On the Genealogy of Temptation: From Simulmatics to Data Sovereignty

Fundacja Dobre Państwo • January 18, 2026 • 🇵🇱 Polski

Introduction

This article examines the evolution of social science ambitions—from pioneering attempts to model behavior by the Simulmatics Corporation in the 1960s to contemporary predictive systems. Jill Lepore calls this company the "missing link" connecting past psychological warfare with today's algorithmic economy. The text illustrates how the attempt to translate decisions into the language of rules changed the perception of human beings: from subjects into objects reacting to stimuli. You will learn how the concept of attention as a resource was born and why the modern struggle for data sovereignty in the EU is crucial for the future of democracy.

Simulmatics: The Missing Link and the People Machine

In the 1960s, Simulmatics defined attention as a finite and modelable resource—a "limited-capacity warehouse." This is where IF/THEN logic was born, in which human behavior is described by a set of precise rules, and free will becomes merely a measure of missing data and computing power. The "what-if men" created an anthropology of the subjunctive mood, where the future becomes a calculation of variants.

The institutional expression of this temptation was the People Machine. It was used in the Kennedy campaign, segmenting the population into 480 voter types. Instead of civic debate, a market of reactions was proposed, where politics becomes a mechanism for managing exposure to stimuli. This reification of the structure of social consciousness meant that democracy began to be perceived as a space for optimization rather than mutual understanding.

Rule-Based Logic, Machine Learning, and the Profit from Prophecy

There is a fundamental difference between classic rule-based logic and modern machine learning. At Simulmatics, rules were explicit and human-authored. Today, a rule is a statistical result emerging from data correlations. This leads to a phenomenon known as profit from prophecy. The business model does not just study demand; it creates it—precise exposure to information shapes preferences, guaranteeing the self-fulfillment of the forecast.

Failure in Vietnam exposed the methodological flaws of such simulations. Systems processing morale data generated "daisy-chains" of misinformation—data was falsified by fear and careerism. This confirms Goodhart's Law: when a measure becomes a target, it ceases to be a good measure. When the engagement metric becomes a tool for profit, the system begins to generate conflict and addiction rather than genuine user interest.

The European Parliament: The Dispute Over Data Sovereignty

The dispute over algorithmic access to data outside the EU is a clash of visions for Europe as both a regulator and a participant in the global game. The EPP takes a pragmatic approach, accepting the dominance of cloud giants while maintaining contractual control. Conversely, the Socialists and Democrats (S&D) view data as a public resource and demand hard sovereignty to protect citizens from the asymmetry of predictive power.

The Renew Europe group focuses on algorithmic transparency and interoperability, rejecting digital protectionism. Meanwhile, the Greens propose a "data ecology," warning against the colonization of attention and the loss of the capacity for self-determination. For them, global clouds are infrastructures of cognitive exploitation that evade European standards of accountability. This debate shows that data is not just technology, but the foundation of modern power.

Unpredictability: The Foundation of Democratic Freedom

In a world dominated by algorithmic optimization, we must rehabilitate unpredictability as a value rather than a defect. A completely computable society loses the ability to critique its own rules and engage in normative learning. Can we regain control over the "attention warehouse" before algorithms completely take over our reality?

Defending the right to error and spontaneity is not nostalgia, but a condition for the survival of an authentic democratic community. The chance to preserve human agency lies in recognizing that a certain margin of unpredictability is essential for us to remain a community capable of self-improvement, rather than just a collection of types reacting to programmed stimuli.

📄 Full analysis available in PDF

Summary

This text provides a deep historical and technological analysis of the mechanisms for modeling human behavior. Beginning with the pioneering work of Simulmatics Corporation and its 'People Machine' project, the author guides the reader through the evolution of the algorithmic prediction economy. The article unmasks the transition from rigid IF/THEN logical rules to contemporary, learning AI systems that monetize human attention. It addresses key issues of data pathology, such as daisy chains and Goodhart's law, which lead to the degradation of social indicators. The finale focuses on the European struggle for data sovereignty, the role of cloud computing, and the political aspects of privacy protection in the era of transatlantic technological disputes. This is a study of how technology is transforming human ontology and the structure of power.

Frequently Asked Questions

What was the People Machine?

It was a tool created by Simulmatics that segmented voters into 480 types to predict their reactions to the Kennedy campaign's political messages.

Why is attention treated as a 'warehouse of finite volume'?

Early Simulmatics models assumed that human perception had a limited capacity, and effective modeling relied on calculating what was currently occupying that space.

What is the main difference between ancient and modern algorithms?

Old systems relied on rigid IF/THEN logic rules entered by experts, while modern algorithms generate rules themselves based on data.

What is the phenomenon of 'profit by prophecy' in the algorithmic economy?

It involves the fact that algorithms not only predict future behaviors, but also actively create them by precisely controlling exposure to information.

What are the risks associated with indicators such as 'morale' or 'engagement'?

When complex human emotions are reduced to parameters, systems begin to optimize for metrics (e.g., clicks) while ignoring real social consequences like addiction or conflict.