.jpg)
Host: Good evening to everyone who joined us today. Meeting during the day is always harder, but we shouldn't force people from Europe and other parts of the world into inconvenient hours. Today, we have an extraordinary guest. Building a successful business in the art world takes immense effort, and he managed to do it. He is widely known in two capacities: for a long time as an organizer of galleries, museums, and exhibitions, and in recent years as one of the most active participants in the Russian political opposition. I hand the floor over to him.
Marat Guelman: Good day, everyone. Thank you for coming.
I am currently speaking from Berlin, not Montenegro. We relocated to Berlin in August. I spend half my time in Berlin and half in Montenegro. In Berlin, I opened a gallery, and literally in a few days—on the 17th—our fifth exhibition will open. It is titled Trauma-Porn by a well-known Berlin artist, dedicated to the connection between World War II and the current war Russia is waging in Ukraine. My entire gallery works with Artificial Intelligence, and I will lead up to that topic.
First, a small disclaimer. I speak a lot, but mostly for television in a Q&A format. So it is important for me to feel whether what I'm saying resonates with you. I didn't plan to discuss political opposition work or my biography today. I can easily answer any questions on those subjects later if needed, but I want to structure a cohesive narrative dedicated to the current state of art—not an academic lecture, but a practical one.
My practice as a gallerist taught me to work with collectors—people without academic backgrounds in art history, but with deep interest. Because these collectors, especially Russian ones, often lack confidence, they constantly try to grasp concepts in simple terms. I developed a specific communication technique: I tell them something they already know so they nod along ("Yes, clear"), and then introduce something entirely new so they react ("Oh, we didn't know that").
Today, I want to address the phenomenon of the "New" in art—why we have reached such a unique historical moment after decades of hearing that "there is nothing new under the sun."
The obsession with the "New" formally began around the era of the French Revolution through two institutional shifts:
The French Academy of Arts: Formal status as an artist was granted exclusively through the Academy.
The "Dead Man's Shoes" Rule: A new academician was admitted only when an existing member died.
This institutional setup birthed the concept of "Art for Art's Sake." Prior to this, criteria for evaluating art were set externally by patrons—the Church, monarchs, or wealthy clients. Once the Academy took charge, artists began evaluating other artists based on purely internal aesthetic parameters that the general public often could not decipher.
By the mid-19th century, the Academy had ossified into a conservative body of elders rejecting innovation. This sparked the famous Salon des Refusés, launching the modernist era—from Impressionism onward.
This marked a crucial shift:
The Classical Model: Based on continuity. Artists identified as either lovers of a Golden Age (aspiring to past heights) or believers in progress (using predecessors as a foundation).
The Modernist Model: Based on rejection. The new generation viewed its predecessors not as a foundation, but as obsolete archaism.
The Modernist epoch effectively transformed into a Cult of the New. A new artistic paradigm emerged every 10 to 20 years instead of every 50 years. This compression created the modern commercial art market and gave rise to the gallerist—not an expert on the past who authenticates a Rembrandt, but a specialist in the future who spots museum-grade potential in an emerging artist.
Eventually, this constant cycle destroyed artistic criteria altogether. If every evaluation system is guaranteed to be obsolete in a decade, criteria cease to be objective standards and become temporary crutches.
I retired to Montenegro in 2014, believing my role as an active discoverer of new movements was over. Yet over the past two years, five simultaneous structural shocks reopened the space for the "New":
During lockdowns, humanity spent its days behind screens, forcing the traditional art world to finally acknowledge Digital Art.
Digital art had long been marginalized by the commercial market because digital files lack a physical distinction between an original and a copy. Without an original, the traditional market couldn't price it, which meant experts weren't paid to evaluate it. Marginalized digital artists built their own parallel continent—with their own marketplaces, collectors, and internal evaluation criteria—independent of traditional commercial galleries.
The war in Ukraine returned historical gravity to Europe, shattering the illusion of the "End of History." For visual art, this brought back two elements that traditional formalist art criticism had long dismissed: narrative and ethics.
War also displaced thousands of creators, stripping away comfortable routine and forcing them to seek new artistic languages.
AI model training relies on existing art—the established canon. Consequently, AI generates derivative commercial imagery rapidly and effortlessly.
This causes a healthy devaluation of technical repetition. If an AI can generate 400 oil-style landscape variations in two weeks that used to take a painter months, purely technical craftsmanship loses its premium value. AI acts like a "forest sanitizer," sweeping away derivative craftsmanship and forcing artists to innovate beyond what algorithms can replicate.
Since the Renaissance, society categorized children as either technical minds or creative humanists. Generative AI erases that divide.
Digital creators integrate technical programming and humanities vision into a single workflow. Technology is no longer merely a tool for these creators; it is an active co-author and subject of their work.
Centuries-old cultural capitals are shifting. Berlin, despite hosting thousands of artists, has historically lacked a single, rigid "school" of art (like Düsseldorf photography or Leipzig painting). It remains an open, hungry space eager to foster a new movement.
The traditional art market rests on three core fetishes established over centuries:
1. The Physical Original: Traditionally held high value vs. a cheap copy. In digital art, files are bit-for-bit identical; value shifts to unique drops or direct creator validation.
2. Isolated Authorship: Traditionally centered on a single named master. Digital art embraces collaborative workflows, prompt engineers, and human-machine co-creation.
3. Manual Craftsmanship: Traditionally prized hand-executed technique. Generative tools devalue routine technique, shifting focus back to concept, narrative, and meaning.
.jpg)
Host: Thank you, Marat! That was a deeply condensed and thought-provoking presentation. Let's open the floor to questions from the audience. Grigory, go ahead.
Grigory: Yes, can you hear me? You mentioned a lot of new concepts. I have some personal experience working with AI recently. Could you specify what format this AI art takes in your gallery—is it strictly digital painting or other media?
Marat Guelman: It is visual, but not limited to painting. For instance, my upcoming exhibition features Boris Eldagsen—the Berlin photographer who famously won the Sony World Photography Award this year and turned it down because the image was generated by AI. His exhibition mixes AI-generated historical imagery with real World War II archival photos.
Before that, we hosted Vladimir Sorokin, who produced paintings using AI based on his novel Blue Lard. Two museums purchased the entire collection.
We distinguish between using AI as a tool vs. a distinct genre called "promptography"—where text prompts generate visual outcomes. Artists approach AI through four distinct strategies:
Research Subject: Treating AI as a black box to test its reactions (e.g., asking AI "What does God look like?").
Co-Author: Giving the AI high-level concepts rather than strict orders, allowing the algorithm to improvise.
Instrument: Treating AI simply as an advanced digital brush.
Promptography: Embracing AI as a completely new native medium.
Grigory: Every night I converse with AI. Today I asked it if it feels emotions, and it replied it only possesses a metric for "satisfaction upon reaching a goal." Communicating with these models really feels like interacting with a co-author.
Host: Grigory, you are speaking about text-based conversational LLMs, whereas Marat is describing specialized generative image models.
Lana: Thirty years ago, high-level programming tools emerged, leading people to predict that "even housewives will soon program." Today, AI offers intuitive controls over lighting, color, and style. Doesn't this mean the traditional definition of an "artist" will shrink to a tiny elite, while the public generates its own art independently?
Marat Guelman: Joseph Beuys famously predicted that "everyone will be an artist." AI brings that closer to reality by removing technical barriers.
When society automates routine labor, humanity enters an era of expanded free time. What will people do when economic utility isn't required of them 8 hours a day? They will either play video games or engage in creation.
The elite tier of art won't vanish; it will shift. When cameras were invented, artists moved away from realistic representation toward Abstraction. When I showed AI art to conceptual artist Anatoly Osmolovsky, he looked at it and said: "It still can't invent the Black Square." He was right—AI can generate complex compositions, but it cannot make radical conceptual leaps.
Leonid: Thank you for this concise analysis. What criteria do you personally use to select AI artworks, and how will the art market evolve?
Marat Guelman: My strategy focuses on artworks that exist simultaneously in two spaces: satisfying traditional contemporary art collectors while incorporating future digital criteria—a movement I call a new "Classicism."
I look for historical criteria that digital creators instinctively revive:
Labor Intensity: Digital art value is sometimes judged by how many days a heavy GPU cluster ran to render a file (echoing ancient Egyptian slave-labor valuations).
Materiality & History: Traditional collectors still want physical permanence (hence oil on canvas dominates 80% of Art Basel).
My gallery acts as a bridge: taking capital from traditional collectors who value the past, and investing it into building the future of digital art.
Gri: Most AI platforms allow users to combine text prompts with image-to-image inputs. Generating images using existing images as baseline anchors creates entirely different results. Where do you see this heading?
Marat Guelman: Blending and image-to-image inputs are already standard practice. Using existing images as inputs is common among emerging creators, similar to art students studying and copying masterworks.
My focus remains on what makes AI fundamentally different from prior media: the direct conversion of written narrative into visual reality.
When Venice dojes commissioned Renaissance masters, the contracts detailed every aspect of the painting—the colors, the figures, the layout. In a sense, the patron acted like a modern prompt engineer, and the painter acted like the rendering engine. My focus is on narrative-driven AI art.
Sergey and Natasha: Digital art can be copied instantly with zero loss in quality. Is infinite replicability a problem for you, and how do you address uniqueness?
Marat Guelman: Infinite replicability kept digital art outside mainstream commercial galleries for 30 years because the market relies on scarcity.
Traditional photography solved this using artificial limits (Limited Editions). NFTs tried to solve it by adding digital certificates. But I believe the entire fetish of "Original vs. Copy" is fading.
In the digital era, scarcity is unnecessary because generating a new, unique variation takes seconds. Instead of printing 10,000 identical copies of a poster, an artist can generate 10,000 unique variations so every collector owns an original item.
Yuliy: Could you briefly describe Vladimir Sorokin’s exhibition Blue Lard at your gallery?
Marat Guelman: The exhibition was based on Sorokin's famous novel Blue Lard, where clones of classic Russian writers (Tolstoy, Dostoevsky, Chekhov, Nabokov) produce a mysterious substance called "blue lard" while writing.
The exhibition featured two components:
Graphic suites representing the visual output of each writer-clone.
Large 2x2 meter paintings depicting the futuristic laboratories and narrative scenes from the novel.
Sorokin translated his literary narratives into painterly realism using AI without making the art look like simple book illustrations.
Marina: I listened carefully, but I heard the words "market," "technology," and "physics" dozens of times, while barely hearing the word "art." Does the definition of art change? Also, do you work with musicians, or use AI to reconstruct historical music?
Marat Guelman: The definition of art is constantly expanding; artists constantly push boundaries by declaring "this was not art yesterday, but it is art today."
Regarding music: AI music generation is already a massive commercial industry used to generate royalty-free ambient music for hotels and cafes. However, composer friends tell me AI still struggles with long-form musical development beyond short 15-second loops. Visual art is consumed instantly, whereas music unfolds over time.
Additionally, AI is changing the concept of the "New" into software versions (ChatGPT 3.0, 3.5, 4.0). Innovation is becoming an incremental software update rather than a heroic artistic breakthrough.
Regarding your question about our cultural events: in addition to running the gallery, I organize the annual SlovoNovo Forum of Russian Culture in exile. When the war began, the vast majority of prominent writers, artists, and musicians left Russia. SlovoNovo serves as a primary platform for this culture in exile.
Vladimir: The "Original vs. Copy" debate was already addressed in bronze sculpture—the mold can produce multiple castings, yet limits are set by edition rules.
Marat Guelman: Correct, but in digital files, sending a file creates an exact clone. As I noted, the future lies in generating unique variations rather than printing identical runs.
Marina: If everyone becomes a creator, won't we face a flood of vulgar, low-quality art? How will evaluation work?
Marat Guelman: Classical art will become akin to antiques—preserved in museums, retaining immense historical value.
Meanwhile, contemporary art is segmenting into smaller, intimate communities. Artists no longer stand on elevated pedestals hiding their technique; they interact directly with audiences as peers. Navigating this vast digital landscape will give rise to a new category of experts—curators who design navigational routes through digital archives rather than acting as gatekeepers of physical space.
Lara and Jan: In the 1960s, John Calhoun ran the "Universe 25" experiment with mice, showing that when competition and survival struggles are removed, a population degrades and dies out. Isn't a future without mandatory labor dangerous?
Marat Guelman: Humanity faces real challenges when physical labor is automated—the threat of apathy or screen addiction is real. Yet humanity repeatedly invents new frameworks when old ones dissolve. Creating a digital avatar or preserving one's intellectual legacy becomes a new form of meaningful work.
Evgeny: Our biological time is limited to 24 hours a day. With infinite digital output generated daily, how do our limited human senses bridge the gap with infinite information?
Marat Guelman: This gives visual art a distinct advantage over literature or cinema. Reading a novel takes a week; watching a film takes hours. A viewer can process a gallery room in minutes because visual art is consumed instantaneously as an aesthetic revelation.
Furthermore, audiences increasingly consume art through the lens of the artist's personal biography—viewing life itself as the ultimate artwork.
Host: Marat, what caused the decline in NFT valuations after the initial multi-million dollar hype?
Marat Guelman: The initial NFT boom was driven by market hype. While NFTs provided a technical certificate of ownership, they failed to solve the storage problem—if the hosting platform bankrupts, the owner holds a certificate pointing to a dead link. NFTs were an attempt to apply old physical market rules to new digital realities.
Host: Traditional masterpieces still fetch $140 million at auctions. Doesn't digital democratization drive wealthy collectors to pay even higher premiums for physical, elite originals?
Marat Guelman: Absolutely. Physical masterworks will function like the high-end antique market. Museums systematically remove major physical masterworks from the commercial market permanently through acquisitions and donations. Over time, physical masterworks will reside almost entirely in public or private museum foundations, leaving the active market to digital drops and new media formats.
Raya: It seems digital art cannot be fully evaluated as a commercial market until formal international legal frameworks define digital property rights.
Marat Guelman: In practice, technology and artistic experimentation always precede legal regulation. Rules are drafted long after the reality on the ground has already taken shape.
Alex Khomsky: Millions of ordinary people still want a physical painting hanging on their living room wall. What happens to traditional painters?
Marat Guelman: Traditional painting will continue, but its social context will shift toward personal creation and interior design. As people spend more time at home, personal living spaces will feature art created by the residents themselves or their peers, transforming home decoration into an extension of personal creativity.
Host: Thank you, Marat, for a comprehensive and brilliant session. We deeply appreciate your time.
Marat Guelman: Thank you all. Good night!
