Artificial Intelligence Recognizes Patterns. But Who Chose What You Never Had the Chance to Choose?

Explainable AI, APUS, digital sovereignty, and Body-Territory.

Have You Ever Lost an Opportunity You Never Knew Existed?

Perhaps a scholarship. A medical treatment. A right you could have exercised. A profession you never imagined pursuing.

You did not reject that opportunity.

You did not compare alternatives.

You did not decide that another path would be better.

It simply never appeared before you.

Now imagine discovering, years later, that the opportunity had been available all along.

What would you feel?

We do not need to agree on the answer.

But perhaps this experience reveals something we rarely consider when discussing freedom: we do not choose among all existing possibilities. We choose among those we manage to perceive.

At BrainLatam, we propose a provocative idea:

You are standardized by the choices you never had.

What if artificial intelligence, instead of expanding our world, is learning to offer us only what our previous behavior makes predictable?

The Machine Knows What You Might Do. But Does It Know What You Could Become?

Imagine a platform that knows ten years of your choices.

It knows what you read, buy, watch, and search for.

Using these records, it calculates which content will probably interest you.

Perhaps it is right most of the time.

You receive convenient recommendations. You save a few minutes. You keep scrolling.

But what happens to everything the system considers unlikely?

A different book, Latin American research, or a professional opportunity far removed from your previous experiences?

The platform can simply stop presenting these alternatives.

It does not need to prohibit anything.

The possibility disappears before reaching your attention.

An important distinction must be preserved: not all personalization reduces autonomy. Recommendations can also reveal previously unknown opportunities.

The scientific question is when algorithmic selection expands our repertoire and when it begins to narrow it.

Neurobucano: Recognizing Patterns Is Not the Same as Knowing Someone

In our scientific dialogue with Neurobucano, we bring together the contributions of Tsang Ing Ren, from the Federal University of Pernambuco, and Edward Hermann Haeusler, from PUC-Rio.

Machine learning makes it possible to identify complex relationships in images, texts, and data.

Explainability seeks to make aspects of the results produced by these systems understandable.

This research is essential.

But imagine an algorithm identifying a pattern in your history and concluding:

"You are unlikely to succeed in this activity."

That is a prediction.

When the prediction begins determining which opportunities will be presented to you, something else has happened.

The system is no longer merely recognizing patterns.

It is participating in the conditions of your future choices.

Who authorized that transition?

APUS: The Possible World Is Larger Than the Perceived World

At BrainLatam, we call APUS the perceived field of possible movements.

It is a conceptual framework we propose investigating, not an already validated neuroscientific measurement.

Imagine three situations.

An opportunity exists, but you do not know about it.

You know about the opportunity, but lack the resources to pursue it.

You know about the opportunity, identify the obstacles, and find ways to overcome them.

External reality may contain possibilities that have not yet entered your perception.

And perceiving a possibility does not mean having the material conditions to realize it.

Here we encounter a decisive question for Personalissimo AI:

Should it merely show us the most probable path, or should it also help us discover paths we have not yet imagined?

Perhaps the value of public intelligence lies less in predicting our next movement and more in revealing movements we did not know were possible.

Who Is Spending the Time of Your Life?

Imagine needing a public service.

You search for information, interpret regulations, consult contradictory pages, and discover that you contacted the wrong institution.

Hours later, you finally identify the appropriate procedure.

Your organism has spent time, attention, and energy finding information that could have been organized in an understandable way.

That time is not abstract.

It is time from your life.

Now imagine a Personalissimo AI saying:

"This path is available."

"This alternative requires an additional condition."

"This information has not yet been confirmed."

"There is another option you may not know about."

"You can challenge this restriction."

It would not invent rights.

Nor should it decide on your behalf.

But it could reduce informational obstacles that currently separate people from real possibilities.

Why should every Body-Territory have to discover alone what public intelligence could already help make visible?

And what happens when that time returns to life?

Perhaps it becomes rest, care, creativity, or simply doing nothing.

We do not need to turn every recovered minute into productivity.

When an Explanation Can Also Hide a Choice

Edward Hermann Haeusler and Cleyton Slaviero investigate computational approaches to explaining artificial intelligence results.

Imagine an AI denying a request and presenting its reasons:

"Your data do not meet the established criterion."

The explanation may be technically correct.

But who selected that criterion?

Are your records complete?

Is another procedure available?

Can you challenge the decision?

In Brazil, Article 20 of the General Data Protection Law (LGPD) establishes rights relating to the review of certain automated decisions and access to information about their criteria, subject to legal conditions.

Explainability helps us understand how a result was produced.

But it does not, by itself, establish the legitimacy of using that result.

At BrainLatam, we distinguish:

Experiment → Result → Interpretation → Judgment → Decision.

An AI identifies patterns.

People and institutions remain responsible for the criteria that transform those patterns into consequences.

DOING IS SCIENCE. JUDGING IS POLITICS.

What If the State Started With You?

Now imagine a transformation in the architecture of the State itself.

Instead of treating each person merely as a recipient of services, the State would recognize every Body-Territory as a constitutive and indivisible unit of its collective existence.

We are not proposing millions of legally independent sovereign States.

We are investigating another way of organizing participation.

In BrainLatam's proposal, the Jiwasa-State would recognize different needs, languages, histories, and possibilities without requiring them all to be reduced to interpretations made by economically powerful representatives.

Personalissimo AI would be a singular interface with this State.

We call this architecture fractal: relevant institutional capabilities would become accessible from each Body-Territory.

But the interface would not own the person's voice.

It should ask before acting, explain before recommending, and allow challenges.

The State would not need to interpret you first and then permit your participation. It could provide the conditions for you to participate in your own interpretation.

Reality Also Changes Between A and B

Remember our discussion about complex systems.

An organization can move from condition A to condition B.

But what happens during the transition also matters.

The same applies to you.

Your previous records describe real experiences.

However, they do not contain all the relationships you may still establish.

At BrainLatam, we propose a computational right to plasticity: preventing our past from becoming a statistically imposed limit on our future.

A Personalissimo AI should be able to recognize:

"Your previous data suggest this path, but other relevant possibilities exist."

This does not mean promising that everything is possible.

It means distinguishing a demonstrated impossibility, a temporary obstacle, and something we simply do not yet know.

An algorithm may know many of the selves we have been. It cannot determine in advance all the selves we may still Become.

But Who Controls This Intelligence?

Paola Ricaurte, Edgar Gómez-Cruz, and Ignacio Siles investigate how algorithmic governance relates to institutions and structures of power in Latin America.

State-operated AI can expand access to services.

It can also produce surveillance, exclusion, and new dependencies.

For this reason, Personalissimo AI would not become legitimate simply because it is publicly operated.

It would require data protection, independent audits, non-digital alternatives, authorization before taking actions, and identifiable institutional accountability.

It would also need mechanisms through which people could discover which alternatives its own recommendations had omitted.

An interface that merely replaces private control with centralized public control does not necessarily achieve the autonomy we seek to investigate.

Real Democracy: Choices That Do Not Yet Exist

Now imagine a community facing problems related to water, transportation, and food.

AI identifies patterns among requests.

But people with similar needs may imagine different solutions.

A statistical cluster does not automatically represent collective will.

In BrainLatam's proposal for Real Democracy, every Body-Territory could learn about alternatives, submit proposals, and challenge institutional interpretations.

Drex Cidadão and Rendimento Bioma add the discussion of material conditions necessary for participation.

These political and economic proposals still require legal assessment, funding, and governance.

Because knowing about an opportunity does not mean being able to pursue it.

And participating in the construction of the State requires more than receiving information.

It requires real conditions to act, disagree, and transform.

Artificial Intelligence Should Give the Future Back

Return to the opportunity that never reached you.

Perhaps it was available.

Perhaps it depended on conditions that could have changed.

Perhaps it was not appropriate for your life.

But you never had the chance to find out.

Now imagine an intelligence that does not decide in advance what deserves to enter your field of possibilities.

It presents paths, explains obstacles, recognizes limitations, and allows you to explore alternatives.

It does not need to capture your attention.

It does not need to fill the time it helped you recover.

It does not need to determine who you will become.

You are standardized by the choices you never had. But you can transform your relationship with the world when previously invisible possibilities become perceptible, accessible, and open to challenge.

Personalissimo AI should not exist to decide our future from past patterns.

It should help every Body-Territory discover paths that its own models cannot yet predict.

Let it come. Let it stay. Let it Be. You may disobey. Let it go.

Jiwasa: Being the State without ceasing to Be.

Latin American Scientific and Intellectual References 

The references follow the article's argument: from patterns identified by machines to possibilities made invisible, from explainability to contestation, and finally to public intelligence that respects every Body-Territory's capacity for change.

1. REN, Tsang Ing — Federal University of Pernambuco. Research in Artificial Intelligence and Machine Learning.

Machine learning identifies regularities and generates predictions, but its performance depends on the datasets and conditions under which models are developed and evaluated.

2. SLAVIERO, Cleyton; HAEUSLER, Edward Hermann (2025). An Approach to Explainability in AI with Abduction.

Computational explainability contributes to understanding algorithmic results, but a technical explanation alone does not establish the legitimacy of the resulting decisions.

3. BRAINLATAM — CIONEK, Jackson (2026). O Big Data não é o mundo (Big Data Is Not the World).

Every model operates on selected aspects of reality shaped by methodological decisions; what was not recorded must not be confused with what does not exist.

4. BRAINLATAM — CIONEK, Jackson (2026). O cérebro não recebe a realidade pronta - ele compara (The Brain Does Not Receive Reality Ready-Made — It Compares).

The APUS concept connects perception, expectations, and possibilities, asking how technological systems influence which paths become visible to us.

5. BRAINLATAM — CIONEK, Jackson (2026). Quando conhecer demais pode retirar autonomia (When Knowing Too Much Can Take Away Autonomy).

Algorithmic personalization can facilitate choices while silently restricting alternatives that do not match previous behavior.

6. BRAINLATAM — CIONEK, Jackson (2026). O direito de mudar o dado sobre mim (The Right to Change the Data About Me).

The proposed computational right to plasticity distinguishes legitimate records of the past from permanent restrictions on future possibilities.

7. BRAINLATAM — CIONEK, Jackson (2026). Uma IA para cada Corpo-Território (One AI for Every Body-Territory).

Personalissimo AI would also be evaluated through obstacles removed, paths revealed, and time returned to life.

8. RICAURTE, Paola; GÓMEZ-CRUZ, Edgar; SILES, Ignacio (2024). Algorithmic Governmentality in Latin America: Sociotechnical Imaginaries, Neocolonial Soft Power, and Authoritarianism. Big Data & Society.

These Latin American researchers examine how automation can reinforce existing structures of power, demonstrating the need to discuss governance and participation beyond algorithmic efficiency.

9. RICAURTE QUIJANO, Paola (2025). Algorithmic Assemblages of Power: AI Harm and the Question of Responsibility. Teknokultura.

The impacts of artificial intelligence must be examined through relationships among institutions, markets, communities, and knowledge systems, including mechanisms of accountability.

10. BRAINLATAM — CIONEK, Jackson (2026). IA Personalíssima - a interface do Corpo-Território com o mundo digital.

The proposed personal interface would help people understand their rights, communicate authorized requests, and expand their possibilities without replacing individual autonomy.

11. BRAINLATAM — CIONEK, Jackson (2026). Estado-Jiwasa - e se a democracia pudesse emergir de quem a vive?.

Body-Territory is presented as a constitutive unit of the State, supporting the hypothesis of continuous political participation without eliminating individual differences.

12. BRAINLATAM — CIONEK, Jackson (2026). Drex Cidadão - Quem recebe o dividendo dos dados?.

The discussion of digital monetary infrastructure adds the material conditions needed to exercise rights to the informational possibilities made available by AI.

BRAINLATAM | NEUROBUCANO | W46/2026 | BLOG 07/08