I was sitting at my desk yesterday, cleaning the lens of a 1970s Leica, when I realized how much we’ve lost the sense of deliberate friction in our work. Lately, my inbox is nothing but noise about how generative ai tools are going to “revolutionize” the creative industry by making us faster, better, and—most importantly—more efficient. It’s a hollow promise. The industry is being sold this myth that speed is the same thing as quality, as if a prompt can ever replace the muscle memory of a designer who understands how a single kerning adjustment changes the entire soul of a typeface.
I’m not here to sell you on the magic or the fear-mongering; I’m here to talk about intentionality. In this piece, I want to move past the shiny marketing fluff and look at how we can actually use these generative ai tools as a way to amplify our own agency rather than surrendering it. I’ll be sharing my honest, no-nonsense perspective on which technologies actually respect the craft and which ones are just shortcuts that lead to a sea of visual mediocrity.
Table of Contents
Large Language Models Explained Through a Human Lens
When we talk about large language models explained in technical terms, the conversation usually gets lost in a sea of tokens, parameters, and neural weights. It feels cold, like reading a manual for a machine that has no soul. But if I look at it through the lens of a designer, I see something different. I see a massive, digital library of human expression—a vast, interconnected web of how we structure thought and rhythm. It isn’t just math; it is a reflection of our collective patterns of communication.
To me, these models function less like a calculator and more like a highly sophisticated apprentice. They don’t “know” things in the way we do, but they have observed the way we weave ideas together. When we explore different generative ai use cases, we shouldn’t be looking for a way to outsource our thinking. Instead, we should treat the model as a sounding board—a way to test the tension of a concept or to find a new way to phrase a brand’s core values before we ever touch a pen to paper. It is about finding that creative friction that leads to something better.
Beyond Automation Ethical Implications of Generative Ai
When we talk about the ethical implications of generative AI, we often get lost in the technicalities of data scraping and copyright. But for those of us who have spent our lives building visual languages, the concern runs deeper than just legalities; it’s about the sanctity of the creative impulse. As I sit here with my sketchbook, I can’t help but wonder if we are trading our unique perspectives for a polished, mathematical average. If we lean too heavily on ai image generation software to solve every aesthetic problem, we risk a future of visual stagnation where everything looks “perfect” yet feels entirely hollow.
The real challenge isn’t just avoiding theft, but avoiding the erosion of intent. We have to ask ourselves if we are using these systems to expand our horizons or if we are simply automating workflows with ai to the point where the designer becomes a mere curator of machine-made choices. We must ensure that as these tools evolve, they remain a scaffold for human expression rather than a substitute for the deliberate friction that makes art meaningful.
Reclaiming Intentionality: Five Ways to Work With AI Without Losing Your Soul
- Treat the prompt as a sketch, not a final directive. Just as I wouldn’t expect a single charcoal stroke to define a complete composition, don’t expect a single prompt to deliver a finished brand identity. Use AI to generate the “roughs,” then bring your own hand back in to refine the edges.
- Curate the output with a designer’s eye for friction. AI tends to smooth everything out into a generic, frictionless sheen that lacks character. Your job is to reintroduce the “human error”—the slight asymmetry or the unexpected color tension—that makes a design feel alive.
- Use LLMs as a sounding board for logic, not just a generator for copy. Instead of asking it to “write a brand story,” ask it to “critique the emotional resonance of this mission statement.” Use it to stress-test your strategic thinking rather than letting it do the thinking for you.
- Maintain a physical paper trail. When the digital tools feel too overwhelming or the AI-generated suggestions feel too hollow, step away from the screen. Revert to your sketchbook and fountain pen to map out the core concepts. This ensures the foundation of your work is built on human intuition, not an algorithm’s probability.
- Audit your tools for “creative drift.” It is easy to become lazy when a tool can mimic a style in seconds. Periodically review your recent work to ensure you aren’t accidentally adopting the aesthetic biases of the models you use. If your work is starting to look like everyone else’s, it’s time to pivot back to manual craftsmanship.
The Intentionality Checklist
Treat AI as a high-fidelity sketchbook, not a finished product; use it to explore the breadth of an idea without letting it dictate the finality of your vision.
Guard the “friction” in your process, ensuring that the soul of your work—the subtle, human imperfections—isn’t smoothed away by an algorithm seeking optimization.
Prioritize tool-building that augments your specific craft, focusing on technologies that handle the repetitive heavy lifting so you can return your focus to the core principles of design.
The Ghost in the Machine
“We shouldn’t be looking for tools that can mimic the final result; we should be looking for tools that respect the messy, beautiful friction of the process—the kind of tools that act as a digital extension of our own intent, rather than a substitute for our own judgment.”
Ingrid Bellamy
The Intentional Path Forward
We’ve spent much of this discussion dissecting the mechanics of LLMs and the heavy ethical weight they carry. It is easy to get lost in the technical noise or the fear of being rendered obsolete, but the core of our conversation remains the same: these tools are neither gods nor replacements. They are mirrors of our own data and, if we use them correctly, extensions of our own curiosity. As we navigate this shift, we must remember that efficiency is not the same as excellence. We can use generative tools to handle the heavy lifting of pattern recognition or initial ideation, but the responsibility for the soul of the work—the nuance, the empathy, and the strategic “why”—remains firmly in our hands.
As I sit here with my sketchbook and a half-empty cup of coffee, I am reminded that the most beautiful things often come from the struggle of the process. Technology will continue to accelerate, and the friction we once felt in physical paste-ups has been replaced by the seamlessness of digital pixels, yet the human intention must remain the constant. Don’t let the speed of the tool dictate the depth of your vision. Instead, use these new capabilities to push past the mundane and reach for something more profound. Let us build a future where we don’t just automate the craft, but use every tool at our disposal to elevate the human touch.
Frequently Asked Questions
How do we maintain a sense of personal style and "human friction" when the tools we use are trained on such massive, homogenized datasets?
We have to lean into the “imperfections” that the algorithms are trained to smooth away. The datasets are built on averages, which is the enemy of character. To find your voice, you must intentionally introduce friction—use the tool to generate a foundation, but then break it. Bring in your hand-drawn sketches, your idiosyncratic kerning, or that specific, non-linear way you compose a frame. Style lives in the deviations from the mean.
Where is the line between using AI as a productive assistant and letting it dilute the intentionality of the design process?
The line is drawn at the point of decision-making. When I use a tool to automate a repetitive task—like resizing a grid or organizing layers—I’m reclaiming time for thought. But when I let an algorithm dictate the soul of a typeface or the tension in a layout, I’ve lost the thread. If you aren’t the one making the final, difficult choice about why a specific element exists, you aren’t designing; you’re just supervising.
As these tools become more integrated into our workflows, how can we ensure they remain instruments for creators rather than just engines for mass production?
We have to treat these tools like a new kind of lens, not a replacement for the eye. To keep them from becoming engines of mindless mass production, we must maintain our role as the ultimate curators of intent. It’s about using AI to handle the heavy lifting of iteration while we guard the nuance, the texture, and the “why” behind every decision. If we stop questioning the output, we stop designing.
