AI running coach vs. a static PDF plan: what actually changes

Strip away the marketing and the difference is specific: a static plan is written once and can't read your training; an AI coach re-reads your actual runs and adjusts. Here's what that means in practice, and what it doesn't.

"AI coaching" is having a moment, which means it's earning both hype it doesn't deserve and skepticism it doesn't deserve. The way out of both is to get concrete about the mechanics. This post compares the two honestly. And yes, we build one of these things, so read with that in mind.

What a static plan actually is

A classic 10- or 12-week plan from a book, a race website, or a PDF is a well-tested progression written in advance for a broad audience. That has real strengths: the good ones are written by experienced coaches, the structure is proven, and nothing about them can glitch. For a runner whose life is predictable, a static plan can absolutely carry them to a finish line. Runners qualified for Boston off photocopied plans for decades.

Its weakness is symmetric with its strength: it was finished before your training started. It doesn't know your week fell apart, that your easy pace has improved, or that you've strung together three tired runs. Every adjustment is on you to figure out, which is exactly the judgment call beginners are least equipped to make.

What an adaptive coach actually does

An AI coach worth the name does three mechanical things a PDF cannot:

Reads your log before writing next week. The week ahead is generated from the runs you actually completed (volume, paces, how the week went) instead of a fixed script. Miss half a week and next week accounts for it automatically.
Updates your paces continuously. Training paces are recalculated as your fitness estimate changes, rather than staying anchored to whatever shape you were in when you started.
Answers questions in context. "Should I run on sore legs?" has a different answer in week 1 than in race week. A conversational coach that can see your history gives contextual answers, not generic ones.

// The short version

A static plan is a good progression that can't see you. An adaptive coach regenerates your week from the training you actually did and keeps your paces current. The adaptation is the product. If a plan doesn't change when your training does, the "AI" label is decoration.

What no app should claim to replace

Honesty cuts both ways, so here is what an AI coach, ours included, cannot do:

It can't examine you. No app can palpate a sore tendon or diagnose an injury. Persistent, sharp, or one-sided pain is a job for a physio or physician, and any coaching tool should say so out loud.
It can't feel the run. Effort data and pace curves are proxies. A human coach watching you run sees form breakdown an app can only infer.
It can't want it for you. No coach of any species has solved this one.

So which should you use?

If your schedule is stable, your goal is modest, and you enjoy managing your own training, a static plan from a reputable coach or book remains a perfectly good tool, and it's usually free. An adaptive coach earns its keep when your life is unpredictable, when you don't yet trust your own adjustment judgment, or when the thing you struggle with isn't knowledge but staying on the wagon. A plan that keeps fitting your real week is dramatically easier to stick to than one you're perpetually behind on.

That last sentence is the honest pitch for joggy, and for adaptive coaching generally. Not that it knows secrets a good book doesn't, but that it keeps meeting you where you actually are.

joggy gives general fitness coaching, not medical advice. Talk to a physician before starting a new exercise program, and get persistent pain checked by a professional.