The Artificial State: between intelligent administration and automatocracy in light of The Rise and Fall of the Artificial State

• • 🇵🇱 Polski
The Artificial State: between intelligent administration and automatocracy in light of The Rise and Fall of the Artificial State

📚 Based on

The Rise and Fall of the Artificial State

👤 About the Author

Jill Lepore

Harvard University

Jill Lepore (born August 27, 1966) is an acclaimed American historian, author, and journalist. She serves as the David Woods Kemper '41 Professor of American History at Harvard University and Professor of Law at Harvard Law School, while also writing regularly as a staff writer for The New Yorker. Lepore holds a doctorate in American Studies from Yale University. Her scholarship spans American political history, law, literature, and the historical methodology and technology of evidence. Lepore is renowned for bringing rigorous historical analysis to broad public audiences. Her seminal works include the Bancroft Prize-winning The Name of War, The Secret History of Wonder Woman, and These Truths: A History of the United States, an influential single-volume history of America. A two-time Pulitzer Prize finalist, her work critically investigates democratic institutions, archival absences, and societal transformations.

Introduction

Modern administration is undergoing a subtle transformation. This is not a robot revolution from science-fiction films, but rather a process of the algorithmization of power.

This article analyzes the concept of the Artificial State, in which the agency of the citizen is replaced by a digital profile. You will learn why reducing humans to data threatens freedom and how to distinguish intelligent administration from automatocracy.

Understanding this mechanism is crucial for protecting democratic foundations in the era of AI.

The Artificial State as the Reduction of Humans to Data

The Artificial State is not a government of machines, but a system in which institutions treat the output of a model as a substitute for independent reasoning. It is a systemic transformation involving the transfer of cognitive functions to algorithms.

The core problem is the reversal of the relationship between human and tool. Instead of adapting technology to people, we are reconstructing humans into a form that is easier for machines to process.

The citizen ceases to be a subject and becomes a record and a prediction. An example of this occurs when an official accepts a scoring result without analyzing the context, thereby forfeiting their role as the normative author of the decision.

Algorithmic Prediction as a Threat to Agency and Freedom

Algorithmic profiling threatens freedom because it treats a probabilistic description of a person as the absolute truth about them. This creates a self-fulfilling prophecy: if a model limits someone's chances of success, the subsequent lack of success is viewed as confirmation of the model's accuracy.

Such reduction strikes at the human capacity for change and new beginnings. Predictive systems preserve the past, whereas political freedom relies on the unpredictability of action.

It is necessary to introduce a right to biographical unpredictability. A citizen cannot be perpetually defined by an image of their own past recorded in data.

Separating Epistemic Competence from Political Legitimacy

The greater analytical efficiency of AI does not grant it the right to make binding decisions. There is a fundamental difference between knowledge regarding the means-end relationship and the legitimacy of the end itself.

The decisions of scoring algorithms are not objective; every optimization contains a hidden political decision regarding an acceptable level of error. AI can calculate a trade-off, but it cannot determine the moral cost of a mistake.

Simply keeping a human in the loop is often a fiction due to automation bias. Real control requires effective contestability—the ability to challenge a result without facing organizational risk.

Conclusion

The greatest threat is not a machine rebellion, but the acceptance of a model of a human as a sufficient substitute for the person. In such a scenario, the formal institutions of democracy remain in place, but they lose their real influence over reality.

We must defend the right to be more than the sum of our data. Protecting agency requires the re-constitutionalization of the human-machine relationship and the primacy of judgment over recommendation.

Only then will AI become a tool that supports the citizen, rather than an architecture to which humans must uncritically conform.

Mind map: The Artificial State: Between Administration and Automatocracy

📖 Glossary

Automatokracja
System, w którym ludzie pozostają formalnie obecni w strukturach władzy, ale tracą realny wpływ na decyzje na rzecz systemów optymalizacji i predykcji.
Automation bias
Psychologiczna tendencja do nadmiernego ufania wynikom generowanym przez systemy automatyczne, nawet w obliczu dowodów wskazujących na ich błąd.
Prawo Goodharta
Zasada mówiąca, że gdy miara statystyczna staje się celem optymalizacji, przestaje być wiarygodnym wskaźnikiem mierzonej wartości.
Problem alignment
Wyzwanie polegające na zapewnieniu, aby cele i zachowania systemu AI były zgodne z intencjami operatora oraz wartościami etycznymi i prawnymi.
Government by scores
Model zarządzania, w którym merytoryczne uzasadnienie decyzji zostaje zastąpione przez wynik liczbowy (scoring) wygenerowany przez algorytm.
Bounded rationality
Koncepcja ograniczonej racjonalności człowieka, wynikająca z limitów pamięci i czasu, co czyni go podatnym na wsparcie systemów AI.

Frequently Asked Questions

What exactly is the Artificial State, and how does it differ from the vision of a government of robots?
The Artificial State is a subtle systemic transformation consisting of shifting the cognitive functions of institutions to systems based on classification, prediction, and optimization. Unlike the science-fiction vision of robot rule, it is not a system without humans, but one in which humans lose their function as normative authors of decisions, and their image is reduced to a form that is easier for a machine to process.
How does algorithmic profiling affect human freedom and the ability to change?
Algorithmic profiling can limit human freedom by preserving a person's past in the form of a probabilistic image, which becomes a self-fulfilling prophecy and hinders a new beginning. These systems may define an individual based on statistical regularities, which strikes at the individual's capacity for unpredictability and the ability to change their own views or behaviors.
Does the greater analytical effectiveness of AI give it the right to make binding political and legal decisions?
No, because epistemic competence (knowledge) is separate from political legitimacy. The cognitive advantage of AI does not grant it the right to assume a normative role and define values, as binding decisions require authorizations that technology itself does not possess.
Why does simply keeping a human in the AI decision-making process not guarantee democratic control over the system?
Simply keeping a human in the loop does not guarantee control due to automation bias—the tendency to consider system outputs more objective than one's own judgment. Real control is impossible when organizational conditions make rejecting an algorithm's recommendation costly.
Are decisions made by scoring algorithms objective and purely technical?
No, decisions made by scoring algorithms are not purely technical because they are based on a hidden political philosophy and accepted norms. An algorithm can calculate the trade-off between different types of errors, but it cannot independently determine their moral cost without a previously defined objective function.
How can AI systems be controlled to avoid axiological errors and technological dominance?
Controlling AI systems requires integrating the humanities into the oversight of objective functions to account for values that are invisible to formalization. It is also necessary to limit the arbitrariness of operators through clear rules and the right to appeal, as well as to implement a polycentric intelligence infrastructure, dispersing power among many autonomous centers.
What are the limitations of the Artificial State theory, and how can the impact of AI on institutional agency be empirically studied?
The limitations of the theory include, among others, an overestimation of Big Tech's coherence, the risk of mistaking similarity of ideas for causal links, and a lack of automation gradation based on the scale of impact on fundamental rights. The impact of AI on institutional agency can be empirically studied by analyzing measurable indicators such as the automation deference rate, institutional reversibility, infrastructural dependency ratio, and examining the actual possibility of challenging system recommendations (effective contestability).
How can the development of the Artificial State be stopped if it does not result from a single malicious will, but from systemic incentives?
Stopping the development of the Artificial State requires the creation of institutions and democratic laws capable of altering the structure of systemic incentives and coordinating the actions of market participants. It is essential to introduce an architecture for the separation of computational power through balancing mechanisms such as audits, competition law, the AI Act, or the control of computing power concentration.
What is the key threat to the citizen in the Artificial State system, and how can it be countered?
The key threat is the political reduction of the human being to a data format (a profile or probability), where a model of the person becomes a substitute for the human, and information transforms into control. This can be countered by protecting the citizen's right to be more than the sum of their data and ensuring the primacy of decision over prediction and the individual over the profile.

Related Questions

🧠 Thematic Groups

Tags:

More in: Szkatułka kosztowności

 Content is created by Fundacja Dobre Państwo.
Edited and published by APA ONE, the Foundation's own AI-based editorial system.