Why are creators so attached to how much they control a work?
The question is especially sharp in code-based generative art and in making with AI. A lot of arguments end in the same place: if some parts of the work are not under my control, is it still my work?
Keep going, though, and control starts to look like the surface. What a creator cannot let go of is a very concrete kind of feedback in the making.
When I make a move and the work changes at once, I know I caused that result. Then I see the change, and decide what to do next. Intention, action, result, adjustment — joined together they form a dense feedback loop.
The resistance at the tip of a brush, turning a parameter in music, moving a few frames in an edit, swapping a word while writing: these details may never be visible to anyone else. For the creator they matter. Each adjustment confirms the same thing: I know why it became this way, and I can feel the relation between my action and the result.
That relation can bring a kind of satisfaction at the level of the nervous system. It may be pleasure. It may only be a plain felt sense: I am acting on this thing.
So many creators say their first audience is themselves. I don’t think that sentence is quite accurate. The creator does not wait until the work is finished, then sit down to watch what they have expressed. From the first second of working, they are operator and perceiver at once.
The work here is a bit like an interface. I do something to it; it feeds the change back to me; I feel that change, then answer it. In that loop I am making the work, and through the work I am perceiving myself.
This also explains why AI generation feels alienating to some creators. The problem is not only how much work the AI does in their place. The more specific change is that a dense sequence of action and feedback is compressed into a few rounds of input, waiting, and selection. The finished work may match my taste completely, but in the body and in the mind I have not gone through enough of “I made it become this.”
If that is so, what a creator needs may not be total control, but causal feedback of sufficient density. Those two things are different.
I don’t have to know where the next line will appear, and I can accept a shape the algorithm gives that I had not imagined. As long as I can still set the rules, draw the bounds, judge the result, decide what to keep, and know when to stop, the relation between the work and me does not have to disappear.
Code-based generative art has been dealing with this all along. The artist may not control every concrete result, but they control the conditions that let results happen. Randomness can exist. Accident can exist. The question is whether the uncontrolled parts, in the end, break the artist’s aim.
So what I think needs asking now is not “can I accept losing control,” but: after I give up some control, can I still tell whether the work has met my aim?
If the answer is yes, that portion of control may never have been a necessary condition for the making to hold.
Another very personal question shows up here: how high is a creator’s intolerance for unexplained residue?
Some people cannot accept parts of a work they cannot explain, cannot trace, or cannot confirm the source of. Even if the result is good, they still feel it is not theirs. Others can accept chance, collaboration, tool bias, algorithms, and misreading. As long as the result still lands inside their aesthetic judgment and their aim, they will claim it.
The amount of uncontrollability each person can tolerate is different. The higher that intolerance, the more they need to confirm, explain, and take responsibility for — and the harder it is to accept the black box in generative AI.
So a lot of argument about whether AI counts as making looks, on the surface, like a debate about authorship. Mixed into it are different tolerances for what cannot be controlled. Some people weigh the micro-process heavily. Others care more about aim, bounds, judgment, and what to keep or cut. These people already have different answers to “why this work belongs to me.”
I now tend to take control apart: what is the work’s aim, which conditions must I decide, which concrete results can be left to chance. That may be more practical than trying to master every detail.
A creator does not have to control the whole process of generating a work. They do have to know what they are willing to accept, and they have to be willing to bear what is left at the end.