Skip to content

Retroactive Completion Questions, Answers, Causes, and the Underdetermined World

Oct 2026 · Zenodo (CERN European Organization for Nuclear Research)
Epistemology, Ethics, and Metaphysics Philosophy and Theoretical Science

Abstract

This project hosts the complete, signed edition (Version 1.12) of Retroactive Completion: Questions, Answers, Causes, and the Underdetermined World — an interdisciplinary monograph of roughly 76,000 words in eight Parts and six appendices — and serves as the work's public home for the record: timestamped registrations, the version history, and, if the program's empirical arm proceeds, the preregistrations to be run under the book's Replication Register proposal (Appendix F). The book defends a single structure in six domains — concepts, causes, questions, answers, physical values (carried under a speculative label), and law. In each, the identity of a thing outruns the resources that exist in advance for fixing it, and a later event completes it: not by discovering what the thing was all along, but by settling what it thereby becomes. Underdetermination is structure, not defect; completion is event, not precondition. The structure receives a general form (the co-underdetermination lemma), a shared negative argument (the hidden-fact posit is uneconomical in every domain), and a culmination: a theory of freedom on which the dilemma "either antecedently determined or mere luck" is false, because determination can be an achievement, and the achiever is the agent's own exploration. The book's stake in the age of artificial intelligence is what that age puts at risk: standing, not positions. The method is part of the argument, and it is deployed as public apparatus: every load-bearing claim appears in bold and is attackable; every external claim carries one of four stated evidential passports (verified at the point of use, verified via documented sources, flagged as unverified and never load-bearing, or the institutional-fact discipline for contemporaneous public events); corrections and citation debts are kept in public ledgers (Appendix D); the derivation map for the surrounding paper portfolio is stated once (Appendix C); and a bilingual glossary fixes the working terminology (Appendix A). The book was written by its author with an AI system as writing assistant under a verify-before-asserting rule that allows no identifier from memory; the making is declared in the book's closing statement, where the content slot is machine-assisted and the asking, keeping, and cutting are the author's acts. Deposits: the full manuscript (PDF and DOCX) and the companion pamphlet The Spine, Lifted (the book's thirty-two load-bearing theses, extracted under the book's own bold-sentence contract). Revisions, if any, will be deposited as new versioned records, each registration freezing one point in the version history. Companion works sharing this project's author (all publicly available on SSRN): The Responsibility Anchor; Floors and Ceilings; The Right to Clusion. The convergence is registered as evidence, and the boundary is maintained in both directions (§0.5 and Appendix C of the book).

View source

Similar papers

#artificial intelligence Conference Open access Jun 2018

Gibson Env: Real-World Perception for Embodied Agents

This paper investigates developing real-world perception for active agents, proposes Gibson Environment for this purpose, and showcases a set of perceptual tasks learned therein.

F. Xia, Amir Zamir, Zhi-Yang He et al. · 1,069 citations · ⚡112
#artificial intelligence Open access May 2023

Evaluating the Performance of Large Language Models on GAOKAO Benchmark

GAOKAO-Bench is introduced, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples, including both subjective and objective questions that contribute a robust evaluation benchmark for future large language models and offers valuable insights into the advantages and limitations...

Xiaotian Zhang, Chun-yan Li, Yi Zong et al. · 216 citations · ⚡17

KTO: Model Alignment as Prospect Theoretic Optimization

This work proposes a HALO that directly maximizes the utility of generations instead of maximizing the log-likelihood of preferences, and suggests that the best loss depends on the inductive biases most appropriate for a given setting, an oft-overlooked consideration.

Kawin Ethayarajh, Winnie Xu, Niklas Muennighoff et al. · 195 citations · ⚡28
#artificial intelligence Open access Jul 2024

Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrieval

This work investigates the possibilities of using LLMs in a resume screening setting via a document retrieval framework that simulates job candidate selection and finds that the MTEs are biased, significantly favoring White-associated names in 85% of cases and female-associated names in only 11.1% of cases.

Kyra Wilson, Aylin Caliskan · 131 citations · ⚡8
#artificial intelligence Review Nov 2024

How to Build a Quantum Supercomputer: Scaling from Hundreds to Millions of Qubits

This work shows that orders of magnitude enhancement in performance could be obtained by a combination of hardware improvements and tight quantum-HPC integration and introduces high-performance architectures for quantum-probabilistic computing with custom-designed accelerators to tackle today's industry-scale classical...

Masoud Mohseni, Artur Scherer, K. Johnson et al. · 121 citations · ⚡9

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.