That Eric Alper

By Mitch Rice

There’s a specific kind of frustration every musician knows. You record something quickly—maybe a melody, maybe half-formed words—and in that moment it feels right. Not perfect, but alive. Then later you hit play, trying to figure out what you actually said.

And it’s gone. Or almost gone.

You catch pieces. A word here, a phrase there. The rest dissolves into guessing. You rewind. Again. And again. At some point, you start filling gaps with something new, and the original idea quietly slips away.

That gap used to be part of the process. Annoying, but normal. Now it’s shrinking, and not because artists suddenly got more organized. It’s because something else started listening.

Talking First, Writing Later

A lot of artists don’t really “write” at the beginning anymore. Not in the traditional sense.

They mumble. They hum. They talk over a beat like they’re explaining something to themselves. Words show up halfway, then disappear again. It’s messy, and that’s the point.

The difference now is that the mess doesn’t stay messy for long.

AI tools take those recordings and turn them into text almost immediately. Not perfectly, not always clean, but readable. Enough to work with. Enough to recognize what was actually there in the first place.

And that changes the order of things. Voice first. Structure later.

The Weird Relief of Seeing Your Own Words

There’s something slightly strange about seeing lyrics you didn’t consciously write.

You said them, sure. But you didn’t sit down and decide on each word. They just came out. Then suddenly they’re on a screen, lined up like they were planned.

Sometimes it’s disappointing. A line that sounded deep turns out flat.

Sometimes it’s the opposite. A throwaway phrase ends up carrying the whole idea.

That moment—when sound turns into something you can read—is where a lot of editing actually begins now. Not before, not during. After.

It’s Faster, But That’s Not the Point

Yes, it saves time. That part is obvious.

What matters more is the shift in attention.

Instead of spending energy trying to catch what was said, artists spend it shaping what’s already there. The focus moves forward. Less decoding, more deciding.

And because of that, the process feels lighter. Not easier exactly, just less clogged.

There’s no need to babysit every idea as it happens. You can let it run, knowing it’ll be there later in some form.

Not Everything Comes Out Right

AI doesn’t really understand what’s being said. It guesses. Pattern matches. Fills in blanks based on probability.

So it gets things wrong.

Names get twisted. Slang turns into something more “formal.” Emotions flatten out. Sometimes a line comes back completely different from what was intended.

And yet… those mistakes aren’t always useless.

A wrong word can sound better. A misheard phrase can open a different direction. It’s accidental, but it works.

Some artists don’t even correct everything anymore. They pick what feels interesting and move on.

The Middle of the Process Feels Different

There used to be a long stretch between recording something and actually working on it. A kind of dead zone.

Now that stretch is shorter. Almost gone.

You record, you transcribe, you tweak. Then maybe you record again. It loops quickly. No big pause in between.

That rhythm changes how ideas develop. They don’t sit untouched for hours or days. They evolve immediately, while they still feel fresh.

Short cycles. Quick reactions.

Sometimes that’s all a song needs.

Doing More With Less

Not everyone has a team. Or time. Or patience for the technical side of things.

This is where tools like AI music transcription quietly become useful. No setup, no complicated workflow. Just upload, wait a bit, and get something you can actually read.

For independent artists, that’s enough.

It removes one more barrier. One less step where things can get stuck. Especially for people who think in sound, not sentences.

And there are a lot of those.

Catching Things That Would’ve Disappeared

Some ideas only exist for a few seconds.

You don’t plan them. They just happen. Usually at the wrong time—when you can’t stop and write, when you’re too tired, when it doesn’t feel “important enough.”

Those are the ones that get lost most often.

Recording everything used to help, but only partially. If you couldn’t understand it later, it didn’t matter that it was saved.

Now it does.

Even a rough transcription can bring back enough of the idea to rebuild it. Not perfectly, but close enough.

And close enough is often all you need.

Sharing Without Explaining

Collaboration used to involve a lot of explaining. “Wait, listen to this part.” “No, go back a bit.” “That line right there.”

It’s slower than it sounds.

Text cuts through that.

You send a recording with a transcript, and suddenly the other person knows where to look. They don’t have to guess what matters. It’s already outlined.

They might still hear it differently, but at least they’re starting from the same place.

That makes feedback quicker. And sometimes more honest.

Not Clean, Not Perfect, Still Useful

There’s a tendency to expect tools like this to be flawless. Especially when they’re called “AI.”

But in practice, they’re not. And that’s fine.

Creative work isn’t about clean input and perfect output. It’s about reacting to what’s there.

A slightly messy transcript still does its job. It captures enough. It gives something to push against.

In a way, it matches how ideas actually form—uneven, incomplete, a bit off.

The Process Feels More Continuous

One thing that stands out after using this kind of workflow for a while is how little separation there is between stages.

Recording blends into writing. Writing blends into editing. Editing loops back into recording.

There’s no clear line where one part ends and another begins.

That can feel chaotic at first. Then it starts to feel natural.

Because that’s closer to how ideas actually move—back and forth, not step by step.

Still Human at the Core

Even with all this, the important parts haven’t changed.

Choosing what to keep. Deciding what a line actually means. Cutting something that doesn’t fit, even if it sounds good. That’s still on the artist.

AI doesn’t replace that. It just gets you to that point faster.

Or maybe just with less friction.

And that’s enough to make a difference.

Where It Settles

This isn’t some dramatic reinvention of songwriting. It’s smaller than that.

But also more practical.

Artists are still doing what they’ve always done—trying to hold onto ideas long enough to shape them into something real. The tools are just better at helping them not lose those ideas halfway through.

And sometimes, that’s the only thing that matters.

Data and information are provided for informational purposes only, and are not intended for investment or other purposes.