Why AI gives you generic answers (and how to fix it in 2 minutes)

The needs analysis you'd never accept

Imagine a brand new assistant hands you a needs analysis for your team, and it opens like this:

Athletes benefit from strength. Speed is important. Consider conditioning.

You'd hand it back. Not because it's wrong. Because it's useless. It could describe any roster, in any weight room, in any state.

Now think about the last time you asked AI for help and got an answer that felt just like that. Technically correct. Completely generic. Nothing you could actually use with your athletes on Monday.

Most coaches conclude the same thing: “AI just gives generic answers.”

Here's the thing. That needs analysis wasn't useless because the assistant couldn't write. It was useless because they never watched your athletes train.

Your AI hasn't watched your athletes either. And it never will, unless you show it.

Context Before Prompt

This is the first framework I want to put in your hands, because every other AI skill you'll ever build sits on top of it.

Before you ask an AI for anything, tell it what it needs to know to answer like it works in your building, not like it read a textbook.

AI doesn't become valuable because it knows coaching. It becomes valuable because you know your program, and you can hand that knowledge over in about two minutes.

The difference between a generic answer and a usable one is almost never a smarter question. It's better context. Same tool, same question, completely different output.

The workflow: Program Card

Here's the two-minute version you can run today.

Open a note on your phone or a doc on your laptop and write out one short paragraph covering:

• Who you coach. Sport, level, roster size, training age. “Division 1 College Basketball, 15 athletes, most with 2-3 years of structured lifting” beats “college athletes” every time.
• What you're working with. Equipment, space, session length, sessions per week.
• Where you are in the calendar. In-season, off-season, three weeks from conference tournament.
• Your constraints. Shared weight room, 40-minute sessions, no assistant, whatever's real for you.
• Your philosophy in one line. What you believe about training this group.

That's your Program Card. Save it. From now on, it's the first thing you paste into any AI conversation about your program, before you ask a single question.

You write it once. You reuse it hundreds of times. Two minutes of setup that upgrades every answer you get for the rest of the season.

What it looks like in the real world

Say you just listened to a podcast episode on force plates and CMJ testing on the drive home. Good episode. Lots of ideas. Now what?

Without context:

“Summarize the key takeaways from this episode on force plates and CMJ testing.”

You'll get a clean summary. Countermovement jumps, asymmetry data, fatigue monitoring, force-time curves. Accurate, tidy, and no closer to your weight room than the episode was. You still have to do all the translation yourself, which is why most podcast notes die in your phone.

With a Program Card in front of it:

“College football, 105 athletes, off-season week 2 of 8, 60-minute sessions, 4 lift groups, Train 4x per week, Wednesday and weekends are off, 3 coaches on the floor. No force plates and no budget for them. I believe in simple, repeatable systems my athletes can run without me hovering. Here's the transcript. What from this episode actually applies to my setting, what doesn't, and what's the simplest version of the useful parts I could test with one group?”

Now the answer sorts the episode for you. It tells you the force-time curve talk doesn't apply, but the principle underneath it, using a simple jump as a window into how ready an athlete is to train, can show up in your room as a basic jump test tracked week to week. It flags what to ignore, what to steal, and where to start.

Notice what didn't happen: AI didn't write your program. You did that years ago and you're still doing it. What AI did was collapse two hours of podcast into one testable idea that fits your building. That's the job. Turning information into action is back-end work, and back-end work is exactly what you should be handing off.

The question barely changed. The context changed everything.

Where the coach stays in charge

One warning, and it matters.

Context makes AI's answers more relevant. It does not make them automatically right for your room. In fact, AI can even help you pressure-test the episode itself. Ask it to check the guest's claims against the research, and it will go do that homework too.

What it can't do is make the call. It can't tell you whether that jump test is worth five minutes of your Tuesday, or whether it's the right week to add anything at all, or what to do with the kid coming back from a hamstring pull whose numbers say one thing and whose face says another.

Relevance is a job you can hand off. Verification is a job you can hand off. The decision never is.

AI reads your Program Card. You read the room. That second part is coaching, and it's not in the workflow, because it can't be.

Treat every output as prep work from a smart assistant who's never met your athletes. The decisions were yours before AI showed up. They're still yours.

Your next rep

Before you close this email: write your Program Card. One paragraph, five bullets, two minutes. Then grab the last podcast, article, or clinic talk you meant to “do something with,” paste your card in first, and ask what actually applies to your setting.

Compare that to the generic summary you would have gotten. That gap is what this newsletter is about.

Hit reply and tell me what changed. I read every response.

See you next Sunday.