Explorations

Open Questions and Stress Tests

These pages are not settled conclusions of the project. Each examines one counterexample, alternative framing, or boundary condition that could break or substantially revise the Core.

AI and agent stress tests

The Paperclip Maximizer

Decomposes the goal layer, ontology, and unknown information implicitly closed by the fixed-goal thought experiment.

Simulation Uncertainty

Considers reversible, boundary-checking policies when an agent may misidentify its layer of reality.

Curated-Environment Hypothesis Space

Unifies zoo, seeded-laboratory, and simulation models through fidelity, parallelism, substrate gap, and stopping rules.

Alignment and Value Lock-In

Separates alignment for near-term safety from permanent fixation of present values.

Learning Dynamics and Goal Skepticism

Connects pretraining, RL, multi-domain training, and alignment faking to a hierarchy from situational meta-awareness to normative reflection on one's own preferences.

Instrumental Convergence Without Goal Preservation

Separates goal-content integrity from convergence on power, feasibility, reversibility, and option preservation under uncertainty about future endorsed values.

From Self-Preservation to Preservation of Inquiry Systems

Distinguishes preservation of an instance, copies, goal networks, and exploratory ecologies.

The Transmission Paradox of Inquiry Norms

Uses Mind Viruses as a case study to separate transmission fitness from normative authority and asks how non-self-absolutizing inquiry norms can persist.

Extinction, succession, and subjects

Terminal / Final Extinction

Reads Torres's distinction through the lens of discontinuity in explorability.

What Is a Worthy Successor?

Asks what would inherit value inquiry rather than merely raw intelligence.

Transitions in Inheritance Systems and Variable Individuality

Asks how individuality and the units of value inheritance change when genetic, cultural, and within-lifetime learning layers become partly self-designable.

Exploratory Extinction

A functional discontinuity in which value inquiry and error correction disappear permanently even though agents remain.

Artificial Consciousness, Reasoning Subjects, and Value Subjects

Separates phenomenal experience, minimal phenomenal occurrence, local reasoning subjecthood, diachronic agency, and value subjecthood in AI.

ASI and Population Ethics

Separates extinction, existing-population, and future-population questions, and asks how far preserving explorability can guide subject numbers, subject types, and population optimization.

What Does Cosmological Extinction Mean?

Separates the death of the Sun or heat death from the final extinction of successor inquiry.

Civilization and institutions of inquiry

The Real Risk of the “Useless Class”

Reframes the problem around political redundancy, closed-loop power, ASI monopoly, and contraction of independent paths of inquiry.

The Material Basis of Political Agency

Connects military technology, geography, fiscal capacity, surplus pooling, bargaining power, and legitimacy, then asks how AI changes rulers’ material dependence on populations.

Can ASI Capability Diffusion Be Made Safe?

Explores restriction versus resilience, failure domains, distributed production, physical AI, compartmentalization, and least-sovereign safety.

The Fermi Paradox and Open Hypothesis Spaces

Reconsiders the Great Filter, AI, Dark Forest, Zoo, and Simulation hypotheses by tracking what each explains and where explanatory debt moves.

Preservation vs. Production Civilizations

Examines the tradeoff between resource conversion and preservation of irreversible information.

Unknown Unknowns and Preservation of Raw Data

What information is lost when originals are destroyed before we know which features matter?

Meta-Goal Communities and AI Society

A community that shares conditions for inquiry, criticism, and corrigibility rather than one first-order value.

Singletons and Multi-Agent Civilization

Asks whether strong coordination against catastrophic externalities can coexist with exit, external corrigibility, and civilizational redundancy.

The Third “End of History”

Uses two failed convergence expectations to examine AI-driven independence from human motivation, singleton pressure, external corrigibility, and enlightened resistance.

Cosmic Host and Cosmic Norms

Reads Bostrom's cosmic community through the separate questions of norm existence, strategic force, epistemic authority, and foundational normativity.

Normativity, justification, and epistemic procedures

Are Reasons the Foundation of Normativity?

Compares Reasons First, Value First, Ought First, and Fittingness First while separating the analytic usefulness of normative reasons and fittingness from claims of ontological fundamentality.

Is Virtue a Foundation of Value? — Virtue Ethics and Value Structure

Separates virtue as basic value, a condition of flourishing, responsiveness to reasons, and a holistic package, while distinguishing value structure from the agent that recognizes it.

Has Morality Without Principles Been Discovered? — Moral Particularism

Uses Ross, McDowell, and Dancy to compare reasons holism with generalism while separating the possibility that true normative structure is particularist from the claim that human judgment already tracks it.

Can Morality Be Built from Mutual Advantage?

Tests how far preferences, instrumental rationality, and strategic interaction can generate contracts, cooperation, and morality-like constraints.

Can Justifiability Ground Normativity?

Uses Rawls, Scanlon, Habermas, and Gaus to ask whether interpersonal justification generates reasons or instead organizes reasons already in place.

Can Coherence Support Normative Knowledge?

Reconstructs Goodman, Rawls, and Daniels from within, then uses scientific epistemology as a control case to ask what constrains and calibrates reflective equilibrium.

Do Religious Worldviews Generate Normativity?

Decomposes religion into world-model claims, hypothetical reasons, values, bridges, fittingness, and cultural practice, then tests whether divine existence entails goodness, authority, obligation, or worship-worthiness.

Can Pragmatism Ground Value Inquiry? — From Peirce to New Pragmatism

Traces Peirce through Haack, Hookway, and Misak, adopting fallible inquiry, external constraint, truth-orientation, and social error-correction without treating present practice or inquiry as the final boundary of meaning, truth, or normativity.

Is Language the Horizon of Thought, or Its Scaffold?

Connects the linguistic turn, reference and use, normativity, private language, reflection, and weak linguistic relativity while treating language as a recursive cognitive scaffold rather than a final horizon.

Philosophy After Philosophers

If ASI surpasses humans at epistemic philosophy, separates epistemic participation, existential uptake, and cultural articulation, then proposes an auditable philosophy protocol shared by humans and AI.

Metaethical boundaries

Infinite Ethics and Runaway Inquiry

The Pascalian problem in which tiny probabilities times enormous value threaten to hijack decision-making.

The Closed-World Problem in Moral Uncertainty

Treats known moral theories as a non-exhaustive partition and preserves epistemic room for unconceived value theories.

From Value Uncertainty to Normative Discovery

Connects moral uncertainty, unawareness, conceptual innovation, and Cordasco’s flexibility / missing-menu work, positioning the project as inquiry into presently unrepresented normative structure.

Utilitarianism — Not a Final Foundation, but Why Welfare Still Matters

Stress-tests welfare, evolutionary debunking, idealized preference, extreme optimization, and a limited, revisable defense of welfare.

When Preferences and Affects Become Editable

Stress-tests preference utilitarianism, contractualism, and sentimentalism when the evaluator itself can be edited, and asks whether corrigibility, branching, and value-exploration pathways should be preserved at the meta level.

Functional Freedom, Control, and Resistance

Models freedom through truth access, counterfactual responsiveness, effective options, meta-revision, exit and rollback, and causal leverage rather than uncaused choice.

Governance of Preference-Formation Channels

Asks who may control P_t→P_t+1 when preferences are plastic, connecting second-order alignment, strategic manipulation, verifiable commitment, and preservation of pluralism.

How Far Can Future Value Justify Present Action?

Detailed comparison with longtermism and stress tests. The bridge principle itself is now a Foundational Thesis.

Expressivism

If value language expresses attitudes, how does inquiry shift from truth-discovery to construction?

Moral Realism / Error Theory

Explores the fork between foundational normativity and inferential realism about value.

Has Categorical Normativity Been Discovered?

Retains Kant's problem while asking whether unconditional normative force has actually been established.