“Art is Dead”: A Rumination on the Future of Creativity in an AI-Curious World

In a studio with the Empire State Building looming across the street, Raul Gutierrez has set up a group of innocuous looking black boxes, each one with pinholes drilled on one side. These boxes are in fact the earliest, simplest type of camera, a camera obscura, and their purpose is to try and record time itself.

Gutierrez has had a lifelong obsession with time and how to represent it visually. “It started with long exposure photography, taking pictures of star trails and solar photographs,” Gutierrez explained. “In my teens I lost many elderly aunts and uncles. I would stare at old photos trying to see through the patina of age and imagine what those images might have looked like in person. In college I read a lot of John McPhee who often focuses on deep time in particular places. The funny thing is the longer the timespan, the more people and the things they’ve created disappear.”

His fascination continued as he contemplated mortality at the start of his adult life. “In my twenties I was again dealing with a spate of deaths of close family members and my projects became introspective. At one point I took a polaroid of myself every day because I was curious if I could see the grief in my own face. I couldn’t then, but I can now decades later." He started long term projects focused on photobooths, where he would make a series of photos with the intention of returning to the location for a photograph at exact time intervals (every 10, 20 years, etc.).

“Startups are all about time, or the lack of it,” Gutierrez says. He's used photography to even visually track the time he's spent developing companies. “In my previous startup I started an exposure the day we moved into our office, and finished it the day we launched. In that case 9 months."

With his latest startup, founded by Gutierrez in 2023, from the beginning he knew he wanted to work together in a studio space with a small team. “I love the friction between ideas and the pace. In startup world, the ticking clock in our heads often pushes us to wake up earlier, work later, and squeeze more minutes out of the day.” His pinhole cameras are a way of marking that time. Primitive and fragile, he fully expects many of them to fail. It’s a process, he realizes, that mimics the nature of building new technologies.

Gutierrez's focus is now all in on Picture Studio, a generative AI company developing a software that effortlessly translates your conceptual ideas into a visual reality within seconds. Since the algorithms don’t know what’s in your head, and because the same prompt might be visualized in hundreds of ways, current AI tools often disappoint. The technology aims to get you from the image in your head to the image on the page much faster, by giving artists a tooling layer that goes far beyond simple prompting.

What distinguishes the company vision from generative AI competitors is its focus: Picture Studio wasn't created to replace creative workers but rather to empower them. This tool allow artists to train models on their own styles and concepts so they can use it to supercharge their output, freeing them to work and complete projects faster than they ever thought possible. It’s an app truly built for creators.

Gutierrez’s mission is to help you instantly render a thought into a vision, nevertheless he appreciates the input that comes from process. Process is an inherent piece, in fact, of the act of creation, and how the Picture Studio product is coming to life.

Alice Zhang, a creative technology intern renders a model of herself in a few of the hundreds of styles Picture Studio is fine tuning

The team is building an extensive database of style experiments they meticulously track as they train the generative AI module to render images more granularly and accurately. The results of this work that was shown to me—the images you can create with Picture Studio—were distinctly refreshing compared to other Gen AI softwares out there. The archive reflects a rich and varied cultural context, compared to the dominant “future sci-fi,” “video game,” and let’s face it, sometimes “apocalypse-core” style results that can feel homogenous in the space. Rather than trying to mimic the look of specific video games or an artist’s style, Picture Studio’s goal is to offer artists a long list of mediums and historical styles in which to express ideas.

And though some may be skeptical of AI 's contributions to the creative field, the Picture Studio team is thinking critically about their tool’s impact on artists. Because of this, Gutierrez and his team hold a more optimistic view of what AI promises the creative world than most.

“I think of our tools as a way to supercharge ideas and enhance existing skills, not a replacement for doing work,” says Gutierrez. “I have creative projects  that have simmered in my head for years, but my own limitations prevented me from getting them done…[sometimes the] world is a blocker to getting there, whether it's software or something visual. So being able to have an idea, and express it exactly as I had imagined is exciting. It’s a new surface for expression. Picture Studio’s hunch is it can ultimately reign as a critical tool—and are made more confident by the historic moments of skepticism about technology within the arts as precedence.

After all, how different is this moment really from other moments of technological panic in the face of art? Numerous moments can be cited where innovations uncomfortably shook the landscape of creativity. After viewing his very first daguerreotype photograph in 1838, the painter J.M.W. Turner famously quoted, “This is the end of Art. I am glad I have had my day.”

And while photography had an immediate impact on the world—mainly how image-making and the collection of art itself became accessible to a wider class of people—many artists were reluctant to consider the photograph a work of art. Similar to the emergence of generative AI, photography was initially seen as eerie in its accuracy, but still crude. Photography compared to painting appeared somewhat disappointing at first— it took at least 10 minutes of exposure to generate an image, and rarely did it render a product close to perfect. But slowly over time, this perception began to shift.

After all, how different is this moment really from other moments of technological panic in the face of art?

Only four years after the introduction of photography, an experimental collaboration between artist David Octavius Hill and chemist Robert Adamson that resulted in one of the first examples of art photography foreshadowed the medium's potential. Fast forward years, and many technological innovations later, photography eventually evolved into a legitimate art form. Artists like Alfred Stieglitz, Gordon Parks, and Diane Arbus advocated for and demonstrated the medium's potency, challenging traditional notions of artistic creation. Meanwhile, many painters' successes assuaged the fears of 19th century artists like J.M.W. Turner. Reactions to photography gave rise to Impressionism, Surrealism, and Abstract Impressionism, while other artists incorporated photography deep into their practice.

Where we stand in the history of generative AI appears to mimic these early days of photography; products are becoming more sophisticated, and the art of perfecting an image is a main concern. The Picture Studio team is interested in building a more intelligent generative system, a tool that operates as a true collaborator with artists and designers looking to bring a specific vision to life. Following our historical trail, what might the timeline of photography hint at for what comes next for generative AI?

Nearly a century after photography’s inception, some artists’ contributions took a markedly postmodern turn—this was inspired, in part, by the pop art movement. King of Pop-Art Andy Warhol’s work, which made idols of recognizable symbols of consumerism, disrupted the hierarchy of high culture and mass media.

What might the timeline of photography hint at for what comes next for generative AI?

Though slippery to define, Postmodernist art arguably establishes, most essentially, a sort of questioning. What constitutes “legitimate,” art and the means in which it is created? Take Warhol’s Factory for example, which mechanized the creation of art. Warhol challenged the perception of art-making from the image of a lone artist in their studio to a bustling assembly line of reproduction and proliferation. This radical notion ushered in a provocative era of art history.

The Postmodernist take also legitimized the referencing and appropriation of familiar media. Photographer Cindy Sherman is a textbook example of a postmodern artist in that her self-portraits are not a reflection of reality, but of caricature. Her famous Untitled Film Stills series, for example, is a study of stereotypes of women that existed in post-war film and popular culture. Fast forward to now and, interestingly enough, Sherman is experimenting with generative AI self-portraits herself, a thought-provoking commentary on identity and self in the age of social media.

(Photo source: Cindy Sherman Instagram)

Depicted well in Sherman’s off-kilter portraits, postmodern photography is characterized by a sort of uncanniness—something that mimics reality while holding it at a distance. As critic Andy Grundberg has reflected, “Postmodern photography implies the exhaustion of the image universe: it suggests that a photographer can find more than enough images already existing in the world without the bother of making new ones.”

Understanding this history, it makes one wonder: is something like generative AI the logical conclusion to an evolving postmodernist trajectory of thought? One could argue it marks either its continuation or its death, depending on how it is made. It is contingent on how teams like Picture Studio imagine the relationship between the tool and the artist.

Perhaps inspired by his original passion for photography, Gutierrez sees the app’s future capabilities not as a printer, but instead a type of camera. This camera goes beyond aperture and ISO, allowing for adjustments in style and mood as well. “When it comes to the potential of the technology," he says, “I think we're going to get to a point where you can move a virtual camera around the room. Where you have an image of yourself, but can say, Okay, I want to be further away, I don't want the background to be blurry, or I want to reposition myself so I'm looking the other way. All that is possible, and very similar to creating with a camera…" He likens this vision for generative AI to a camera in manual mode, where changing aperture by even one millimeter can have dramatic effects on an image's output. "The same goes here," Gutierrez says. "you could change a few key words, and suddenly, everything is rendered differently. It’s the ultimate form of play for people who think in images.”

And as many know, generative AI would generate nothing if it weren’t for the rich human history it has to reference across the web. “The models that power AI image-making are still relatively crude in their understanding of the world and how to render it, but they are improving exponentially,” Gutierrez shares.

For now, many gen AI solutions are communicating one-dimensionally— think about it like one person trying to explain an image to someone on the other end of a telephone. What if we had better ways to provide the person on the other end context clues, and even ask how they might visually interpret the scene?

The Picture Studio founder says that “as [AI image-making models] asymptotically converge, the tooling layer used to control them will become ever more important, because even a perfect understanding of a prompt can’t account for the multitude of ways that idea might be expressed." His belief is that in order to evolve, these tools require new ways of thinking about workflows. Generative AI can begin as it does now with a prompt, but then ought to move fluidly to new sets of visual editing tools that allow the artist to contort the image to their liking, and do things one would have a hard time imagining was even possible with the software.

The power of Gutierrez's technology lies ultimately in its flexibility of command. Picture Studio is a tooling layer that harnesses the power of image models. It is model agnostic and can easily switch between any number of open source and commercial models, but most importantly, it allows creative people and creative companies to create private models based on their own IP.

Practically speaking, this means artists can create a small set of images they use to generate a much larger body of work, or they can use generated images to extend and enhance their work. Models of people, characters, styles, and objects can be combined with ideas in prompts fluidly in ways that allow for rapid iteration of ideas.

Images created with simple prompts from models trained on 1920s Loteria cards.

Though it sounds sci-fi, Gutierrez also hints that a generative AI accurately handling a request to incorporate symbolism is a closer reality than you might imagine. This deeper level of communication marks a new phase of the evolution of AI; one that asks, what does AI have to say?

It’s all a bit overwhelming to comprehend. And for some, this conclusion may sound quite eerie. But it’s important to think about how J.M.W. Turner may have been feeling at his first glance of the daguerreotype—did his prophecies come true? It’s possible that these softwares merely add color to the landscape of creative tools rather than illustrating predecessors’ demise. After all, darkrooms still exist despite the emergence of Photoshop, and though the Kindle is a ubiquitous tool, libraries are exploding in popularity. As it is that technology disrupts the natural, expected order of established creative process, it shares the value that art movements are marked by ideas that shake up the world.

For Gutierrez and his highly creative team, the product they are building is a game-changing tool for their creative practices. “People feel threatened, because [new technology] does replace certain things,” he shares. “When portrait photographers became a thing, some people would rather have a photograph than somebody's painted interpretation of something. But it also opened up entire new worlds. The best painter’s work became more valuable, but a certain class of painters were replaced by photographers. And with AI image making, it feels to me that a lot of the same kind of dynamics are in play.”

The technology can feel unwieldy because, as a whole, we understand very little about it at the moment. But perhaps we ought to keep an open mind about what it might evolve into. “Over the last few months as we’ve built Picture Studio, "I’ve started thinking of models as encapsulations of idea space," Gutierrez reflects. "[They are] a dimension, just like time that can express those ideas in ways that tend to reveal the invisible. Just as placing a camera on a ledge and pointing it out at a building for a few months can reveal the movements of the sun and blend everyday moments into a seamless whole, the models tend to reveal the essence of an idea."

Gutierrez gives the example of attempting to draw his wife Jenn, who he proclaims he's always had trouble doing accurately. But after creating a model of her, prompting the software to generate sketches and line drawings of her face, he says, "the hidden geometry of her face, the things that make Jenn Jenn were revealed. The generated images, in turn, taught me how to draw her by hand.”

“And speaking of Jenn," Gutierrez recalls, "the other night I found her on my computer creating images on Picture Studio. She was visualizing myths her Korean grandmother told her as a child and was expressing them in the mode of the technicolor movies her grandmother loved. This left her teary… they were images she had been holding in her mind since childhood and had never had the tools to express them. 

Gutierrez is excited about what comes next. “Images are the currency of modern life. The web, social media, all run on them. We have a chance to make the tools that will help this generation of creative people and companies tell their stories. Seeing the images people create is exciting to me. It’s like peeking into idea space. The feeling is the same as opening one of my camera obscura boxes to see what the accumulation of days has rendered.”