Is AI Prompting like Web Design or like Fashion Design?

On separating form and function.

AI text prompting is everywhere now. You've likely encountered it, perhaps even used it today. But what does it truly entail?

At its essence, prompting is a dialogue with an artificial intelligence. You input information, and it generates a response. The outputs can range from insightful to unexpected.

The challenge lies in directing these outputs. Can we distinguish between content and style? Is it possible to separate the substance from its presentation?

This is where the process becomes both complex and intriguing. Consider two design disciplines: web design and fashion design. They approach the separation of form and function differently. Web design often compartmentalizes these elements, while fashion design tends to intertwine them.

Where does AI prompting fall on this spectrum? Does it resemble the structured approach of coding a website, or the integrated process of creating a garment? This article aims to explore these questions. We'll examine the nature of AI prompting, drawing parallels with different design philosophies. Through this exploration, we may gain insights into AI, design principles, and the evolving landscape of human-machine interaction.

Let's begin our investigation.

Web design: separation of form and function

Let's talk web design. It's akin to constructing a digital sandwich. HTML serves as your filling - the substance, the core content of your website. CSS? That's your presentation layer. It's the arrangement, the visual appeal, the way it's all displayed. This separation is a clever approach. It allows you to modify the appearance without altering the content, or vice versa. But it's not merely about aesthetics.

This division serves practical purposes: faster loading times, easier maintenance, and responsiveness across various devices. One style sheet can govern multiple pages. However, it's not always a perfect system. Occasionally, HTML and CSS become intertwined, resembling spaghetti code. But the principle remains valid. Form and function can exist independently in web design.

Fashion design: form and function intertwined

Fashion design, on the other hand, is a different realm entirely. Imagine attempting to separate a sweater's insulation properties from its fluffiness. Challenging, isn't it? In fashion, form and function are deeply interconnected.

The fabric isn't just for show: it's the structural element, the essence of the garment. Consider a wedding dress: the silk isn't merely decorative. It provides structure, flow, and defines the entire piece. Or take activewear. The elastic material isn't just about comfort. It's about performance, durability - the garment would fall apart without it. While minor alterations are possible, such as changing a button or adding a pocket, the overall design is largely inseparable from its function.

This intrinsic connection between form and function in fashion design creates a unique constraint. It compels designers to think holistically, where every decision and every element plays a crucial role in the final product.

Basic Prompting Techniques

Let's get back to AI prompting and consider some techniques that involve form and function.

First, consider chain-of-thought prompting. It's a method of asking the AI to demonstrate its reasoning process. Similar to how a math teacher might ask you to show your work, this technique encourages the AI to think more systematically, step by step.

Output primers are another useful tool. They act as initial guidance for AI responses. For instance, you might instruct the AI to "Start your response with a joke" or "Provide a list of three items." This gives the AI a starting point for its output.

Multimodal prompting expands the AI's input beyond text. By incorporating images, sounds, and other data types, it's akin to giving the AI more sensory information to work with. More diverse input often leads to richer, more nuanced output.

But here's the crux of the matter: can we truly separate content from style in AI prompting? It's a challenging proposition. Sometimes, it works well. You might request "facts about elephants in the style of Shakespeare," neatly packaging content and style. Other times, it's more like trying to un-scramble eggs.

Sure, I’ll do it.

Prompt: in <100 words, give facts about elephants in the style of Shakespeare.

O, fairest elephant, with wrinkled hide, 
Thy massive form doth roam the savannah's pride. 
Thy tusks, like ivory scimitars, do shine, 
And with thy trunk, thou dost entwine and climb. 
Thy memory, a treasure trove of old, 
Doth recall the paths of seasons past, and tales of yore told. 
Thy social bonds, a testament to love and might, 
Do bind thee to thy herd, in endless, gentle night.

‘Un-scrambled Eggs’

Moving on. Prompting Culture’s ongoing experiments with short story prompts illustrate this complexity. I've tried various approaches, sometimes focusing on plot points while leaving style open, and other times specifying a writing style but leaving the story vague. The results have been mixed.

In one instance, I prompted for corporate environmentalism in the style of Roald Dahl. The environmentalist message came through, but according to one commenter, Dahl transformed into Dr. Seuss - an unexpected twist.

Another time, I asked for James Patterson’s take on office drudgery. The result? Murder, of course. Not quite the clever observations and cynicism I had envisioned.

The challenge lies in how AI processes information. It doesn't compartmentalize "content" and "style" the way we might. Instead, it works with a complex web of probabilities and patterns.

So, does AI prompting align more closely with web design or fashion design? Probably the latter. In some cases, form and function can be cleanly separated. But mostly, they're quite intertwined, and changing one will impact the other. The key lies in experimentation, trial and error, creativity, and patience. Ultimately, prompting isn't just about directing the AI to perform specific tasks. It's about discovering where it truly shines and appreciating it for its unexpected and innovative outputs.

Recent Developments

It's 2025, and AI prompting's evolved. Mega-prompts are normal now. These aren't just longer instructions; they're intricate frameworks. Think of them as detailed blueprints for AI, complete with context, examples, and specific guidelines. A mega-prompt for a medical diagnosis might include patient history, symptom patterns, and even recent medical literature.

Adaptive prompting's gaining traction. AI systems now adjust their queries based on user responses. It's like a conversation that gets smarter with each exchange. This tech's particularly useful in customer service bots, which can now handle complex inquiries with increasing finesse.

AI-generated prompts are slowly moving into focus. We're using AI to optimize prompts for other AI systems. It's meta, but effective. So effective that I can’t imagine using anything else for generating my illustrations. It’s as simple as prompting (after your final AI-assisted article rewrite):

Prompt: Now describe in great detail a simple yet characteristic scene from our story, so that our AI image generator will make us an excellent illustration. Include style elements. Use comma separated style. Don’t use names. 75 words max.

Outside of illustrations, AI-generated prompts can analyze vast datasets of successful prompts, identifying patterns and generating new, highly effective instructions. There’s lots of potential there.

Real-time optimization has become a game-changer. Prompt engineers now get instant feedback on their prompts' effectiveness. It's like having a co-pilot that suggests tweaks as you write. This technology's dramatically reduced the trial-and-error phase of prompt crafting.

Zero-shot prompting is making waves. AI can now tackle tasks without specific examples, relying on its broad knowledge base. This advancement's particularly impressive in language translation and content generation tasks.

The field's moving fast. What was cutting-edge last year is now standard practice. AI prompting in 2025 is more nuanced, more powerful, and more integrated into our daily tech interactions than ever before.

Future Implications

Looking ahead, AI prompting's set to shake things up even more. No-code platforms are on the horizon. Soon, anyone with an idea can craft complex AI interactions, no programming skills needed. Some might even describe this as a sort of ‘democratization’ in action.

Personalization's the next frontier. Imagine AI that responds not just to general user input, but to your specific communication style. You could compare it to having a digital assistant that knows you better than you know yourself. Maybe current models do this already, it’s hard to keep track sometimes.

Cross-domain integration is gaining momentum. We're seeing AI that can juggle multiple specialties at once. A single system might soon handle everything from medical diagnoses to financial forecasting.

The real game-changer? AI that can explain its own reasoning. Transparency's becoming key. We're moving towards systems that don't just give answers, but show their work. These advancements aren't just cool tech. They're reshaping (to use a wonderful AI term) how we interact with AI, blurring the lines between human and machine intelligence. The future of prompting is not simply about asking the right questions, but more about a real dialogue with the machine.

AI prompting sits in a fascinating space. Sometimes it's like web design - clean, separable, modular. Other times, it's pure fashion design - messy, integrated, where every element touches everything else.

We're not just talking about a technical process. We're exploring a new language. A way of communicating with machines that's part science, part art. Part HTML, part silk dress. The boundaries are blurry, the potential is enormous, and we're just getting started. Want to play? Start prompting.