The SCORE Check
Your check for whether a response deserves your trust, and what to do when it doesn't.
Basics to Breakthroughs is a practical course for choral conductors who want to work more effectively, think more critically about these tools, and decide for themselves where AI belongs in their musical and professional lives.
You know AI isn't going away. You've probably tried ChatGPT, Claude, Gemini, or Copilot.
Sometimes it's astonishing. Sometimes it's confidently wrong. Sometimes you think, I could have just written this myself.
And underneath all of that are bigger questions.
That's what this course is about.
The goal isn't to remove yourself from the work. It's to become more intentional about where you belong in it. Every lesson runs through the same habit: AI produces a draft, and you make the call.
Organizing. Transforming. Summarizing. Generating possibilities. Finding patterns in information you already have.
Taste. Context. Relationships. Interpretation. Responsibility. Artistic judgment.
One payment. Everything opens the moment you enroll.
The course is built on a small set of frameworks that stay useful when the products change. They're the part you keep.
Five rungs for turning a vague request into one AI can actually do well, and a rule that comes with it: don't climb higher than the task requires.
Your check for whether a response deserves your trust, and what to do when it doesn't.
You bring the task. The AI interviews you. Together you climb the ladder.
The questions you run before you hand AI anything that matters.
Goal, context, capabilities, guardrails, approvals, done. The page you fill in before any system carries your work forward.
What to ask about any new AI product, starting with the job it was built to do.
Keep, build, explore, stop, share. What you'll do with all of this after the last video.
“Give me a rehearsal plan for this piece.”
You'll get something. It will be generic, because the model has to guess everything that matters.
“Plan a 45-minute rehearsal for my 38-voice SATB community choir. We perform in five weeks. The altos are insecure in measures 43 to 58, the tempo keeps rushing in the B section, and tonight's priority is the vowel on ‘alleluia.’ Here are my score notes.”
Now it's working with your ensemble, your timeline, and your ears.
“Before you finalize it, tell me what you're assuming about my singers, what you'd need to know to do this better, and one alternative sequence.”
Now it's showing you its reasoning, so you can decide.
The Prompt Ladder is Module 2, and every module is yours from the first day.
Turn a semester of rehearsal notes into patterns you can actually see.
Reverse-engineer a past concert to understand why the program worked.
Treat your music library as information you can search and plan from.
Pressure-test a concert theme before you commit to it.
Draft family and singer communication that sounds like you, from three things you've already written.
Set up one workspace for concert communication so you stop re-explaining your program every time.
Compare several ways to introduce a musical concept before you walk into the room.
Build a repeatable process for auditions, tours, recruitment, or concert logistics.
And sometimes the right answer will be: don't use AI for this.
Each module is about forty-five minutes of short video lessons, plus one assignment that leaves you with something you'll use.
How the course works, what it is not, and a short AI Audit of where AI already shows up in your week.
What AI is, what it doesn't have, why a neural network works like a choir, and why a musician was there at the start.
Why output looks generic and how to fix it. The five rungs, the SCORE Check, and the Meta Prompt, back in the rehearsal room.
You are already surrounded by data. Put your rehearsal notes, library, and programs to work, and see why the bigger opportunity isn't generation.
Privacy, fairness, transparency and accountability. FERPA and approved tools. Scores, texts and copyright. Where your judgment stops and governance starts.
When a chat should become a Project, how to make it sound like you, how to build a small tool and connect it, and how to hand work off with guardrails.
The information you've never been able to study, how to judge any new tool, why music generation is a different problem, and your capstone.
You don't finish with more AI knowledge. You finish with your own way of working with AI.
Self-paced · Lifetime access · No live sessions to schedule around
Most AI courses add ethics at the end. This one teaches it as a skill, in the middle of the course, where it shapes everything after it.
And you'll write your own AI Use and Transparency Statement, rather than borrowing someone else's blanket rules.
Module 4 is open from day one, not locked behind the first three.
I'm not an AI engineer teaching musicians. And I'm not a conductor who discovered ChatGPT last month.
My work has lived where choral conducting, teaching, business, and applied AI meet. So this course isn't built around what AI companies say these tools can do. It's built around what conductors actually need to do.
Beth Philemon · Founder, ChoirBaton
Six modules of short video lessons, about forty-five minutes each, plus Before We Start and a permanent Toolkit. One assignment per module and copyable prompts throughout.
All twenty-three handouts in one place, including a Tool Landscape that's updated every quarter so it stays current.
Everything opens the moment you enroll. No cohort and no schedule. Lifetime access.
One-time payment · Lifetime access
No. Curiosity and a willingness to try things is the whole prerequisite. The course is built for conductors, not technologists.
Claude is the main tool we work in, and its free tier is enough to start. The frameworks transfer if you'd rather use ChatGPT or Gemini, because the course teaches the architecture, not the brand.
There are copyable prompts in the lessons where they help. But the course teaches the skill of writing them, because a list goes stale and a skill doesn't.
Yes. It's built for directors in all of those settings, and Module 4 speaks directly to how data and approved-tools rules differ between them.
Module 4 covers exactly this: what to put into an AI tool, what to keep out, FERPA and the approved-tools rule, and how to write a transparency statement you can defend to a parent, a colleague, or an administrator.
Not yet, and you should know that before you enroll. Every lesson, handout, and assignment is in place. Video is being released module by module, and each one appears in your course the moment it's done. You pay once and get all of it.
Each module is about forty-five minutes of video plus one short assignment, on your own schedule, with no live sessions. What you build pays that time back in rehearsal prep and admin work you're no longer doing twice.
You do need enough understanding to decide how these tools belong in your work. Basics to Breakthroughs gives you that foundation, and a practical way to keep building from it.