Craft is often defined as “an activity involving skill in making things by hand.” In art, however, craft goes beyond manual labor because it gives the work meaning. When someone masters a skill, it is magical to see them use it as a form of expression. Craft isn't just about manual labor. It is about engaging with the process intentionally, about understanding the medium, its limitations, and its possibilities.
When someone masters a craft, they work with the limitations of their medium. Working with a medium means making conscious choices. The final piece is the result of many choices, all made to achieve a result that best expresses what you wanted to say or show.
While these limitations are obvious in physical art mediums, they have always been less clear in digital art. From its earliest days, the limitations of digital art mediums have always been changing. At first, the constraints were writing code to draw pixels. Later, they became the size of the canvas, the available colors, the number of polygons that could be rendered, the realism of simulated light, or the input devices used to manipulate the work.
The digital art toolset has always evolved alongside technical progress. The goal was to create more, faster, and at higher quality. Today, quality is rarely a limitation. We have CGI that is indistinguishable from filmed footage. We have highly immersive video games in every imaginable art style and genre. When we see bad CGI now, it's not a lack of tools; it's because the artists weren't given enough time.
Technological improvement has always been a steady progression. Some steps were larger than others, but they were predictable and understandable. That's not to say tools always have to be fully adopted. People still use Microsoft Paint to draw realistic portraits, not because there aren't tools that make it easier, but because working with limitations can be a meaningful challenge in itself.
This steady progression is now changing. Technology is no longer just improving tools; it is making a leap. Where previous advances reduced the number of steps to reach a result, generative AI has the potential to remove those steps entirely.
With more of these tools emerging every day, resistance from artists is growing. Eliminating or simplifying a single step in your process is perceived positively. It saves time while keeping the artist in control. Skipping the entire process, however, feels like a loss of control. This results in rejection of AI use as a whole.
This is where the main tension of digital art becomes visible. From the very beginning, tools in digital art have been about reducing limitations. Better tools increased what artists and teams could create, allowing them to more fully realize their vision. Digital art has always been shaped by its tools. Every new tool drives a new wave of creative expression.
Photoshop made images possible that could never exist before. 3D rendering enabled entirely new kinds of films. Now, generative AI seemingly removes all creative constraints. The question becomes what happens to artistic control when outcomes no longer require a process.
A fair and important critique of generative AI is the training material used to build the models. Violation of copyright and the importance of consent deserve discussion. At the same time, they are separate from the practical reality that artists are facing. Generative AI models already exist, and there is no practical way of reversing their presence. As long as these models are commercially valuable, attempts to ban them are unlikely to succeed.
Art is not a sport. In sport, rules are explicitly defined and enforced. Certain shoes can be banned from running competitions because all participants agree to the same constraints. Art, especially in a commercial context, is different. While competitive environments still have rules and expectations, tools can't simply be forbidden because they change how work is produced.
Some artists may be able to continue working with their current processes for niche audiences. For most, adaptation will be unavoidable. Choosing not to engage with generative AI isn’t a moral stance, it’s a creative constraint. It is similar to deliberately working with a limited toolset from earlier eras of digital art.
Adapting your process does not mean prompting an AI model to generate a final image and claiming it as one's own. This approach eliminates the artistic process entirely, removing the decision-making and craft required to realize a vision that goes beyond a first idea.
This is not to say that AI will never be used in this way. Producing a usable result with a single action is extremely valuable in many contexts, especially where speed or scale matters more than artistic voice. But it does not have to be the only way generative AI is used.
AI can also be used as a tool within an artistic process. It can be used to enhance creative exploration, allowing artists to test ideas and iterate at a higher speed. It can remove tasks that would otherwise require large amounts of repetitive labor. Smaller teams can take on more ambitious projects, and work that could once only be tackled by large studios can become accessible to the indie teams that we love for their unique ideas and creativity.
Tools that integrate AI into an intentional creative process have barely been explored so far. Most existing AI tools are designed for people who do not want to invest time in mastering a craft. Their goal is to minimize input and optimize for results, not for control and decision-making. Preserving an artistic process requires tools that can be used with intention. That means shaping AI into a new kind of toolset.
These two directions already exist in software engineering. On the one hand are vibe coding tools, aiming to eliminate the need for an engineer entirely. They reduce decision-making to defining the outcome. On the other end are tools built for engineers. Examples are copilots that remove repetitive work, agents that handle well-defined tasks, and review tools that speed up releases. These tools don't replace the engineer; they enhance them. As a result, smaller teams can now tackle larger projects in shorter timeframes.
A similar split is possible and desirable in digital art. Allowing people to create media without a process is valuable in many contexts. At the same time, work that benefits from a larger vision, deep subject knowledge, and exploration can use AI tools as an extension of the artistic process rather than as a shortcut around it.
The goal of tools in digital art is not only efficiency. It is also about enabling new possibilities, whether it's creating specific styles, achieving higher realism, or creating more complex scenes.
The real value of generative AI as a tool lies in allowing artists to create what was previously out of reach.
Digital art has always been defined by its tools. The next step is deciding who those tools are built for.
Rejecting generative AI outright is neither useful nor realistic. Choosing not to use it can be a valid creative decision. It is a creative constraint that will shape the process and outcome. But refusal will not change how these tools evolve.
If craft should stay important to digital art, artists will have to step up to use the tools and to actively shape them. The goal is not to prevent automation, but to design processes that preserve intentional control, creative exploration, and decision-making.
We are still in the early stages of this transformation. That makes generative AI not a threat, but an opportunity. Artists have the chance to consciously define how AI fits into their process and shape it into tools that expand creative expression rather than replace it.