On Wednesday, April 10, the IEAT 2025 Residents’ Exhibition began, promoted by the Institute of Advanced Transdisciplinary Studies (IEAT) at UFMG. The opening ceremony featured a presentation by Professor Marcelo Azevedo Costa, from the Department of Production Engineering.

With the theme Artificial intelligence, data science and people: the challenges of managing productive and resource sectors in the digital age and in the era of extreme factors, the presentation addressed how emerging technologies, such as artificial intelligence and data science, can transform the management of strategic sectors, such as energy and water resources.

Marcelo Costa began his speech with an interactive proposal, inviting the audience to introduce themselves and reflect on their physical and emotional well-being. The dynamic, inspired by practices used in technology laboratories in which he participates, aims to humanize relationships in work and research environments, in addition to encouraging listening and interaction between teams.

The professor emphasized that strategies like this are part of the planning of experiments, which are specific actions, aimed at local contexts and with reduced groups, but with potential for expansion. “It’s about thinking of solutions that, if successful, can be applied in larger communities,” explained Marcelo when detailing a similar experience he had during his postdoctoral studies at Linköping University, in Sweden, in 2018. During the presentation, Marcelo also discussed the use of statistical models in decision-making processes, warning of the need to consider human and contextual variables and not treat statistical results as absolute truths.

An enthusiast of artificial intelligence, but also a critic, Marcelo ended his presentation by talking about the nature of artificial intelligence, which, as a statistical model by nature, uses the average to summarize a set of data. According to him, it is promising that artificial intelligence and data science can assist in decision-making in strategic areas, such as energy supply and water management, but caution is needed. “Let’s take, for example, large-scale language models (LLMs) that simulate conversations, develop responses and replicate human language. “In fact, LLMs, which became popular with ChatGPT and DeepSeek, for example, are large machine learning models that feed on huge volumes of data, but are subject to biases and failures, so they should always be used critically and with discretion,” he emphasizes.

The IEAT 2025 Residents Showcase continues with presentations by the other professors selected for this year’s program. The full program is available below:

23/04/2025 – 09h30

Eduardo Magalhães Ribeiro – Institute of Agricultural Sciences
Topic: Family farming, water and drought in the semi-arid region of Minas Gerais

07/05/2025 – 09h30

Bruno Souza Leal – Social Communication (School of Philosophy and Human Sciences)
Topic: Everyday catastrophes: analytical explorations of the articulations between textualities, events and temporalities

28/05/2025 – 09h30

Glaucia Muniz Proença Lara – Linguistics (School of Arts)
Topic: Migrants and refugees in Brazil: between life narratives, laws and public policies

04/06/2025 – 09h30

Fabrício Rodrigues dos Santos – Genetics, Ecology and Evolution (Institute of Biological Sciences)
Topic: Science of Brazilian natural and cultural heritage

18/06/2025 – 09h30

Luciana Diniz Silva – Medical Clinic (School of Medicine)
Topic: Historical, cultural and social analysis of the body image obtained from the collection of the Memory Centers of the Federal University of Minas Gerais: eugenics and resistance to ethnic diversity