Tube Paint Moved the Skill Instead of Removing It

In 1841 John Goffe Rand put oil paint in a tin tube. Painters no longer had to prepare their materials. They had to read light outdoors, under conditions that changed while they worked. One requirement replaced another. That is what tools do when they work.

What a painter did before 1841

A painter in 1820 ground pigment, mixed it with oil, and used it the same day. Extra paint went into pig bladders sealed with string. To get at it later, you pierced the bladder with a pin, squeezed out what you needed, and tried to reseal it. Bladders leaked. Paint dried. The process was inconvenient enough that painters mixed only what they expected to use in a session.

In 1841 John Goffe Rand patented a collapsible tin tube with a screw cap. Paint went in from the open bottom, the bottom was crimped shut, and what remained stayed workable because the tube collapsed as paint was used. Winsor and Newton licensed the design almost immediately. Within a decade the tube was standard.

The preparation that had always preceded painting was gone.

The tube did not make painting easier, it made it faster

What replaced the preparation was not easier. Painting outdoors, under light that moved and weather that changed, required faster decisions about color than studio work ever had. The difficulty shifted from preparing materials to reading a scene.

Jean Renoir, in his 1962 memoir of his father Pierre-Auguste Renoir, recalled the painter saying that without paint in tubes there would have been no Cézanne, no Monet, no Sisley or Pissarro, nothing of what journalists would later call Impressionism. The Linda Hall Library, which holds a Winsor and Newton tube from the mid-nineteenth century, notes that painters had been working outdoors for decades before the tube arrived, and that portable easels and paint boxes mattered just as much. The line from tube to Impressionism is not as clean as the memoir makes it.

What is clear is what changed once painters could leave the studio without the preparation problem. Working under moving light put a premium on speed and on judgment. Color relationships had to be read in real time. The skill of preparing materials did not transfer to that demand. A different skill was required, and the tube was what made the demand unavoidable.

Warhol ran a production line and it required skill

Gerard Malanga pulled most of the silkscreen prints that became Warhol's canonical work in the 1960s. In principle the entire process, from source photograph to signature, could occur without Warhol touching the work.

Warhol chose the subject. He chose the source photograph. He decided the color relationships across each frame. He could look at what Malanga produced and say whether it was right. The execution was delegated. The judgment was not.

That distinction is what made the Factory functional. Malanga could pull a print. He could not decide what the print needed to do. Warhol could, because his decade as a commercial illustrator had built exactly that capacity. The production line worked because the person running it understood what he was producing.

I discuss this period in the module on postwar art from Paris to New York.

Delegation requires knowing what you delegated

Warhol would ask visitors to the Factory for opinions on color choices. This was not uncertainty about his own work. It was a working method. A director asks a cinematographer whether a shot achieves what the scene needs. The director does not have to operate the camera. The director has to know what the image needs to do, and the judgment about whether it does that cannot be handed to anyone else.

Two things together made Warhol's delegation work: a clear sense of what the work was trying to accomplish, and the training to evaluate whether an output accomplished it. Without either one the delegation breaks down. The person at the end of the chain is no longer specifying and approving. They are hoping.

Where the Factory analogy breaks for AI retouching

Warhol delegated execution to someone who could execute. If Malanga produced something Warhol did not want, Warhol could see it, say so, and have it done again. The feedback loop between what was specified and what was returned ran through people who understood both ends of it.

A photographer using an AI retouching tool receives a result. Whether that arrangement resembles the Factory depends entirely on whether the photographer can evaluate what came back. If they understand what a skin tone adjustment is doing, what a color grade is assuming, which parts of the result reflect the image and which reflect a model trained on different material, then the delegation is functional. They can look at the output, decide whether it is right, and specify what to change if it is not.

If they cannot evaluate the result, there is no Warhol in the room. There is a person who applied a tool and accepted whatever arrived. The output may look right. It may look right in a way that fails when the light changes, or when images from the session are placed side by side, or when a client asks for one specific thing to change. The gap between applying a tool and understanding it is invisible at the moment of application. It shows up later.

This is not an argument against AI retouching tools. It is an argument about what has to come before them. The tube did not make the Impressionists. It found painters who already understood what they were looking at, and it freed them to look at it under harder conditions.

Tools that relocate skill versus tools that absorb it

The tube removed the skill of preparing materials. It exposed the skill of reading outdoor light as the new requirement. One category of knowledge became unnecessary. A harder category became urgent.

The silkscreen process removed the skill of hand-painting each canvas. It exposed the skill of specifying, proofing, and approving a production run as the new requirement. Warhol had that skill. The Factory worked.

A tool that appears to absorb difficulty entirely is a different thing. The paint by number kit removed composition, value relationships, and color decisions from the painter's hands. It produced a finished picture. It did not produce a painter. The person who completed it learned nothing about why it looked the way it did, because every decision that produced the look had been made in advance by someone else.

A purchased preset works the same way. An AI retouching tool can work either way. The output arrives regardless. The question is whether the person receiving it can read it.

The test is whether you could do it without the tool

Rand's tube was not a threat to painting, and no one seriously suggested painters should continue grinding pigment once tubes were available. The question is not whether to use the tool. It is whether the photographer using it understands what the tool is doing well enough that they could, in principle, produce the result without it.

Warhol could have painted each canvas by hand. He chose not to, because the production process was part of the argument the work was making. That was a decision, not an abdication.

A photographer who understands a color grade can apply it with a tool, adjust it precisely when it is wrong, and explain to a client what it is doing and why. The tool is an extension of their judgment.

A photographer who cannot evaluate the result is in a different position. The tool is not extending their judgment. It is substituting for it. The difference is invisible on a good day and shows up on a bad one.

The tube freed painters to work outdoors because they already understood what they were looking at. The Factory worked because Warhol already understood what he was making. The tool lands differently depending on who picks it up.

Where This Comes From

History of Photography, Module 08: Modernism, Dada, Surrealism, and the Birth of Fashion Photography

Modern Art, Module 05: Post-War, From Paris to New York

History of Photography, full course outline

Modern Art, full course overview