The Role of Design in GAI Development: A Weave of People and Technology

In recent years, the advancement of generative artificial intelligence, through initiatives like ChatGPT and, more recently, DeepSeek, has popularized a new way for people to interact with computational systems. Natural Language Processing (NLP) is increasingly occupying more space in areas such as business, healthcare, education, and services. However, as chatbots become more frequent on our screens, the technical challenges in their development face increasingly complex questions: How to validate the usefulness of this interaction mode in a project? How to ensure the reliability of the information sought? These are some of the questions that arise. It becomes more evident every day that establishing relationships between business needs, human expectations, and technical requirements is no trivial task.

This text presents the argument that, despite having a history in computer research and development, design practices play a fundamental role in the development of systems that use natural language processing in conversational interfaces. It brings a vision that connects technology with meaning, placing people and their communities at the center of the project.

As a starting point to understand the role of the designer, let’s go back to the basics and explore two complementary ways of thinking about information: the quantitative view of Claude Shannon and the relational view of Gregory Bateson.

Two Concepts of Information: Engineering and Anthropology

Shannon’s View (Engineering and Computing):Claude Shannon defined information as the reduction of uncertainty, focusing on the amount of data that can be transmitted efficiently and accurately. His approach is essential for the technical development not only of chatbots but of all computational systems to date, which rely on mathematical and statistical models to process data—in our case, natural language—optimize responses, and operate reliably.

Bateson’s View (Design and Meaning):Gregory Bateson, on the other hand, defined information as "a difference that makes a difference," highlighting the relational and contextual impact of information. For Bateson, information is meaningful only when perceived within a system—that is, for our understanding in this text, when it makes sense in the human and social context (in broader terms, which we won’t delve into here, he also applied the concept to communication, social psychology, and biology).

These two views, two understandings of the world, provide the tools that represent the necessary complementarity in chatbot projects: while Shannon lays the technical foundation, Bateson reminds us that interaction must promote transformation in the human context.

The Connection Between Business, People, and Technology

An effective chatbot is thus woven into a framework that operates in three main dimensions:

1. Business: It must meet the organization’s strategic objectives, such as improving customer or employee experience, promoting public well-being, reducing costs, or generating revenue.

2. People: It must create meaningful and useful experiences, responding to the real expectations and needs of people in their communities.

3. Technical: It must be technically robust, efficient, and scalable.
   

The Role of the Designer in Weaving This Framework

Within a project, the design team is the mediator that weaves relationships between these dimensions, ensuring that development is not only technically feasible but also useful, human, efficient, and strategic. In a schematic and simple way, we can outline how this happens:

1. At the Start of the Project:

   • Designers deeply investigate people’s needs in their uses, within their communities, understanding business objectives and mapping the differences that make a difference in the context of chatbot use (Bateson).

   • This includes user research, interviews, and journey mapping to understand what truly matters to the target audience.

2. Collaborating with Engineers:

   • Combining Shannon’s view to build the analysis and processing of available data. Designers work with engineers and scientists to determine which AI models (generative or not) are most suitable for the context and desired experience.

3. Creating Meaningful Interactions:

   • While engineers optimize models to handle large volumes of data, designers ensure that the chatbot communicates clearly, empathetically, and in alignment with the human context.

4. Continuous Iteration:

   • Designers promote testing and improvement cycles, ensuring that the chatbot not only functions but resonates with users and generates positive results.

Why Multidimensional Approaches Matter

Often, chatbot projects fail because they focus solely on technological challenges (which are not small) or business objectives, neglecting the human experience in their habits and needs. Integrating a design approach ensures that the chatbot:

• Addresses real problems, creating value for both users and the business.

• Responds to the cultural and social nuances of the audience.

• Operates within business needs and technological approaches in an ethical and responsible manner.

Conclusion: Returning to the Starting Point

Technological development is intrinsically linked to human existence itself. In a world where such computational technology advances rapidly, chatbots need to be more than efficient systems—they need to promote the construction of experiences that produce appropriate meanings.

By understanding information not only as something that "transmits" (Shannon) but as something that transforms (Bateson), teams composed of designers, scientists, and engineers enable the creation of projects that impact lives and drive change. The future of chatbots lies in collaboration, in the community, and the reflections brought here seek to remind us that, in the fabric of all technology, the main threads are those of people and their communities.