When AI Makes the First Move

Theme

For most of the generative AI boom, the machine responded to a human first move. Now it is starting to make that move itself. In editing, that is consequential because a first cut is not merely assembly: it decides which moments sit beside one another, what gets emphasized, and what story begins to cohere. In May, the International Documentary Association examined what changes when filmmakers encounter their footage through AI-mediated sorting and selection, arguing that those systems can begin shaping which stories feel possible before an editor has formed an interpretation of their own. The new creative question is no longer just what happens when machines can make. It is what happens when the work arrives already carrying a point of view.

Signal

A July paper on human-AI co-creation, accepted to ICML’s 2026 workshop on Human-AI Co-Creativity, offers useful language for what should happen next: reduce repetitive friction, preserve reflective friction. The latter is the resistance that forces a creator to explore, reconsider, and discover something they did not already know they were looking for. That distinction gets closer to the real opportunity of a machine-made first cut. If the machine can take on the friction of sorting and assembling, it creates more room for the human to encounter a different kind - the friction of disagreement. What did the system make obvious? What did it flatten? Which strange connection did it miss? Where does instinct pull the work somewhere else? Adobe’s Firefly Quick Cut makes that shift concrete, using scene detection, shot selection, and audio analysis to assemble supplied footage into a first draft. The interesting possibility is not the draft itself, but what happens when it becomes provocation rather than precedent. As the first plausible version gets easier to produce, original sparks may increasingly begin in the collision between the machine’s interpretation and the choices a human makes against it.

Pulse

• Duck, Duck, Robot: Hugging Face unveiled Microduck, a $399 open-source, waddling, roller-skating duck you train with reinforcement learning — physical AI at a hobbyist price point. (Yes, we're buying one.)

• You Sound Familiar: Nicola Coughlan, Hugh Bonneville, and Matt Lucas are demanding UK laws against AI voice cloning, calling it "an existential threat to our entire industry."

• The Rough Draft Button: Gemini Omni 1.1 Flash extends AI video scenes to 40 seconds and adds a cheap 360p sketch mode before committing to 4K — the first plausible version, now a product feature.

We’ll continue these conversations at the Artist and the Machine Summit in LA this December, where Adobe joins as our Presenting Creative Partner to lead the way in the future of human and machine collaboration.

Now that everyone is back at their desks from summer, get your Early Bird ticket while they’re still available and we’ll see you there 😉 

If you’re creating something aligned with Artist and the Machine or you’d like to recommend work shaping this space, we’d love to see it for the chance to be featured. Reach out to us at [email protected].

Til next time,

Artist and the Machine.

xx