Artificial intelligence can broaden access to answers, but it does not distribute equally the tools needed to understand them, question them, and build on them. In a world of abundant information, logic, mathematics, science, and data analysis can determine who uses the machine to think more effectively—and who merely accepts what it provides.

The same tool does not produce the same opportunities

Two people can access the same artificial intelligence and obtain completely different results. Not because one knows a secret command, but because they have distinct approaches to formulating problems, comparing information, interpreting data, and recognizing when an answer seems plausible but doesn’t hold up. The next educational inequality may arise precisely from this gap.

Access to AI may become widespread without its benefits being distributed equally.

Those with a broader repertoire are able to make connections, adjust criteria, explore possibilities, and turn an answer into new questions. The difference lies not only in access to technology, but in what each person is able to bring to the table when faced with it.

To ask better questions, you need to know more

Good questions don't come from curiosity alone. You need to:

• To know in order to recognize a contradiction,
• Comparing quantities,
• To be skeptical of a conclusion,
• Admitting that doesn't make sense.

Therefore, teaching students to ask questions does not mean downplaying the importance of content. The broader the repertoire, the wider the range of possible questions.

See also:

• When the Future Became Too Small
• Fact or fiction? The challenge of seeking the truth in the age of deepfakes
• Law No. 15,436/2026: A New Approach for Highly Gifted and Talented Students

The machine talks like a person, but it works according to a different logic

The interface is simple: we type in a question and receive an answer in natural language. Behind the scenes, however, there is data, probabilities, patterns, weights, mathematical models, and statistical decisions.

The challenge in education is not to teach students to converse with AI as if there were a person on the other end. It is to help them understand why a convincing answer might be wrong, why an average can hide differences, why correlation does not imply causation, and how graphs, samples, and criteria can alter a conclusion.

The more natural the conversation with the machine becomes, the more important it is to understand the logic behind how it generates responses.

Stiglitz: Access to information does not mean equal opportunity

Joseph Stiglitz, a Nobel Prize-winning economist, helped demonstrate that differences in access to and understanding of information lead to economic and social advantages.

AI can bridge part of that gap by expanding access to answers. But a new asymmetry may remain: the difference between those who receive information and those who are educated enough to interpret it, question it, and turn it into an opportunity. Inequality may simply shift to a different area.

PISA and School Accountability

The PISA standards already treat mathematics and science as ways to interpret problems, analyze data, work with evidence, and make decisions—not merely as content to be memorized. With the inclusion of media and computational literacy in PISA 2029, this connection becomes even more evident.

Working with AI is directly linked to:

• reading,
• logic,
• mathematics,
• statistics,
• science,
• research,
• argumentation,
• creativity
• ability to compare.

The role of the school will not be merely to teach students how to get better answers. It will be to ensure that they have enough knowledge not to accept just any answer and, above all, to use the system to get the most out of it. 

About the author:

*This text does not necessarily reflect the opinion of Bett Brasil.

Share on social media: