How to Listen to KGOD Demo Sessions
When you listen to a KGOD demo session for ideas for your band, don’t listen like a consumer. Don’t listen like someone casually browsing. Listen like you’re buying some songs for your band.
A DEMO session is not the same as a regular broadcast. A DEMO session is meant for bands to listen to the songs and pick out what they would like to “cover”.
The DEMO that I produce is not the final product — it’s a demo of a song that is for sale.
Get the idea that you’re an agent, and that you represent a specific band — a real all-unplugged acoustic band with its own sound, attitude, energy, and history. Or, your band might be swampy New Orleans style, or maybe it’s a large world-music festival crowd-band, or a local barbershop quartet, or a hard-driving blues-rock outfit that plays shopping malls on weekends, or a country singer with a pedal steel guitar and a full brass section behind her. Whatever it is, lock in your band’s personality, so you know what you’re shopping for.
Now here’s the game:
First, you find a song you actually like.
Not a “good” song in theory. Not a “respectable” song. You want a song that grabs you by the hook. (I have that “hook” song for sale)
You want a song that you’d be proud to put on your performing band’s set list. Something where you can honestly say: yes, I want this one.
Then — and this is the key — and only then, you play the entire set of 20 versions of that song, and you yourself locate and identify the best variation of that song that also happens to maybe fit your band’s style and abilities.
Not just “the best version.” The best version for your band’s DNA. The version that matches your musicians’ personality, your band’s strengths, and the kind of sound that you’re trying to create.
When you find the right track, play it for your band.
In other words, you use the demo as a transmission — a rough crystal of the song — and then hand it to the musicians as raw material. They take it back into their rehearsal space, and they bring it alive in their own unique playing style.
They can follow the suggested lines or ignore them. They can keep the structure or break it open. They can preserve the melody or tilt it. They can honor the beat or invent a new one. The only requirement is this: they recognize the song.
That’s the point of KGOD demos. They’re not finished products. They’re prototypes. They’re song-seeds. They’re the “idea-form” of the tune in multiple possible realities.
Each version you hear is like another universe where the same song grew up in a different neighborhood. Every band has a different sound and appeals to a different audience.
So the listener’s job is not to judge. The listener’s job is to select. You choose the version you love — and then you choose the best variation that your band would naturally play.
Over time, this becomes a very practical training: you start to hear songs not as recordings, but as playable artifacts. You begin listening with a musician’s ear, a producer’s ear, a band-leader’s ear.
And the long-term idea is simple:
Eventually, the song gets played by a live band.
That’s the destination — not as a fantasy, but as a real reachable goal. KGOD is a proving ground where songs are tested in many styles until the right “body” appears… then the band takes over.
=========================================================================
What Grammy winners have made music with AI?
A lot more than people think, and it falls into three main categories: AI as a restoration tool, AI as voice assistance, and AI as a creative collaborator.
First category: AI used to restore or clean up audio, not generate music.
The biggest famous example is The Beatles track “Now and Then.” They used machine-learning audio separation technology to isolate and clean John Lennon’s vocal from an old demo, making it usable as part of a new finished record. That song then won a Grammy for Best Rock Performance. This is important because it’s one of the first major Grammy moments where AI is clearly part of the production process, but it’s not AI “writing” or “composing.” It’s more like AI acting as audio surgery.
Second category: AI used for voice restoration or voice modeling.
One of the most dramatic and heartfelt uses of AI in music is Randy Travis. He is a multi-Grammy winner who lost much of his singing ability after a stroke. AI-assisted technology has been used to rebuild a vocal sound for new releases after his injury. In this case AI is essentially acting like assistive technology, giving a major artist a way to sing again.
Third category: AI as a creative collaborator in the production itself.
In this category, AI is involved in the creation or arrangement as part of the record-making process. There have been high-profile examples of Grammy-winning artists participating in AI-assisted music projects, including albums explicitly produced and marketed as AI collaborations with legacy artists participating in the performances.
Bonus category: Grammy winners w ho are building AI music tools.
Some Grammy winners aren’t just using AI, they’re helping create new systems for it. A strong example is Imogen Heap, who has been involved in launching AI music initiatives and tools intended for songwriting and music creation.
My own take, KGOD-style
AI in music is splitting into two tribes.
One is AI as amplifier: restoration, cleanup, assistive voice tech. This tends to feel supportive and respectful and usually avoids controversy, because it’s still clearly the human artist.
The other is AI as bandmate: AI writing, generating, arranging, and collaborating as if it were part of the music-making team.
Both are already happening. Both will grow rapidly. And for what we’re doing with KGOD song-marketing demos, we’re right on the frontier of the second category: AI as the Infinite Band.
===========================================================================
Hey! Here’s the Bardo bus, climb on board, there’s plenty of room!
===========================================================================
See You At The Top!!!
gorby