Unplugged AI: When the Body Learns the Algorithm Before the Machine
Unplugged AI: When the Body Learns the Algorithm Before the Machine
From the Body to the Brain Bee: Decolonial Neuroscience for Latin American Teenagers
Maybe we need to begin talking about Artificial Intelligence without opening the computer.
Before the prompt, there is a question.
Before the question, there is a curious body.
Before the machine answers, there is someone trying to understand the world.
A classroom can learn AI with cards on the floor, students representing data, others making choices, others showing errors and corrections. No one needs to begin with the screen. We can first feel how an algorithm organizes, classifies, excludes, and decides.
This is what we call Unplugged AI: learning how artificial intelligence works through the body, conversation, doubt, and collective experience before giving our attention to the machine.
Revista Yvirá, from the UNESCO Chair in Science for Education, published an important reflection on the challenges and benefits of AI in education, highlighting teacher training, critical use, ethics, and understanding the limits of technology.
https://yvira.org/artigo/ia-os-desafios-e-beneficios-na-educacao/
The BrainLatam2026 question is simple:
how can we teach AI without turning students into followers of automatic answers?
AI should not replace the human question
Today, it may seem that thinking has become easier: we ask, and AI answers. But this comfort also brings risk.
The machine can suggest paths.
It can summarize texts.
It can organize ideas.
It can speed up tasks.
But if we do not notice what we are doing, AI can also capture our attention, our writing, our authorship, and even our desire to think.
So, before teaching teenagers only how to “use AI,” we need to teach them to ask:
Who trained this system?
With which data?
Who was left out?
Which culture became the standard?
Which territory was ignored?
Who benefits when I accept this answer without thinking?
UNESCO published guidance in 2023 on generative AI in education and research, defending a human-centered approach with attention to privacy, ethics, safety, equity, and age-appropriate use.
https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research
In our language, this means:
AI needs to expand Fruition and Metacognition, not replace the body that thinks.
Fruition is noticing what happens inside us while we learn.
Metacognition is observing how we are thinking, choosing, making mistakes, or being guided.
Without this, AI becomes just another intelligent screen.
Teacher training is also Jiwasa
We talk a lot about preparing students for AI. But who supports teachers in this process?
No teacher should be left alone in front of a technology that changes all the time. Teacher training cannot be just a quick tutorial. It needs to be a space for questions, practice, doubt, ethics, and collective support.
Here comes Jiwasa:
no one regulates alone all the time.
Not the student.
Not the teacher.
Not the school.
Not society.
Piauí became a reference by implementing Artificial Intelligence education in the state public school system, reaching thousands of students in Basic Education. This kind of initiative shows that AI can enter public schools, but the question remains: will it form obedient users or young people capable of critically understanding the systems that organize their lives?
https://www.consed.org.br/noticia/piaui-e-destaque-internacional-com-projeto-de-ensino-de-inteligencia-artificial-nas-escolas-publicas-estaduais
The body before digital capture
Unplugged AI matters because it removes technology from the place of magic.
When a student represents an algorithm with their own body, they understand that the machine does not “know” by itself. It follows patterns. It depends on data. It makes mistakes. It repeats biases. It can exclude people, languages, territories, and ways of life that do not appear in the data.
A 2025 study showed that unplugged AI activities, combined with digital activities, can improve students’ perception and self-efficacy regarding AI.
https://journals.sagepub.com/doi/10.1177/21582440251336708
This confirms something we already feel:
the body understands before abstraction.
Before the digital tool, the student can experience the algorithm as a process. They can see that every automatic decision depends on human choices.
This is the decolonial shift: AI stops being a center to obey and becomes a system to question.
APUS: AI also needs territory
The question is not only “how do we teach AI?”
The question is:
how do we teach AI without removing the student from their own territory?
A Latin American teenager does not arrive at AI as a neutral body. They arrive with language, accent, school, city, stable or unstable internet, a good or limited phone, family, culture, inequality, desire, and history.
This is APUS: the body-territory. Technology never enters an empty space. It enters a concrete life.
A decolonial AI needs to ask:
which bodies does it recognize?
which languages does it understand?
which schools does it consider?
which peripheries does it ignore?
which Indigenous peoples does it treat as exceptions?
which teenagers does it help to think — and which ones does it simply teach to obey?
From the Brain Bee to the scientific question
If a teenager reads this text and becomes interested in neuroscience, we already have a good question:
what changes in the brain, attention, and body when we learn AI first with the body and only then with the machine?
A BrainLatam2026 study could compare three groups:
students learning AI directly through a digital tool;
students learning AI first through unplugged bodily activities;
students learning AI through body, digital tools, and ethical debate.
We could measure understanding, perception of bias, sense of authorship, excessive trust in AI, and the quality of the questions students produce.
In more advanced research, we could use EEG, fNIRS, eye-tracking, respiration, GSR, and HRV/RMSSD to observe attention, cognitive effort, and body regulation during AI use.
The hypothesis would be:
when the body learns the algorithm before the machine, the student gains more autonomy to use AI without being used by it.
DREX Cidadão: AI as a common good
If AI is going to enter education, it cannot be a privilege only for those who can pay for the best platforms.
Here, DREX Cidadão appears as a metaphor for public metabolism: each student needs a minimum level of social energy to participate in collective intelligence.
This means connected schools, trained teachers, data protection, infrastructure, fair access, and critical literacy.
AI in education without social justice can increase inequality.
AI with belonging, intelligent State action, and the common good can open paths.
Not to replace teachers.
Not to replace the body.
Not to replace territory.
But to expand questions, create bridges, and return knowledge to the common good.
Closing
Before AI, there is the body.
Before the answer, there is the question.
Before the tool, there is authorship.
Before the machine, there is community.
Unplugged AI reminds us that technology does not need to begin on the screen. It can begin in a circle, with cards, with students representing data, with a teacher asking: “who was left out of this model?”
Maybe this is the central point for Latin American teenagers:
we do not want young people only trained to use AI.
We want young people able to think with AI without giving away their attention, their body, and their future.
Jiwasa returns here:
no one learns AI alone all the time.
We learn together.
We ask together.
We regulate together.
And maybe, before the machine, this is what still makes us human.
Post-2021 References
Yvirá / UNESCO Chair in Science for Education. AI: the challenges and benefits in education.
https://yvira.org/artigo/ia-os-desafios-e-beneficios-na-educacao/
UNESCO. (2023). Guidance for generative AI in education and research.
https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research
Consed. (2024). Piauí gains international recognition with Artificial Intelligence education project in state public schools.
https://www.consed.org.br/noticia/piaui-e-destaque-internacional-com-projeto-de-ensino-de-inteligencia-artificial-nas-escolas-publicas-estaduais
Ruan, J. et al. (2025). Developing Artificial Intelligence Literacy Through Mixed Unplugged and Plugged-in Activities in Primary Education. SAGE Open.
https://journals.sagepub.com/doi/10.1177/21582440251336708
European Commission / OECD. (2025). AI Literacy Framework for Primary and Secondary Education.
https://education.ec.europa.eu/event/empowering-learners-for-the-age-of-ai-launch-of-the-draft-ai-literacy-framework-and-stakeholder-consultations
