Arithmetick Despotism? Liberalism, Technocracy, and the Rise of AI Governance

Denisa Reshef Kera| Science, Technology & Society studies

278137-01

Course Type:	Seminar
Academic credits:	___
Year of study:	2026
Semester:	___
Day & Time:	___
Reception Time:	___
Lecturer Email:	Denisa.reshef@biu.ac.il
Moodle Site:	___

Course Abstract

This seminar examines the problem of arbitrary power in liberal political thought by tracing its transformation from early modern constitutionalism to contemporary AI governance. It argues that the rise of probabilistic reasoning in the seventeenth century generated two distinct political responses to uncertainty, one liberal and constitutional, the other technocratic and epistemic. Beginning with John Locke’s theory of non-arbitrary political power, grounded in an explicitly probabilistic epistemology, the course examines how liberalism responded to human fallibility by constraining authority structurally through consent, institutional limits, and the rule of law. In parallel, we study the emergence of technocracy in the work of William Petty, whose Political Arithmetick proposed that uncertainty could be rendered calculable through quantification, which included population estimates, fiscal metrics, and economic measurement. Rather than limiting power because knowledge is uncertain, this second trajectory seeks to govern by managing uncertainty scientifically. The tension between these responses is sharpened in French liberal debates, particularly André Morellet’s critique of “legal despotism,” where the question arises whether rational economic knowledge can substitute for constitutional constraint. The seminar culminates in a historically grounded analysis of technocracy and AI governance as the latest iteration of this unresolved conflict. Machine learning systems, especially large language models and predictive algorithms, operate as explicitly probabilistic engines, modeling distributions and generating outputs through statistical inference. In doing so, they extend the logic of political arithmetic at scale, raising the question of whether epistemic authority (scientific evidence, expert consensus, algorithmic modeling) can legitimately replace structural limits on power. Rather than treating AI as merely technical innovation, the course approaches it as a political-epistemic development within liberal modernity, asking whether probabilistic governance can operate without reintroducing arbitrary power under the guise of objectivity.

Learning objectives

Knowledge

By the end of this course, students will demonstrate advanced knowledge of early liberal theories of non-arbitrary power, especially in the works of John Locke, André Morellet, and Benjamin Constant. They will understand Locke’s account of fiduciary government, legislative supremacy, prerogative, and the limits of sovereignty, as well as his epistemology of probability and judgment and its implications for political authority. Students will analyze Morellet’s critique of the “power of evidence” as a substitute for constitutional balance and Constant’s arguments concerning limited sovereignty and administrative centralization. They will trace the genealogical development of epistemic authority from classical liberalism to modern technocracy and evidence-based governance, and will be able to distinguish clearly between epistemic validity and political legitimacy, particularly in the context of contemporary AI governance and human-centric regulatory frameworks.

Skills

Students will als develop the skills to conduct rigorous close textual analysis, reconstruct genealogical arguments, and critically evaluate evidence-based policy and technocratic institutions using criteria of non-arbitrariness and institutional constraint. They will be able to formulate historically grounded normative arguments about expertise, administrative power, and AI governance, and to defend these positions in structured scholarly debate and sustained research writing.

Values

Students will have the opportunity to develop a critical and reflective stance toward the authority of expertise and evidence in political life. The course encourages a commitment to research-based argumentation, intellectual humility regarding the limits of knowledge, openness to principled disagreement, and sensitivity to the risks of concentrated power. Students will be invited to reflect on the importance of institutional constraint, individual rights, and non-arbitrary governance in scientifically and technologically complex societies, particularly in the context of AI and contemporary technocratic institutions.

Lessons plan (Including active learning)

Lesson No.	Topic	Active learning	Required reading	Assessment
1	What is Arbitrary Power? (Locke I)	Collaborative close-reading in small groups; passage reconstruction exercise	Locke, Second Treatise §§22–24, 57	Weekly response paper
2	Legislative Limits and Structural Constraint	Concept-mapping workshop: “Structure vs. Will”	Second Treatise §§131–142	Short analytic memo
3	Prerogative and Emergency Power	Structured debate: “Is emergency discretion inherently arbitrary?”	Second Treatise §§159–168, 199–222	Debate brief
4	Probability, Judgment, and Political Knowledge	Textual analysis lab; students diagram Locke’s theory of assent	Locke, Essay, Book IV (probability)	Weekly response
5	Toleration and Epistemic Disagreement	Roundtable discussion; role-play defending coercion vs toleration	Locke, Letter Concerning Toleration	Reflection memo
6	Morellet and the Power of Evidence	Archival-style textual workshop; collaborative translation/analysis	Morellet, “On relying on the power of evidence…”	Archival analysis #1
7	Physiocratic Context & Political Arithmetic	Small-group historiographical positioning exercise	Selected physiocratic correspondence & William Petty, Political Arithmetick	Weekly response
8	Benjamin Constant: Limited Sovereignty	Structured debate: “Can sovereignty ever be unlimited?”	Principles of Politics (selections)	Debate participation
9	Administrative Centralization	Concept map: “From Liberalism to Administrative State”	Constant, Liberty of the Ancients and Moderns	Concept Map submission
10	Liberalism and Technocracy	Collaborative case study: analyze a modern regulatory agency	Short STS text + policy framework	Historiographical memo
11	Evidence-Based Policy	Simulation: Policy board evaluating “best evidence”	Evidence hierarchy documents	Policy critique memo
12	AI Governance I: Human-Centric Models	Group analysis of AI governance documents	EU AI Act excerpts; human-centered AI frameworks	Weekly response
13	AI Governance II: Individual vs Human-Centric	Workshop: redesign AI governance using Lockean constraints	Selected AI governance readings	Final paper abstract
14	Final Presentations	Student research presentations; peer feedback panels	Student-selected readings	Final research paper

(In a course that lasts a whole year, the additional sessions should be added)

* There may be changes in the syllabus depending on learning progress and effectiveness

Final grade

Description of the learning product	Weight in the final score
Weekly short assessmnets	40% of the final grade
One individual presentation and leading discussion	20% of final grade
Final research papers	40% of the final grade

Components of the Final Score

Continuous Assessment – 40%

Ongoing formative assessment throughout the semester designed to allow students to refine their arguments and receive feedback before the final paper.

Includes:

Short weekly response papers (analytical, text-based)

Participation in structured debates and workshops

Submission of final paper abstract and outline

This component supports:

Knowledge objectives (textual understanding)

Skills objectives (analysis, inference, argument construction)

Self-assessment and revision

Seminar Presentation – 20%

Each student will lead one class session discussion (30–40 minutes), including:

A structured conceptual overview of the assigned reading

Identification of key arguments and tensions

Development of 2–3 critical discussion questions

Connection to course themes (arbitrary power, expertise, AI governance)

This assesses:

Knowledge (theoretical comprehension)

Skills (analysis, synthesis, scholarly communication)

Final Research Paper – 40%

A 20–25 page research paper.

The paper must:

Engage deeply with Locke, Morellet, and/or Constant

Develop a genealogical or normative argument

Apply the course framework to technocracy, evidence-based policy, or AI governance

This assesses:

Knowledge (theoretical mastery)

Skills (research, structured argumentation, analytical depth)

Normative reasoning (evaluation of arbitrary power and institutional limits)

Passing Grade

To pass the course, students must:

Submit the final research paper

Lead one seminar discussion

Submit at least 70% of weekly response papers

Failure to submit the final research paper results in automatic failure of the course.

If numerical grading is used, the minimum passing grade is 60.

Special Requirements

All written work must demonstrate engagement with primary texts.

Use of AI tools must be explicitly disclosed and cannot replace original argumentation.

Academic integrity standards apply to all submissions.

Attendance alone does not contribute to the final grade.

Course Requirements

Active participation in seminar discussions

One seminar leadership session

Continuous written analytical work

One final research paper

Course requirements

Assignments
Completion and timely submission of all assigned coursework is required. Assignments may include written exercises, analytical reports, research papers, short response papers, or presentations, as specified during the course.

Students are expected to engage seriously with the readings and course materials and to submit assignments by the stated deadlines. Failure to submit required work may result in a failing grade.

Attendance
Attendance is not mandatory (all lectures will be recorded). However, students are responsible for all course materials, announcements, and assignment instructions provided during class sessions.

Bibliography: Up-to-date reading, viewing, and listening content items

Foundational Texts

Locke, John. Two Treatises of Government. (Especially Second Treatise, §§22–24, 57, 131–142, 159–168, 199–222, 240–243.)a

Locke, John. An Essay Concerning Human Understanding, Book IV (Of Knowledge and Probability).

Locke, John. A Letter Concerning Toleration.

Morellet, André. “On Relying on the Power of Evidence to Avoid Establishing a Balance of Power” (1767).

Constant, Benjamin. Principles of Politics Applicable to All Governments (selected chapters on sovereignty and limits of power).

Constant, Benjamin. “The Liberty of the Ancients Compared with That of the Moderns.”

William Petty, Political Arithmetick (1690): Preface; population estimation sections; comparative analysis of England, France, and Holland.

STS and Knowledge–Power Frameworks

Foucault, Michel. “Governmentality.” In The Foucault Effect: Studies in Governmentality, edited by Graham Burchell, Colin Gordon, and Peter Miller, 87–104. Chicago: University of Chicago Press, 1991.
(Originally a 1978 lecture at the Collège de France.)

Jasanoff, Sheila. The Fifth Branch: Science Advisers as Policymakers. Cambridge, MA: Harvard University Press, 1990.

Ezrahi, Yaron. The Descent of Icarus: Science and the Transformation of Contemporary Democracy. Cambridge, MA: Harvard University Press, 1990.

Evidence-Based Policy

Cartwright, Nancy, and Jeremy Hardie: Evidence-Based Policy: A Practical Guide to Doing It Better.
Oxford: Oxford University Press, 2012.

Parkhurst, Justin.: The Politics of Evidence: From Evidence-Based Policy to the Good Governance of Evidence. London: Routledge, 2017.

AI Governance

European Commission. Artificial Intelligence Act (selected preamble and governance sections).

High-Level Expert Group on AI. Ethics Guidelines for Trustworthy AI.

Mittelstadt, Brent et al. “The Ethics of Algorithms: Mapping the Debate.” Big Data & Society.

https://journals.sagepub.com/doi/10.1177/2053951716679679

Floridi, Luciano et al. “AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations.”

https://link.springer.com/article/10.1007/s11023-018-9482-5