AI for Video, Not for Music

I have used AI for more than five music videos and a lot of promotional material. I do not use it to make the music.

That line is deliberate, and I can tell you exactly where it is and why.

The practical reason

I have no marketing budget. None. A decade of this has come out of my own account, and the money went to recording, mixing, mastering and paying features.

A music video at the level I can picture — the one in my head — costs more than everything else combined. So for most of ten years the honest options were: no video, or a video that does not resemble the idea.

Now there is a third option. For pennies on the dollar I can get an absurd amount of my actual vision, and then some. Not a compromise of the idea. In some cases more than I would have gotten with a budget, because the budget version would have required somebody to talk me out of the expensive parts.

The video for “Sauce That” won five awards — Best Music Video from both the jury and the audience, plus Best Comedy and Best Experimental. That is not an argument that AI is good. It is evidence that the tool was used for something.

Why it still feels creative

This is the part people assume is not true, and it is the reason I will keep doing it.

Directing one of these is a long argument with a machine about what you meant. You are making thousands of choices — framing, pacing, what is absurd enough, what is too much, what to cut. The taste is entirely yours and the work is real. It is closer to editing than to ordering.

If it felt like pressing a button I would not bother. It does not. Which is the whole test.

Where the line is

The music stays human, and the reason is completely selfish.

Writing and making the music is the fun part. It is the reward. It is the thing I have been doing since an iPad in the back of a bowling alley at twenty-two, and learning piano by ear so I could build what I heard.

Handing that to a machine would be trading away the only part I actually want to do, in order to go faster toward... what, exactly? The point was never to have an album. The point was to make one.

I said “not yet” when I wrote this down, because I try not to make promises about a technology moving this fast. But I cannot currently imagine wanting to.

The honest summary

Visuals were a wall I could not pay my way through, and now I can get over it for almost nothing while still making every creative decision myself. The music was never a wall. It is the reason for all of it.

Anywhere I can create freely and cheaply and it still feels like creating, I am going to do it. That is the whole philosophy and it is not more principled than that.