Boosting Business Models with AI: Scalable Design Systems

One of the most revolutionary aspects of generative AI as a content production tool is autonomy. Generative language models, whether for text, image, sound, or video, propose a paradigm shift in that any individual can instantly generate a detailed report on the military conflicts that occurred between 1939 and 1945, without even knowing about the winners, losers, or culprits.

Now, we are not historians, but we like to make history, and if we can build Rome in a day, we do it.

‍

Here comes AI.

What do we mean by autonomy, and where do we, as professionals, believe we can establish a differential line in the execution of a tool that democratizes operational capacity without differentiating between expert and novice? Or at least it seems that way.

Let's start at the beginning… autonomy. By definition, it is "the condition of someone who, for certain things, does not depend on anyone." So, when we talk about it in business terms, we think of execution, optimizing times, and talking about time implies talking about resources, and if we talk about time and resources, it is inevitable to think about scalability.

So, if we associate business with design, we can think of the ideal of autonomy when any participant in the execution of a project has the virtue of developing multiple tasks and growing as a versatile, independent individual, and even better, intervenes and favors other areas. Thinking about the health of a business, the self-sufficiency of its participants is a key component, necessary for thinking about survival and growth. Artificial intelligence, within this ideal of self-management, disseminates cross-area knowledge to elevate and enhance the execution spectrum of each one. The line between a marketing team, a design team, and an art team is increasingly thin.

As a small mention, by autonomy, we do not mean limiting the capacity for team meetings. Not at all. We believe in shared knowledge, that a group can do more than an individual, and that in the mix of the background of a group of people lies the richness of being. In itself, the place that these types of tools occupy is raising the bar in areas of ignorance, to take cross-area meetings and debates to the next level.

Now, if AI is a democratic technological component, both in knowledge and execution, and both we and our hypothetical competitors have access to the same, how do I use it as a differential component? How to add value to an element that levels knowledge and makes us all executors where we were not before? Where is the value of the specialist at this moment?

From Paisanos, we can exemplify it with a Case Study: zityhub. The client approached us with the aim of renewing their website and communicating their value proposition and differentials more attractively, along with the need to identify a new graphic direction that enhances their identity, generates recognition in the market, and gives a corporate image based on their principles and virtues. Moreover, according to a small product structure and a busy Marketing and Design team, an underlying need that emerged from their team was to train and provide tools to those teams involved in generating graphic material. The challenge lay in ensuring that any member of the content teams could generate it autonomously, even if the graphic direction included elements that exceeded their expertise and execution capacity. And also, how to maintain consistency, coherence, and quality of what is generated along the way? How to be an omnipresent art director after the handoff and delivery?

‍

Here come the experts

At the process level, we always start from the base given by our moodboard. What is a moodboard and why is it fundamental to our process? A moodboard is a visual and compositional panel that originates from the aspirational brief of our client, the main concepts of their brand, and an extensive graphic reference search. It defines the product identity in visual terms—how we imagine its typography, its compositional margins, its colors, miscellanies, etc.—and even more importantly, in conceptual terms, what we want to convey with it. This panel is fundamental as it will be the one that defines the general atmosphere and will act as a sieve for the selection of graphic elements. Translating to AI, the moodboard will define the set of attributes we will assign in the prompt when executing tools like MidJourney (which, by the way, we use).

For this case, some main concepts that permeated the brand were efficiency, innovation, intelligence, flexibility, fluidity, technology, humanity, and corporateness.

So, with the direction and concepts defined, we have an aesthetic framework. Now we need to structure the solution: the dynamic prompt. This prompt will be the new art director, our guide, so that anyone who executes it creates curated pieces in line with what we designed during the course of the project. This statement should be closed enough to delimit consistency and coherence, but open in subject to be versatile enough for daily use.

‍

‍

‍

This is one of several prompts we designed during the work process. I say several because although the prompt is dynamic, we categorized the brand’s needs so that each prompting structure matches the expected outcome. That is, for the “photography” category we had one prompt, for the “renders” category another, and so on according to the need.

So, if we have different prompt structures according to the different types of subjects, the visual consistency between pieces, as we said previously, will depend on the qualifiers and descriptive elements to maintain the design line. Let’s call qualifiers all those attributes that determine the style. Elements such as: color tone and temperature, lighting, materiality, shot type, etc., etc. In themselves, they are fundamental for our prompts to generate atmosphere and personality, and where the general coherence cross-prompt can be obtained. On the other hand, the statement structures will allow the language system to interpret in the best way all the compositional load and are fundamental for a scalable result.

‍

Recap

So, I quote the triggers of this entire article: How do I use AI as a differential component? How do I add value to an element that levels knowledge and makes us all executors where we were not before? Where is the value of the specialist at this moment?

Responding from our solutions:

• We found that artificial intelligence empowers teams to be, develop, and exceed their limits.
• We identified in the ease of use and low entry barrier, a space to promote the use of this framework to all teams and be autonomous within their usual workflow, eliminating the time of shared operation in areas where it was not so necessary. At the same time, we have been, are, and will be part of this empowerment in a judicious manner.
• We found that the value of the specialist lies in technique and exploration. In knowing the semantic structure proposed by language models in-depth, to adapt different structures to various outcomes.
• We found the differential in the knowledge of art history and design discipline, to infuse that statement to which we all have access with style.
• We found value in the designer’s background, in their language, in their uniqueness, in their poetry. If they can do it, they can teach it, and if they teach it, they add value to the product and the person accompanying it.
• We found added value in the detail, in being meticulous and exhaustive, falling in love with the process… at least until the roadmap is finished. Just kidding, we will always give a little more.

Te dejamos el caso completo acá 🧉