There is a particular kind of despair that arrives not with a bang but with a podium. You know the moment. The screen fills with flags, the crowd applauds on cue, and something in your chest begins the slow process of shutting down. You could argue with what’s being said. You could fact-check in real time, pull up the contradictions, build the case. But sometimes the body already knows what the mind is still debating: this is not a conversation. This is a performance designed to exhaust the people it cannot convince.
So you turn off the television.
What you do next is the more interesting question.
The Instrument You Already Own
What I do next is this: I open a piece of software, feed it a text, and hear my own voice sing back to me. Not a simulation of a generic voice. My voice — trained on recordings of myself, shaped by my own particular timbre and cadence — transformed into melody. I am singing to myself, essentially. Except I’m not. I’m doing something stranger and more deliberate: using a vocal clone to write spirit songs for the moments when the political world becomes genuinely unbearable.
The song in question here is my adaptation of “This Little Light of Mine,” a traditional African American spiritual that has been doing exactly this kind of work — keeping people spiritually intact in the face of hostile power — since at least the 1920s. Its scriptural grounding is Matthew 5:14–16: you are the light of the world. Its command is simple. Let it shine. But the freedom song tradition that gave the hymn its second life in the 1960s knew something deeper: that the act of singing was itself the resistance. Fannie Lou Hamer didn’t lead this song at SNCC gatherings because it was catchy. She led it because the voice, raised and sustained in the face of threat, was proof that the threat had not succeeded.
My adaptation doesn’t hide this inheritance. It extends it.
No kings can steal this flame / I’m gonna let it shine / Their thrones will burn in shame
This is not Sunday school theology. It’s the freedom song tradition updated for a specific political moment, the imagery of royalty and crowns doing work that would have been legible to both the psalm-writers and the Mississippi sharecroppers who sang in church basements while waiting to be arrested. The verse that follows — Not in their halls / Not under crowns / Not bowed beneath the banners / I walk — and still it glows — is structurally a refusal. Not a lament. Not a prayer for rescue. A declaration of continued walking.
What the Technology Actually Does
Here is where the vocal clone complicates things, and where I want to think slowly.
When you train a model on your own voice and use it to sing back to yourself, you have done something that sits at the intersection of several overlapping questions. The practical one: AI voice synthesis has become cheap, personal, and surprisingly good. The emotional one: there is something about hearing your own voice as an instrument — freed from the limitations of your actual singing range, your breath control, your self-consciousness — that produces a different kind of catharsis than simply listening to someone else’s music. The voice is the most intimate instrument we have. It carries identity in ways that a guitar doesn’t. To hear yourself sing something you couldn’t otherwise sing is, in some sense, to encounter a version of yourself that has more courage than the one currently sitting on the couch, waiting for the speech to end.
The philosophical question is harder. When Fannie Lou Hamer sang, the stakes of her singing were inseparable from her body, her presence, her Mississippi accent, her willingness to be in that room and raise that voice at that moment. The voice was testimony because it was irreducibly hers, risked by her, in conditions designed to silence her. The vocal clone produces a different kind of testimony. It is still my voice, still my words, still my adaptation of a tradition I understand and am taking seriously. But it is the voice without the risk of the moment. The flame without the wind.
I raise this not to diminish what I’m doing — I don’t think it diminishes it — but because the tension is worth naming. The question isn’t whether AI voice cloning is legitimate. It is. The question is what we’re doing with it, and what it requires from us in return.
Surviving Without Numbness
The political exhaustion that sends me to my software instead of my television is not a failure of civic engagement. It might be the opposite. Staying engaged with systems designed to wear you down requires active management of your own interior state. The people who stay in the fight longest are not the ones who white-knuckle through every speech. They’re the ones who have learned, as the tradition has always known, that there are forms of sustenance that are not negotiation and not argument but something more like prayer.
Music has always been this. The spirituals weren’t escapism; they were maintenance. You sang to keep your sense of self intact enough to keep working. The freedom songs didn’t replace organizing — they made organizing possible by reminding people, at the level of the body, that they were not alone and not defeated.
What’s new here is the specific texture of that sustenance. I train a model on my voice. I write my own adaptation of a two-hundred-year-old spiritual. I press play and hear myself — a better-singing version of myself, an unafraid version — declare that no throne can steal my flame. It is a deeply personal technology being used for a deeply old purpose.
There’s something right about that.
What the Turned-Off Television Means
I keep coming back to the act of turning off the television. Not ignoring the speech. Not pretending it isn’t happening. Deciding — consciously, deliberately — that I will not submit myself to something designed to grind me down, and choosing instead a form of sustenance that serves my capacity to remain clear-eyed and intact.
That is a small act of self-governance. It might be the most important kind.
The song says: They told me hush / I sang instead. The vocal clone is the means. The spirit is the point. And the point, as it has been since Matthew wrote it and the enslaved sang it and the sharecroppers carried it to the jails and back, is that the light does not ask permission to shine.
It just does.
I just do.
<iframe data-testid=”embed-iframe” style=”border-radius:12px” src=”
width=”100%” height=”352” frameBorder=”0” allowfullscreen=”“ allow=”autoplay; clipboard-write; encrypted-media; fullscreen; picture-in-picture” loading=”lazy”></iframe>
Tags: vocal clone technology, freedom song tradition, “This Little Light of Mine” adaptation, political exhaustion and creative resistance, AI voice synthesis personal use
