Generative artificial intelligence has profoundly changed the way people perform tasks and conceive ideas. Currently, the technology is in the phase of 'AI agents,' where it executes actions for us, surpassing mere guidance. The effectiveness of this execution depends on variables such as usage cost, the quality of the provided instruction, and how the tool was developed.
Following the exponential growth of ChatGPT in late 2022, the portal Olhar Digital interviewed several experts to analyze the consequences of AI in daily life.
Roberto “Pena” Spinelli, a physicist from USP and a Machine Learning specialist from Stanford University, classified AI as 'the most powerful and impactful technology of all time.' He observed that the assistant phase, which occurred between 2022 and 2025, was just the beginning; the true advancement would come in 2026, when these AIs become capable of performing tasks directly on the computer.
Pena pointed out that while AI eliminates cognitive friction and simplifies complex processes, there is an inherent risk. However, he also argued that the intellectual use previously exclusive to humans forced the development of the brain and cognitive skills. He warned that by removing this workload, a cognitive deficit is natural to occur.
Geovanni Borsetto, a communication and marketing specialist at Paschoalotto, stated that the advance of generative AI—that which produces texts, audio, and images—'changed the starting point of almost everything.' According to him, generative AI drastically reduced the initial difficulty of any project, such as overcoming writer's block or creating the first draft.
When asked whether this shortening leads to mental stagnation, Borsetto replied that the tool does not make anyone stupid, but rather reveals each individual's way of thinking. He highlighted that those who already practice critical thinking use AI to enhance their ideas, freeing up operational time to focus on strategy. Conversely, those who seek shortcuts without thinking risk atrophying their mental muscles.
Pena reinforced this idea, suggesting that if AI is used as an intellectual exchange partner, it can broaden reasoning and allow for the understanding of complex subjects in less time. He compared this to having an 'Einstein friend in your pocket,' but warned against the temptation to let it solve everything.
The general conclusion of the experts is that generative AI is merely a resource, neither intrinsically good nor bad. Dr. Heitor Felipe, a physician and neurology professor at Afya Educação Médica Goiânia, listed clear benefits, such as vast reach and high capacity to optimize numerous processes. As disadvantages, he cited the danger of excessively delegating to the tool, which can result in the loss of a sense of agency.
The doctor also emphasized that some type of mental effort is essential for learning, and handing over this effort entirely to AI can harm the process. Borsetto added that AI has introduced shortcuts in both the workplace and studies, areas that traditionally demanded cognitive effort. He stressed that the effort dedicated to structuring reasoning is what guarantees the real understanding of a topic, and it will be necessary to 'reintroduce purposeful effort,' similar to what happened with nutrition.
When questioned about the ideal use of AI, the experts presented distinct guidelines. Erich Silva, an automation project specialist and COO of Lecom, advised using AI with a defined objective, remaining active in the reasoning process. He recommended using it to save time, but always reviewing the results, and asking the tool to point out flaws instead of just confirming ideas.
Geovanni Borsetto defined healthy use as treating AI as a thinking partner, not an oracle. It should be used to streamline operational tasks and test concepts, but keeping authorship and final decisions in the user's hands. He exemplified that by refining a draft generated by AI with one's critical eye and personal context, the result is more productive and refined, transforming AI into a multiplier for those who already possess good thinking skills.
Roberto “Pena” Spinelli concluded by suggesting that, in situations like an interview, he could use AI to debate his own ideas, find new perspectives, question biases, and refine thoughts, obtaining faster and better answers while improving his own cognitive capacity and expanding his worldview.