Published on: 2025-04-30

Breaking the Graphical Interface

A new paradigm in how to build products and services.

A decade ago, while getting lost in the world of Iron Man, I found myself fascinated by Jarvis, Tony Stark's virtual assistant who handled tasks as if he were a superhuman with a degree in multitasking. Now, in the middle of 2024, it seems that fiction has become our new daily reality. Although we are still far from having a Jarvis that makes us coffee, technological advances have brought us closer to a world where dialoguing with virtual agents is part of our routine. These not only understand our questions but also act on simple voice commands.

Think about it: what would it be like to have a Jarvis at home? Something like saying, "Hey, order some empanadas on Rappi," and done, without touching a button. We are heading towards a future where chatting with technology will be as simple and natural as asking a friend for a favor... or, as happens in my house, where I act as Jarvis for my wife (yes, even superheroes have their domestic chores). In this article, we will delve into how Large Language Models (LLMs) are reinventing graphical interfaces, transforming interaction with machines into something as everyday as asking your mom for advice or telling your dad a joke. And not only that, we will see how they are radically changing product development, forcing us to rethink our strategies and approaches.

What is an LLM and how does it work? The magic behind ChatGPT.

Surely many of you have already had an encounter, let's say, "intellectual" with ChatGPT. Maybe you have asked it about grandma's recipe or how to solve that annoying error in Excel. Behind that ability to answer almost anything lies a Large Language Model, or LLM for its acronym in English.

What exactly is an LLM and how does it achieve those feats that sometimes leave us speechless? Imagine a gigantic library where, instead of shelves with books, we have digital data. An LLM is like a very diligent librarian who has "read" all those books (or rather, has analyzed millions of words on the internet), and not only that but has learned to understand how we use language in different contexts.

When you ask ChatGPT a question, this digital librarian searches through everything it has learned to give you the best possible answer. It creates new responses based on that vast accumulated knowledge. It's like conversing with the most informed person in the world, but through your computer or phone.

Agency in LLMs and its impact on graphical interfaces.

When exploring the world of LLMs, we encounter a key term: 'agency'. Beyond simply answering questions, these models can perform actions based on the commands they receive. This is revolutionizing our graphical interfaces, moving them from the screen to everyday conversation.

Let's think of a practical and very current example: scheduling a medical appointment. Before, this involved a series of steps: finding the office number, calling (and often waiting on hold), speaking with a person, and finally agreeing on a time. But with LLMs, this can be greatly simplified. Imagine sending a direct message via WhatsApp or Instagram to an intelligent virtual assistant: "I need an appointment with my primary care physician for Tuesday afternoon." The LLM, acting as an intermediary, processes your request, checks availability, and confirms the appointment, all within the same messaging platform we already use daily.

This approach has several advantages. First, it reduces friction for the user: there is no need to install new apps or learn complicated interfaces. Communication takes place through platforms already known and massively used. Second, efficiency is improved: the LLM handles requests quickly and accurately, freeing up time for more complex tasks. And third, accessibility is increased: people of all ages and technological levels can easily interact with these technologies.

We are witnessing how LLMs are transforming interaction with technology, moving from complex graphical interfaces to natural conversations on platforms already integrated into our daily lives. The agency of these models is redefining the way we do everything from the simplest tasks, like sending a message, to the most essential ones, like taking care of our health.

In addition to these conveniences, this advancement presents a challenge to product creators and developers: to think of solutions that are more human, more conversational. It's no longer just about touch screens or dropdown menus; now, the challenge lies in creating experiences that feel like talking to an old friend.

Practical applications and use cases: T-Bit and the world of charter flights.

In the business world, Large Language Models are opening up a range of possibilities. An example of this is what we do at our company, T-Bit (www.t-bit.io). We specialize in creating innovative solutions using LLMs, and one of our most notable success cases is with our client Pikzit (www.pikzit.com), a company that offers an alternative for sharing charter flights.

Pikzit faced a common challenge in the aviation industry: the need to inform customers about the specifics of charter flights, which are often misunderstood or unknown by the general public. Frequently asked questions included concerns like "What exactly is a charter flight?", "Is it more expensive than a commercial flight?", and "How can I book one?". To address this, at T-Bit, we trained an LLM-based bot specifically with these frequently asked questions.

The result was a virtual assistant capable of interacting with Pikzit's customers fluidly and naturally. Through a simple and friendly conversation, the bot provides clear and accurate answers to user queries, improving the customer experience and optimizing the service process. This solution not only increased customer satisfaction but also relieved the workload of Pikzit's customer service team, allowing them to focus on more complex and personalized tasks.

This Pikzit case demonstrates how LLMs can be applied practically and effectively in the business world. It is not just a tool for answering questions, but a comprehensive solution that enhances the customer experience and optimizes internal company processes. At T-Bit, we are committed to exploring and expanding these frontiers, taking artificial intelligence and LLMs to new heights in the business world.

Closing the circle: From cinema to reality.

Reflecting on all this, I return to the image of Jarvis in Iron Man. What was once a cinematic dream is now taking shape in our reality, with an authentically Latin American touch. We are living in an era where technology not only assists us but also understands and anticipates our needs, even if it is still learning how to make coffee.

An invitation to business transformation in Latin America.

But what does all this mean for us in Latin America? The answer is simple and exciting: opportunity. For both established companies and emerging startups, the era of artificial intelligence and LLMs opens doors to automate processes, optimize resources, and, above all, to innovate.

At T-Bit, for example, we help companies take their first steps into this new world at no initial cost. We offer LLM-based customer service chatbots that allow companies to start automating their interactions without prior investment. Production use is based on consumption, meaning costs are aligned with the company's growth and scale.

This accessibility is crucial. It allows business teams to spend more time thinking about how to improve their businesses, instead of constantly "putting out fires." Imagine a team that, instead of spending hours answering frequently asked questions, can focus on developing new strategies, exploring markets, or even simply enjoying a well-deserved break.

Looking towards the future with optimism and preparation.

So, while I continue to dream of the day my own Jarvis prepares my breakfast, I invite you to reflect: How can LLMs transform your business or your daily life? In Latin America, the opportunity to integrate artificial intelligence into our companies is no longer a luxury, but an accessible and promising reality. The question is no longer whether we should adopt this technology, but how and when we will do it.

The future is here, and it is full of possibilities. Let's ride this wave of innovation and transform not only our businesses but also our lives.