|
Who made this?
on machine-made pictures, the people who feel robbed by them, and this building's own walls · an essay from the house
So the question in the title is not rhetorical for us. It is the question a visitor is entitled to ask of any picture here, and the one this essay tries to answer honestly: what changes when a machine took part, and for whom. two true reactions There are two reactions to machine-made pictures that people often treat as opposites. Both are reasonable, and we want to give each its room. The first is grief, and it is mostly not confusion. The image programs learned from very large collections of pictures taken from the open web, most of them made by people who were never asked and are not paid. Illustrators found their own names working as style prompts. Some of the work that once paid for a life of drawing, the book covers and the concept art and the small commissions, is now done in seconds. When an artist says this feels like theft, the claim is not that the program copies pixels; it is a claim to have been part of something without being counted.2 Courts in two countries are still deciding what the law says about it, and the open questions are listed on the timeline. Whatever they decide, the loss of a livelihood is a loss, and the people who feel it are not wrong to feel it. The second is wonder, and it is mostly not naivety. Something that did not exist a few years ago now turns a sentence into a picture that most people cannot tell from a person's. People who could never draw use it to show what they meant. Some artists use it as one more instrument, the way earlier artists used the camera, the plotter and the engine Lovelace imagined. Zhou and Lee found that artists who adopted the tools made more and were favourited more, and that their best ideas got bolder even as the average got blander.3 To feel astonished at that is not to approve of everything around it. Many people feel both on the same afternoon. We do. what the label does Psychology has a narrower thing to add. When the same picture is called a person's in one condition and a machine's in another, people like it more as a person's, find it more profound, and would pay more for it.4 When they cannot see a label, they often cannot tell, and in one study of poems they preferred the machine's.5 It is tempting to call this a bias and stop. We do not think it is only that. Part of what a person values in a work is the act that produced it: the effort, the choices, the life behind the hand. The experiments show how much that story weighs. They cannot show that it should weigh nothing. The research page on how people respond to machine-made art sets the studies out. choosing, making and originating Three older answers to the question are still the useful ones. In 1843 Lovelace wrote that the engine "has no pretensions whatever to originate any thing" and could do only what it was ordered to do;6 on that view the author is whoever gave the orders, and the hard modern case is a program whose orders were learned from millions of other people's pictures. In 1917 The Blind Man defended Fountain by saying that whether its artist "with his own hands made the fountain or not has no importance. He CHOSE it."7 On that view, choosing and framing are enough, and the US Copyright Office now agrees in part: it protects a person's selection and arrangement of generated images, but not the images.8 And in 1859 Baudelaire feared that photography would trespass on "everything that has value only because man adds something of his soul to it".9 His claim was never about what machines could do. It was about where value comes from, and no better machine can answer it. what we owe a visitor [the house's view] We do not think the right response is to hide which is which. The tour already hangs one real painting in a quiet room, Moonlight, Strandgade 30 by Vilhelm Hammershøi, an empty room with a window and moonlight on the floor, and its label says exactly where the photograph came from.10 A planned wing of the tour, called "Who made this?", is designed on the same rule: each picture on its walls says what made it, and the drawn ones say they were drawn by script. That is our small answer to the question in the title: tell people what made a thing, as exactly as you can, and then let them look. Hawthorne wrote a story in 1844 about a watchmaker who builds a mechanical butterfly so fine that it seems alive. A child crushes it in his hand, and the watchmaker does not mind, because he has already had what the work was for: “When the artist rose high enough to achieve the beautiful, the symbol by which he made it perceptible to mortal senses became of little value in his eyes while his spirit possessed itself in the enjoyment of the reality.” We find that consoling and also not enough. The watchmaker made the butterfly himself. The question this century asks is what is left to the maker when the butterfly can be had for a sentence, and the honest answer is that we do not know yet. We would rather say so than pretend our walls answer it. in the tour: a wing planned for the tour, called "Who made this?", will have a room called “This Building” ··· notes
Sources
see also
|
e should begin with our own walls. Some of what hangs in this building was made by machines. The paintings and prints are photographs of public-domain works, taken and released by museums; the credits list every one. Almost everything else you see here, the buttons, the plates, the plans, the walls of the tour, was drawn by script: by programs, written for this site, that put down lines and colours according to rules. The music in the tour was made with Suno, a program that generates songs, and the credits say so.
![[welcome]](../img/btn_welcome.gif){kind=small} |
![[research]](../img/btn_research.gif){kind=small} |
![[collections]](../img/btn_collections.gif){kind=small} |
![[institute]](../img/btn_institute.gif){kind=small} |
![[about]](../img/btn_about.gif){kind=small} |
rev. Oct 24 2026