years of bodybuilding experience. Another was a trained hybrid athlete pivoting to powerlifting and Olympic lifting. The third was preparing for a Mont Blanc summit with years of climbing and running behind them.
All three got:
- “Focus on technique”
- RPE 6-7 caps
- “Establish movement patterns”
- Conservative loading
The powerlifter with 3 years under his belt was told to use an empty barbell for technique work. The alpinist preparing for a 4,808m summit was given “baseline assessments” for movements he’d been doing for years.
The plans weren’t wrong. They were just… polite. Patronizing, even.
Garbage In, Generic Out
Here’s what one athlete’s profile looked like:
Name: "Jason"
Goal: "muscle building"
Frequency: 4 days/week
Duration: 60 minutes
Equipment: (none listed)
Benchmarks: (none)
Events: (none)
Context: (none)
With this input, what can any system—AI or human coach—actually do?
It generated a perfectly reasonable hypertrophy program. 4-day split. Progressive overload. Deload weeks. All the boxes checked.
But it was indistinguishable from a Jeff Nippard PDF, a Juggernaut AI template, or any of the thousands of free programs on Reddit. There was nothing to work with.
No tension. No competing demands. No context about what makes this athlete’s situation unique.
Generic input produces generic output. This isn’t an AI problem—it’s an information problem.
What Makes a Plan Actually Useful?
Compare that thin profile to what we’d need to generate something genuinely personalized:
Identity & Context
- How long have they been training?
- What program are they coming from?
- Why muscle building—aesthetics? Sport base? Returning from injury?
Competing Demands
- Do they play recreational sports on weekends?
- Is their job physically demanding?
- Do they have young kids disrupting sleep?
- Do they travel frequently for work?
Training Reality
- What does a bad week look like? (Skipped sessions? Burnout? Injury?)
- What does a good week look like?
- When do they feel uncertain about training?
Current Capacity
- What are their actual numbers right now?
- What were they doing last month?
- What’s working? What’s not?
With this information, the system knows:
- Whether to start conservative or continue at their current level
- Which sessions are “fragile” (likely to be skipped)
- What interference patterns to expect
- When to push and when to back off
Without it, every athlete gets the same safe, polite, forgettable Week 1.
The Entry State Blind Spot
The biggest gap in most AI fitness tools—including ours until now—is the assumption that every athlete is starting from zero.
But athletes arrive in different states:
| Entry State | What They Need | What They Usually Get |
|---|---|---|
| True beginner | Conservative baseline | ✓ Correct |
| Trained, new to sport | Aggressive on known patterns, conservative on new skills | ✗ Treated as beginner |
| Returning from break | Ramp-up protocol | ✗ Treated as beginner |
| Transferring programs | Continuation with adjustment | ✗ Treated as beginner |
| In-season athlete | Maintenance or taper | ✗ Treated as beginner |
| Post-injury | Graduated reintroduction | ✗ Treated as beginner |
When a 3-year bodybuilder transitioning to powerlifting gets “Week 1: Baseline Establishment” with RPE 6 squats, the system has failed them. Not because the plan is dangerous—but because it’s insulting. It ignores everything they bring to the table.
The Real Question We Should Be Asking
Most onboarding flows ask: “What’s your goal?”
That’s the wrong question. Goals are easy. Everyone has one.
The right question is: “What makes your training complicated?”
That surfaces the tension:
- “I play soccer on Sundays so Mondays are always rough”
- “I travel for work one week per month”
- “I tend to overtrain and burn out every 6-8 weeks”
- “My left shoulder has been sketchy since last year”
This is the information that transforms a generic plan into a useful one. It tells the system where the friction is—and friction is where coaching actually matters.
Beyond Planning: The Decision Moment
Here’s the uncomfortable truth we’ve been wrestling with:
Afitpilot is very good at planning. We’re not yet good at arbitration.
We can generate a thoughtful macrocycle. We can periodize intelligently. We can select appropriate exercises and progression schemes.
But the moment that actually matters for most serious athletes isn’t the plan. It’s the decision.
“I have a plan. It says heavy squats today. But I slept 5 hours, my knee feels off, and I’ve got a big presentation tomorrow. What do I actually do?”
No plan survives contact with real life. The value isn’t in the document—it’s in the decision support when things get complicated.
A good plan with no arbitration is just a well-formatted guess. A mediocre plan with good arbitration keeps athletes training consistently and sustainably.
We’ve been optimizing the wrong thing.
What We’re Changing
1. Entry State First
The first question is no longer “what’s your goal?” It’s “where are you coming from?”
- Training consistently → No baseline week needed, continue at current level
- Returning from break → Ramp-up protocol, not patronizing baseline
- Transitioning sports → Aggressive on known movements, conservative on new skills
- Starting fresh → Current conservative approach (actually appropriate)
2. Surfacing Tension
We’re explicitly asking what competes for time and energy:
- Other sports
- Physical job demands
- Sleep disruption
- Travel patterns
- Stress cycles
And what failure modes look like:
- “I skip Thursday when work gets busy”
- “I tend to push too hard and burn out”
- “Life interrupts and I lose the rhythm”
This tells the system which sessions are fragile, which days to protect, and when to proactively suggest modifications.
3. Current Capacity, Not Just Goals
If you’re not a beginner, we want to know where you actually are:
- Recent lifts
- Run times
- What you’ve been doing for the last month
- What’s working and what’s not
This prevents the insulting “baseline week” for trained athletes and gives the system concrete numbers to work with.
4. Enrichment for Existing Users
For athletes who signed up with thin profiles, we’re adding a prompt:
“Your plan started conservatively. Want to personalize it based on your training history?”
One quick flow to capture entry state, recent training, and current benchmarks—then recalibrate the plan.
5. Inline Calibration
After completing (or skipping) sessions, a simple question:
“How was that? Too easy / About right / Too hard / Skipped”
“Too easy” triggers: “Tell us about your recent training so we can calibrate.” “Skipped” triggers: “What happened?” (surfaces failure modes in real-time)
The plan learns from execution, not just from initial setup.
The PMF Question
We’ve been asking ourselves who Afitpilot is actually for. The honest answer:
Not beginners. They need a plan, any plan, and consistency. The market is flooded with solutions for them.
Not people looking for motivation. That’s a different product.
Self-coached competitive athletes who already train 5-6x per week, already have structure, but don’t know when to override it.
Their pain isn’t planning. It’s execution under fatigue, competing demands, and real-life interference.
The question they’re asking, often silently:
“I have a plan. But today feels wrong. What do I actually do?”
That’s not vague. That’s a single moment. And it’s the moment we need to be useful for.
What This Means for Your Plan
If you signed up recently and got a conservative Week 1 despite having training history: we hear you. It’s not personal—our system just didn’t know any better.
We’re rolling out profile enrichment prompts over the coming weeks. Take 2 minutes to tell us where you’re coming from, and we’ll recalibrate.
If you’re signing up fresh: you’ll see a new onboarding flow that asks better questions. Not more questions—better ones. The kind that surface what makes your training complicated, not just what your goal is.
And if you’re the kind of athlete who’s ever stood in the gym parking lot thinking “my plan says X but something feels off”—that’s the moment we’re building for.
Not the plan. The decision.
We’re building Afitpilot for athletes who don’t need to be told to train—they need help deciding what to do when training gets complicated. If that’s you, we’d love your feedback. Reply to this post or reach out directly.
Leave a Reply
You must be logged in to post a comment.