Experts participating in the IA Conference Brasil, an event organized by Alura, analyze the latest developments in technology that are expected to grow next year
According to a Google study, 46% of Brazilians used Artificial Intelligence (AI) in some way in the last 12 months. In addition, McKinsey reinforces the role of this technology in today's society in the corporate sphere, revealing that 72% of national companies have already used it in their operations in 2024, which represents an increase of 17% compared to the previous year.
For Fabrício Carraro, Program Manager at alura, the largest ecosystem of education in technology, this growing integration of the tool into everyday life and organizations demonstrates its potential to transform the country.”We are no longer talking about a futuristic concept that arouses fear, but rather a reality present in our daily lives.", it says.
In view of this, the executive and other participants in the 2024 edition of IA Conference Brazil, the largest AI event in Brazil, which brought together the main references and leaders in this market, wants to prepare Brazilians for what is yet to come. Therefore, the experts listed the 4 biggest AI trends for 2025. Check it out:
- Integrating Generative AI into everyday work in different careers
A projection from Bloomberg Intelligence (BI) reveals that the Generative AI market will reach US$1.3 trillion by 2032. According to Jomar Silva, NVIDIA's developer relations manager for Latin America, Graphics Processing Units (GPUs) play a fundamental role in this growth process, as does high-performance hardware for creating solutions based on large language models (LLMs).
“The analysis of the complexity, timeframe and cost of the alternatives for implementing LLMs demonstrates the growing interest and investment in the development of this technology, consolidating it as a key element for innovation in various sectors in the coming years”, explains the specialist.
Guilherme Silveira, co-founder of Alura, reinforces this reasoning: “Connecting computing, creation and inspiration seemed unimaginable a few years ago, and Generative AI has proven to be a viable path to new possibilities”, he adds.
- Open Source LLM Models
As Generative AI gains scale and investment, a parallel movement is gaining momentum: the open-source provision of Large Scale Language Models (LLMs). This openness allows developers and researchers from around the world to access, modify, and contribute to the evolution of these tools. A notable example is Llama 3.1. With enhanced natural language processing and text generation capabilities, the model becomes a powerful tool for a variety of applications, from chatbots and virtual assistants to translation and content creation.
Alberto de Paola, Partner Engineer at Meta, emphasizes that open source models promise to democratize access to cutting-edge technologies, accelerating the creation of innovative solutions in different sectors. “By opening the LLMs code, we are opening up a range of possibilities so that brilliant minds around the world can contribute to the evolution of this technology and create innovative solutions that benefit us all,” he comments.
- IAOps: AI Operations
“IAOps” or “AI Operations” is a set of practices and resources that aim to optimize the complete life cycle of AI systems development and operation, enabling companies to develop, deploy and operate AI models in an efficient, reliable and scalable way. Therefore, it is a focus of attention for several segments, since it reduces costs and increases process agility.
Fabio Guimarães, an IT specialist at Banco do Brasil, says that these practices can generate real value for the business, but they require a holistic vision and a strategic approach to be maintained. “It is possible to integrate DataOps, ModelOps and DevOps tools to ensure efficiency, quality and scalability. And, to top it off, the integration of different areas, such as FinOps, helps to build a robust and sustainable platform,” he highlights.
- SQL Copilot
Finally, it is worth noting that design, optimization, monitoring and troubleshooting tasks related to Microsoft Azure's Structured Query Language (SQL) database, SQL Databases, now have a powerful ally: SQL Copilot. Antonio Claudio Michejevs Padilha, Data & Applied Scientist at Microsoft, explains that this LLM-based assistant has the potential to help with these and many other demands.
“It is a technology capable of collecting details about your environment, such as configurations, historical query data and database schemas to assist in diagnostics and answer general questions, as well as creating scripts for the user”, he concludes.
Future of AI
And thinking long term, not just 2025, experts highlight that there is still a technology that will also be a trend, but in a more distant future.
- Quantum Machine Learning (QML)
Quantum computing uses the principles of quantum mechanics to perform calculations, which allows information to be processed differently from classical computing. In other words, the algorithms in this universe must continue to develop by exploiting phenomena such as superposition and entanglement to promote complex operations on data.
“The combination of quantum computing with advanced Machine Learning techniques is revolutionizing areas such as optimization and data processing, generating several opportunities for the future of technology”, highlights Ana Paula Appel, AI Engineering & Quantum Ambassador at IBM.
EN 
PT 
