Technological development is currently progressing at a rapid pace. In this modern era, technological sophistication is no longer limited to its use in strategic areas; society can now easily and massively access and utilize the convenience that technology provides.

Technology is no longer merely an input receiver. Now, through the development of Artificial Intelligence (AI), technology actively serves as an interactive "discussion partner" capable of supporting our performance. With its capability to generate varied responses relevant to user needs and context, AI has now permeated not only the world of education but also the financial sector, healthcare, and customer service.

But what is the secret behind AI's ability to be a "discussion partner" that masters various fields and communicates effectively? In this article, our friends at UNDIRA will get to know the part of AI that enables it to "speak directly" to users: the Large Language Model, or LLM.

In short, the LLM itself is fundamentally a component of Machine Learning within a program. An LLM is a language model trained using a massive amount of data sourced from various users and origins. This language training is based on enabling the AI to understand and generate natural responses from various types of content provided.

LLMs function effectively thanks to a specific "artificial neural network" architecture called the Transformer. The basic working principle of an LLM is relatively simple. Imagine the LLM as a librarian who has read and understood vast amounts of research and study content. When someone asks a question, the librarian can skillfully provide a clear answer corresponding to the query—this is the fundamental working principle of an LLM.

The implementation of LLMs is already widespread in daily life. Some of the current benefits of LLMs include:

1. Boosting Productivity

The use of LLMs like ChatGPT, Copilot, and Gemini as aids in translating and simplifying text can directly enhance productivity. For example, if there is a relatively thick research paper or document with complex context, the presence of LLM technology in AI applications can simplify the understanding of the text in a short amount of time.

2. Generating Code for Programming

In the world of programming, various code compositions correspond to intended commands and functions, but not every formula can be memorized by an individual. With LLMs that possess the resources and understanding of programming languages, their existence can accelerate the software development process as needed.

3. LLM as a Bridge for Global Accessibility

LLMs act as a bridge connecting various community groups, overcoming obstacles caused by language differences and accessibility challenges faced by people with disabilities. The implementation of LLMs is widespread across various services to ensure the context of content or conversation is conveyed accurately: Google Translate is available for rapid language translation, while live caption technology on video conference or streaming platforms guarantees easy access for the Deaf community. Furthermore, applications like Be My Eyes demonstrate the role of LLMs in providing audiovisual assistance to individuals who are blind or have low vision.

Through their potential to manage and comprehend data sequences, the presence of the Large Language Model is a tangible solution for the modern world, not only to simplify tasks but also to unlock renewed innovation derived from human imagination that cannot be expressed simply. For UNDIRA friends who are interested in the world of computing and delving into Large Language Models, the Informatics Engineering Study Program is the best choice for you. Informatics Engineering at UNDIRA not only offers extensive laboratories but also features competent lecturers in the field.