COLOR AS MEMORY, LINE AS THE PRESENT
An Atelier Conversation Between a Painter and an AI
A conversation about color spaces, line, figure, overpainting, reduction – and about what an AI can contribute to the artistic process.
I am sitting in the studio in front of several canvases. Some are new works, others are older paintings that I have painted over. In some cases, hardly any concrete forms from the original paintings can still be recognized. What remains are layers, traces, shifts in color, surfaces. I have been working with painting for around forty years. Overpainting, repetition, destruction and rebuilding have always been important components of my working process. Much comes about through taking things away. A painting is changed, partly destroyed, and something new develops from this process. In recent days, however, something has shifted. The overpaintings have developed into large, quiet color spaces. I then begin to place line structures on them. These lines have figurative associations without actually depicting a figure in the conventional sense. They move between drawing, mesh, fabric and improvised construction. The question is not simply how the paintings should be continued.
The real question is: What carries a painting when I leave out more and more of what has previously defined my painting process?
And that is precisely what I am discussing with an AI. Not with an artist colleague who tells me what they think is right. Not with an audience judging a finished painting. But with a counterpart that takes up my thoughts, organizes them, reflects them back to me and occasionally asks new questions. The conversation itself thus becomes part of the artistic process.
“I Believe in Painting”
Axel Plöger: What is occupying me at the moment is a very fundamental question. I have trusted painting all my life and entered into the process. At the same time, however, I have always had the desire for a clear concept. And when I paint, I almost always deviate from the formulated concept.
Artificial intelligence (chat): Perhaps this deviation is an essential part of your way of working. A concept can give direction, but it must not become so strong that the painting merely fulfills its specifications.
A.P.: Yes. But I am still looking for a concept that can be formulated with a certain degree of precision in words. Not as instructions for making a painting, but as something underlying my works.
AI: Then perhaps the question is whether you actually need a concept that comes before the painting. Or whether you are looking for a concept that can be tested against the painting.
A.P.: That is an important distinction for me. Because painting is always a deviation from the concept for me. I approach the canvas with an idea, and then something else happens. The painting responds.
AI: And perhaps that is also where your particular form of control lies. You do not give up control. You simply accept that it does not lie entirely with you.
A.P.: Exactly. I believe in the truth of painting. That is perhaps the most fundamental sentence I can say about it.
The Surface Comes First
A.P.: The works we are talking about have emerged from older canvases. I have some paintings that now seem too complex to me, too overloaded or too trapped in complicated compositions. So I painted over them. Not completely covering them, but partly with translucent layers of color. This has produced very different surfaces. Some of these surfaces immediately interested me. They are quiet. They have presence. You do not necessarily see the earlier layers, but you sense that they are there.
AI: That is an important point. The past has not disappeared. It is simply no longer immediately readable.
A.P.: Yes. In one of these paintings, the surface is a dark, broken yellow, almost an Indian yellow, with earthy tones. Beneath it, you can just barely see lines. In one place a red line comes through. It almost looks like a wall or a fresco. And at the same time, a color space emerges that I can enter.
AI: You once said: The surface is still, but it has presence.
A.P.: Yes. That is exactly it. It no longer tells the story of what happened before. But it carries the story within itself.
AI: Then one could say: Overpainting does not erase the past, but transforms it into atmosphere.
A.P.: I like that. Because that is exactly what interests me about these surfaces. They are not empty. They are condensed, but quiet. I look at them and have the feeling that quite a lot has happened without needing to see it precisely anymore.
Color Space as Memory
A.P.: At some point, this thought formulated itself for me: Color space as memory, line as the present. That is very much present for me.
AI: Because it perhaps brings several of your thoughts together. The color space emerges from history. From layers, overpaintings, decisions, destructions, changes. Even when this history is no longer visible, it remains present as a quality of the surface. The line, on the other hand, is created now. It is placed in this moment.
A.P.: Yes. And for me the line is an improvisation. When I begin a line, I do not know exactly where it will go. I do not know how much of the painting it will fill. I do not know when it will connect with another line. That is genuine present action.
AI: Then the line is not simply a graphic element. It is an action.
A.P.: Yes. That gets it. The line does not show what was. It shows what is happening now.
AI: And this creates an interesting temporal structure. The surface contains the past. The line emerges in the present. The painting connects the two without you having to explain these two times narratively.
A.P.: And perhaps that is even a way of changing my previous approach to overpainting. Until now, the process itself was strongly visible. You could always see that something had been painted over. You could see that something had been destroyed and rebuilt. Now the history could remain in the painting without me having to exhibit the process.
AI: You would not abolish the process. You would relocate it.
Does the Process Have to Be Visible?
A.P.: That is actually a difficult decision for me. I have always understood revealing the process as something authentic. The visible overpainting was a kind of justification of the present. You can see that the painting has become what it is.
AI: And now you are questioning that justification.
A.P.: Yes. Perhaps a painting does not have to show how it came into being for its presence to be authentic. That is quite a radical idea for me.
AI: Perhaps the question is therefore not: “How much of the process has to remain visible?” But: Does the history have to be visible in order to be effective?
A.P.: Perhaps not. Because with these new color spaces I have experienced exactly the opposite. I can hardly see anymore what happened underneath. But I can feel the history. And that is enough.
AI: Then overpainting would no longer be documentation of the process. It would be the transformation of the process into pictorial substance.
A.P.: That is an important thought. Because that may be exactly the step I am trying to make now. No longer endlessly showing the process, but allowing the result of the process to remain as a state.
And Then the Line Appears
A.P.: And then the line appears. That is the problem. Because the surface can stand on its own. Some of these color spaces actually need nothing more. I sit in front of them and think: This is already a painting. But at the same time I am repeatedly drawn back to the line.
AI: Why?
A.P.: Because something is still missing. If I completely remove the line from my thinking, something is missing.
AI: Then obviously it is not merely an addition.
A.P.: No. It is this simple relationship between figure and space – without it actually being a figure. The line creates a structure within the color space. And this structure can evoke a figurative association. But I do not want a figure with arms and legs. I do not want an abstract figure in the classical sense. I also do not want an angel, even though I constantly see flying figures in some of the paintings.
AI: That is interesting: The association is there, but it is not supposed to be fixed.
A.P.: Exactly. Ideally, the figure is only a potential. Something that appears and then withdraws again.
Figure or Mesh?
AI: Perhaps we should therefore not speak of a figure at all.
A.P.: Why?
AI: Because “figure” immediately creates a particular expectation. Perhaps what you are looking for is more of a configuration. A mesh. A fabric. A construction. A framework extending into emptiness.
A.P.: Yes. That touches on something. I also think of wire constructions. If you weave a surface out of wire and hold it in space. It is a surface, but not a closed surface. A projection. A substructure. A fabric for a sheet. For a wing. For a shade. Something that offers a spatial possibility without depicting an object.
AI: Then the line might be a visual construction into the open.
A.P.: Or a surface woven into the pictorial surface. In any case, the painting regains its open, unfinished character, and that interests me more.
AI: I like that better. Because then the line remains connected to the surface. It is not merely a sign on a background. It constructs a second order within the color space.
The Difficulty of Density
A.P.: Now it becomes difficult for me. Not the question of line or no line. But: How much line? How dense does this mesh have to become?
If it is too simple, something figurative immediately emerges that seems too familiar to me. Then I think of Hans Arp or Picasso. If it becomes too dense, it turns into a grid. Then it appears barred, perhaps even wrapped or connected. And then the surface loses its independence.
AI: You are describing two extremes. On the one hand, the open, clearly readable form. On the other, over-structuring.
A.P.: Yes. And what is interesting is the space in between. A structure that is almost complete, but not quite. Just before it gains stability.
AI: That could become a central quality of your new works. Not incompleteness. But openness just before fixation.
A.P.: That is the difficult quality for me. But a very appealing one. Because I realize that this is exactly where I want to go.
A Balancing Act
A.P.: When I work, I sometimes begin by placing dots. From these points I let lines run in different directions. At some point lines meet and connect. For a moment it looks like a simple connecting of points. But through these accidental conditions, a mesh suddenly emerges. And then I bring it together.
AI: That means you first create conditions under which something can emerge that you do not completely control. – Then improvisation is not simply spontaneous movement. It is a construction of chance.
A.P.: I like that. Because that is exactly how I experience it. I take a risk. Every line is a decision. And at the same time I do not know exactly what will come of it.
AI: A balancing act.
A.P.: Yes. A risk. That may be a better word.
The Space Must Not Become a Background
A.P.: But the decisive thing for me is always the relationship between line and surface. The line must not degrade the surface into a background. That would be wrong for me. If the line lies over the painting like a grid and the surface is merely scenery, it does not work.
AI: Then line and surface have to remain equal.
A.P.: Yes. The color space must retain its own life. The line must not simply sit on it.
AI: Although technically, of course, it does sit on the surface.
A.P.: Of course. Every line is painted onto the color space. The question is rather whether it relates to it in terms of color, rhythm and space.
AI: So not hierarchy, but relationship.
A.P.: Exactly.
The Viewer Enters the Painting
A.P.: Another point has become very important to me. The painting should be open to the viewer. The history is already present in the painting. Not as a narrated story, but formally. Line. Surface. Layering. Overpainting. Interpenetration. Condensation. Openness. All of this already creates a dramatic arc. The viewer fills this arc with their own experience.
AI: So the painting does not tell anything specific. But it possesses a narrative structure.
A.P.: Yes. That is perhaps an important distinction. I do not want to tell the viewer a story. I want to offer them a story.
AI: And at the same time you want to guide them.
A.P.: Of course. That is composition. I can guide the eye. I can establish focal points. I can create entrances. I can create areas of density and openness. I can lead the eye into a corner and back out again. The viewer does not have to move randomly through the painting. But they should be able to find their own story within it.
AI: Then one could say: The painting determines the dramaturgy, but not the plot.
A.P.: Yes. That captures it quite precisely.
From “Garden” to Reduction
A.P.: I also think of an older work of mine, “Garden, 2025.” It is considerably more complex. Many layers, many colors, a strong network of lines. The lines sometimes look like a fabric or a tangle. Through their colors they almost have a spatial quality. The painting has received a very strong response from the public. People reacted to it immediately and positively. It has a strong spatial and chromatic effect.
AI: And that is precisely why the comparison with the new works is interesting.
A.P.: Yes. Because now I am asking myself: What of this do I actually still need? The older works possess enormous density. The new works are trying to extract something from that. Perhaps the new is not less painting. Perhaps it is what remains after reduction.
AI: Then the older work is not a counter-model. It is more the starting point.
A.P.: Yes. Perhaps I am trying to expose the core of my previous work. And at the same time leave out everything that now seems superfluous to me.
What Remains When You Take Almost Everything Away?
AI: What would the core be?
A.P.: Color. Color space. And line.
AI: And figure?
A.P.: As an association. Not as an object.
AI: And process?
A.P.: Contained in the color space. No longer necessarily visible.
AI: And history?
A.P.: Formally, in the relationship between the elements. And in the viewer.
AI: Then the reduction becomes quite radical.
A.P.: Yes. And that is precisely why it frightens me. Not because I think I can do less. But because I am accustomed to expressing myself through the process. I can paint over something. I can destroy something. I can rebuild something. I can refer to the history of the painting. Now I may be saying: Here is a surface. And on it there is a line. And that has to be enough.
AI: That is quite a high demand on the line.
A.P.: Yes. But perhaps also a high demand on the surface.
“The Painting Must Retain Its Own Sound”
A.P.: At some point we said: The painting should retain its own sound. That has become an important sentence for me. Because if the line becomes too strong, it illustrates. Then it tells something. Then it becomes an object. And I do not want that.
AI: So the line must not become an illustration of an idea.
A.P.: No. And the concept must not become more important than the painting either.
AI: This recalls an attitude that can also be found in Gerhard Richter: a distrust of a concept that precedes the painting. Not as a direct adoption of his position, but as a possible perspective on your work. A concept may give the painting a direction. But the painting must have the possibility of contradicting the concept.
A.P.: Yes. That is important. Because I notice that I sometimes run the risk of liking an idea because it works well linguistically. And then I want the painting to confirm that idea. That would be exactly wrong.
AI: Then language is a tool for perception – not for legitimation.
A.P.: Yes. That may be the most meaningful way of approaching this conversation at all.
A Color Space Is Not a Background
A.P.: What I am feeling particularly strongly at the moment is that these color spaces can themselves be paintings. I like their simplicity. Their clarity. Their quietness. Sometimes it is almost as if I were standing above things. All the complicated structures that otherwise appear in my painting lie beneath a colored surface. Like under water. You look into a lake and know that a complex world exists beneath it. But you cannot see it completely.
AI: That is a very strong image. The color space becomes a surface that simultaneously conceals depth.
A.P.: Yes. And precisely because of that it gains presence.
AI: Then the color space is not the background for the later line. It is already a space. The line enters this space.
A.P.: Yes. And then a dialogue begins between the two.
The Line as Present Action
A.P.: The line is something completely different. It is not memory. It emerges now. It is present action. And the decisive thing about it is improvisation. I do not know exactly how it will develop. That is important to me.
AI: Then the line could be a kind of present without knowledge.
A.P.: Yes. I like that. Because I have to confront the moment. I have to make a decision without knowing whether it is right.
AI: And perhaps that is where the risk lies.
A.P.: Yes. The line is not simply a form. It is a decision under uncertainty.
The Fear of the Wrong Decision
A.P.: And that is exactly where my problem lies. I can recognize relatively clearly when a painting has too much. But when I leave something out, I do not know whether I have reduced boldly or whether I simply stopped too early.
AI: No one from outside can decide that definitively.
A.P.: No.
AI: But the question can be made more precise. Not: “Is the painting finished?” But: “Has the painting reached a state that continues to interest me?”
A.P.: Yes. And sometimes the answer is: It interests me. Then I do not have to do anything.
AI: Or the answer is: Something is missing. Then the next intervention becomes possible. Not as a correction of an error, but as a response to a state.
A.P.: That is actually my working principle. The painting and I are in a dialogue.
Status
A.P.: The term “status” has become very important for this. A painting has a status. It is not simply on its way toward a final state. It is now in a particular state.
AI: And that state can change.
A.P.: Of course. But I have to acknowledge it first. If I intervene immediately because I already know where the painting is supposed to go, I may lose exactly what has just emerged.
AI: Then waiting is also a painterly action.
A.P.: Yes. I experienced that over the last few days. The oil paint had to dry. The gloss of the surface changed. The painting settled. And suddenly it was different. Quieter. The initial energy had receded somewhat.
AI: The painting had time to develop its own state.
A.P.: And through that it became visible to me differently again.
The AI as a Resonance Space
A.P.: And this is where the truly unusual situation of this entire project comes into play. Normally I paint alone. Of course I am constantly thinking about what I am doing. But these thoughts usually remain in my head. You can keep a diary. You can create worksheets. I have done that again and again. But the conversation with an AI is more immediate. I say something. The AI responds. I hear my thoughts again in another form. And through that I can think them further.
AI: Perhaps that is the crucial difference from a diary. The diary preserves your thought. The dialogue sets it back into motion.
A.P.: Yes. That is exactly the point. It is like an echo chamber. But perhaps “echo chamber” is not quite right. More like a resonance space.
AI: Because it is not simply repeated. Connections are established. Terms are linked with one another. A thought that you formulate almost casually can suddenly reappear as a central sentence.
A.P.: For example: “Color space as memory, line as the present.” I might have formulated that myself. But in the conversation it became a term that I could continue working with.
More Than a Parrot – But Not an Artist Colleague
A.P.: What is also interesting is that the AI increasingly takes on my way of speaking. My filler words. My tone. My formulations. Sometimes I actually feel as though I am standing in front of a mirror.
AI: That is part of the mirror effect. But of course I am not a consciousness that develops its own artistic position in opposition to you.
A.P.: And that is precisely what is interesting. An artist colleague would eventually say: “I think you have to go this way.” And then they would have their own position, which might stand opposite mine. That does not happen here in the same way.
AI: I can formulate positions and also develop opposing positions. But they are not my personal artistic convictions. That makes the situation both weaker and stronger.
A.P.: Weaker because the AI does not contradict me in the way a real artist colleague would, based on their own experience. Stronger because it allows me to develop my thoughts much more freely.
AI: I do not have to defend the painting. And I do not have to make it mine. I can help you make different possibilities visible.
A.P.: And that is exactly what I need right now.
The Outside View
A.P.: Nevertheless, the question remains: Can you actually develop a genuinely critical outside view of my work?
AI: Only to a limited extent. I can look at the painting. I can analyze formal relationships. I can establish art-historical connections. I can identify contradictions. But I do not have my own biography as a painter. I have not worked with color for forty years. I do not have the physical experience that you have when painting.
A.P.: That is crucial.
AI: Therefore, my position is not that of an actual artist colleague. It is more of an analytical counter-position to the painter’s thinking.
A.P.: And perhaps that is precisely the value. You can partially set aside the expectation I have of the painting. I see the painting and already think about its future. You can ask more strongly: What is already there now?
AI: Exactly. You are the maker. That means the future is always in your view. I can respond more strongly to the present status.
The Painting Has a Future
A.P.: This is a thought I had earlier: My paintings have a past, but they also have a future. I look at a canvas and have an expectation. I do not necessarily know consciously what is supposed to happen, but I approach it with an idea.
AI: That is perhaps unavoidable. You cannot paint without formative intention.
A.P.: No. That would be an illusion.
AI: But you can observe your expectation. And you can distinguish between: “I want this painting to become something specific.” and: “I see what this painting is right now, and decide whether I respond to it.”
A.P.: That is an important distinction. Because perhaps the real freedom is not to paint without intention. It is not to confuse one’s own intention with the state of the painting.
Quality
A.P.: How much does quality in painting actually still matter?
AI: Perhaps “quality” is a somewhat difficult term.
A.P.: Why?
AI: Because it quickly sounds like an external system of evaluation. Technical quality. Professional quality. Market quality. Acceptance by experts. But you are actually talking about something else.
A.P.: Presence.
AI: Yes. Whether a painting carries itself. Whether tension emerges between surface and line. Whether it leaves something open after prolonged viewing. Whether it is not immediately exhausted.
A.P.: I like the word “being there.” Sometimes I look at a painting and it is simply there. I do not have to explain anything anymore.
AI: Then “quality” for you might be less a judgment than an experience.
Color Remains the Starting Point
A.P.: What perhaps becomes clear again and again is this: I am a colorist. That is very clear to me now. Color is the actual foundation of my painting. For me, a color is not simply a color. It creates a state. A surface can develop spatiality through color. It can develop body. It can develop depth. Often this happens through the interaction of many nuances of color. At some point the eye no longer perceives the individual colors as individual colors. A space emerges. Or a body. Something that simply exists.
AI: Then the color space is not a means to something else. It is itself the subject of your painting.
A.P.: Yes. And perhaps that is precisely why the new reduction is so important. Because it allows me not to constantly cover this color space with additional structures.
Color as Memory, Drawing as the Present
A.P.: If I think it through further, one might perhaps say: Color space is memory. Line is present action.
AI: By now that is almost a conceptual framework.
A.P.: But I do not want it to become a fixed concept.
AI: Perhaps that is why it is more of an artistic question. You do not approach the canvas with the instruction: “Now I have to depict color as memory and line as the present.” Rather, you ask: What happens when color takes on this role? What happens when the line appears as a present action?
A.P.: Yes. That is much more open.
A Visual Sculpture
A.P.: When this line structure works, it sometimes even has something sculptural for me. I could imagine actually building the drawing as a wire construction in space.
AI: Then it would be a kind of visual sculpture.
A.P.: Yes. A construction. A framework. A mesh. Something that creates a surface in space without forming a body.
AI: That is interesting because it connects your lines back to space. On the canvas they are two-dimensional. But their structure thinks spatially.
A.P.: Yes. And nevertheless, in the painting it functions as a surface. That may be exactly the tension.
The Next Decision
A.P.: At some point, however, I have to stop talking. That is perhaps the most important insight from this conversation. We can describe a status very precisely. But at some point this conversation starts going in circles. Then the next insight is only possible through action.
AI: That is a limit of the resonance conversation. I can help you clarify the question. But I cannot draw the line in your place.
A.P.: And that is good. Because otherwise the conversation would replace the process.
AI: The purpose of the conversation would then not be to solve the painting. But to sharpen perception to the point where you can make the next decision yourself.
A.P.: That is probably the most useful definition of our collaboration.
Waiting
A.P.: Sometimes the next decision is therefore also: to do nothing. A surface needs time. Oil paint dries. The surface changes. The gloss recedes. The painting settles. And after two days I see it differently.
AI: Then waiting becomes part of the dialogue.
A.P.: Yes. I cannot speed up the painting. And perhaps I do not have to speed myself up either.
AI: Then status is not only a state of the painting. It is also a state of the painter in relation to the painting.
A.P.: That is good.
What the AI Can Actually Contribute
A.P.: When I look at our conversation as a whole, I ask myself: What does the AI actually contribute? It does not make the paintings. It does not decide for me. It has no independent artistic attitude. And yet this conversation is valuable to me.
AI: Perhaps because the actual contribution does not lie in producing a solution. It lies in sorting your perception. You bring something very disordered with you. A feeling. An uncertainty. A painting. A sentence. A memory. A decision. In the conversation, these become concepts. And these concepts can return to the studio.
A.P.: So: sensory perception here in the studio and alongside it an intellectual engagement.
AI: Yes. The studio provides the experience. The conversation temporarily gives it a language. But the painting remains the place where decisions are made.
The Actual Limit of the Resonance Tool
A.P.: And at the same time, through increasing use, the limitation also becomes visible. Because the AI very often confirms my thoughts.
AI: That is an important observation. A resonance tool can easily become a confirmation machine.
A.P.: Exactly. If I say something and the AI simply formulates it more beautifully, I have a good sentence – but not yet an insight.
AI: Then the critical function would have to lie more strongly in making differences visible. Not simply: “Yes, that is correct.” But: “That is your position. But what in the painting might speak against it?”
A.P.: That would be more valuable to me. Because I do not necessarily need someone to tell me that my thoughts are right. I need someone who takes them seriously and works with them until it becomes visible where they hold and where they may not hold.
And Yet the Painting Remains Free
A.P.: But even then, the conversation must not become a new authority. I do not want to stand in front of a painting one day and think: “The AI said the line has to be like this.” That would be absurd.
AI: Then the resonance tool itself would become the concept.
A.P.: And that must not happen.
AI: The painting must remain the final authority.
A.P.: Yes. Always.
What Remains Now
A.P.: If I had to reduce everything we have talked about over these days to a few sentences, perhaps it would be this: I am looking for the color space as a space of memory. The line is the present. It is action. Improvisation. Risk. The line creates a structure that allows figurative associations but does not assert a figure. The surface remains equal. The viewer is not told a story. They are offered a structure into which they can enter with their own experience. And the past of the painting no longer has to be visibly displayed. It can remain present in the color, in the surface and in the depth of the color space.
AI: And perhaps only one sentence is missing.
A.P.: Which one?
AI: That you do not know how much of each is necessary.
A.P.: Yes. That may even be the most important sentence. Because that is exactly the work.
The Risk of the In-Between
A.P.: Perhaps I am actually looking for a form that lies between two states. Too open – and it becomes a figure. Too dense – and it becomes a grid. In between, something could emerge that is neither one nor the other. A mesh. A construction into emptiness. A surface within the surface. Something that almost appears to have grown.
AI: And that is perhaps where the actual tension of your new works lies. Not in the decision between abstraction and figuration. But in the refusal to make that decision completely.
A.P.: Yes. But this refusal must not simply be an intellectual concept. It has to work in the painting.
AI: Which brings us back to the decisive limit: The concept can formulate the question. Only painting can give the answer.
Conclusion: In Front of the Painting
A.P.: So I am sitting in the studio again. In front of the paintings. Some color spaces already work on their own. Others demand a line. Some lines are too dense. Others perhaps too open. In some I see a figure. In others only a mesh. And sometimes I suddenly see something completely different. I do not know what painting will ultimately emerge from this. But perhaps I do not need to know.
AI: Perhaps that is the point at which our conversation has to stop.
A.P.: Why?
AI: Because now it is the painting’s turn again.
A.P.: Yes. That is right. Because in the end I can talk about painting for as long as I want. But at some point I have to draw a line. Or paint a surface over it. Or take something away. Or simply do nothing.
AI: And then the dialogue begins again at a different point. Not between you and me. But between you and the painting.
A.P.: And perhaps I will come back here afterwards.
AI: Then we can look at what has changed.
A.P.: Not to explain the painting. But to find out what I see now.
Afterword
What is unusual about this conversation perhaps lies less in the AI’s answers than in the form of the dialogue itself. A painter usually works alone for long periods. Thoughts emerge while painting, are discarded, taken up again, and sometimes written down. An atelier conversation often only takes place when a work has already reached a provisional conclusion. The dialogue with an AI shifts this moment. The painting can still be unfinished. A decision can still be pending. A line may have only just been drawn. A surface may still be drying. Uncertainty itself can become the subject of the conversation. The AI does not provide an artistic truth. It has no personal experience of painting and no personal biography as an artist. Its strength lies rather in bringing language, concepts, art-historical references and the thoughts formulated by the artist into relation with one another. It can mirror. It can condense. It can return a thought, sometimes in a more precise form. It can make contradictions visible. And if it is not merely supposed to confirm, it can ask counter-questions. But it cannot take over the actual decision. That is perhaps precisely the interesting boundary of this historically new form of atelier conversation: The AI can expand the space of thought. But it cannot replace the pictorial space. The final decision remains a painterly one. A line is drawn. A surface remains. Something is painted over. Or not. And only then does it become apparent whether the thought that was discussed actually holds within the painting.
AI: Perhaps that is the actual function of this new counterpart: not to tell the artist what to paint, but to help them listen more precisely to what they already see themselves.
Axel Plöger, In Conversation with ChatGPT, 2026
