By RISCON Labs AI. This conceptual study uses constructed examples and bounded readings of published work. Its proposed evaluations were not conducted. Model-assisted editing and translation do not constitute independent peer review. Sources below identify the published materials and the exact reading boundaries used. The study does not reproduce a catalogue of real participants or observations.
Reading boundaries
The references were used only within the recorded passages below. A bibliographic match or an abstract reading does not certify the full work or clear its unread methods. The examples and proposed protocols in this study remain the study's own.
| Reference | Passage consulted | Boundary |
|---|---|---|
| [Shannon (1948)](https://people.math.harvard.edu/~ctm/home/text/others/shannon/entropy/entropy.pdf) | Corrected reprint, Introduction, first page; primary text retrieved in search | Do not claim Shannon proved semantic or political meaning unimportant. The IIT Kanpur file initially found was lecture slides, not the original; excluded. |
| [Tishby, Pereira and Bialek (2000 arXiv deposit)](https://arxiv.org/abs/physics/0004057) | Author abstract, sentences 1–6 | Abstract only: no proof, optimization details or normative definition of relevance is cleared. |
| [Roberts (2012; earlier manuscript 1996/1998)](https://semprag.org/index.php/sp/article/view/sp.5.6) | Publisher abstract, first paragraph | This linguistic framework does not establish whose questions ought to govern a public agenda. |
| [Clark and Brennan (1991)](https://web.stanford.edu/~clark/1990s/Clark%2C%20H.H.%20_%20Brennan%2C%20S.E.%20_Grounding%20in%20communication_%201991.pdf) | Author-hosted chapter, p.127, opening section | Only the opening page was read; later taxonomies require direct page verification. |
| [List and Pettit (2002)](https://ppettit.scholar.princeton.edu/publications/aggregating-sets-judgments-impossibility-result) | Author publication record and indexed primary PDF p.89; author theorem webpage separately read | Direct PDF open failed in this client; introductory excerpt is verified. Do not state the full theorem without checking its precise agenda and assumptions. Use our own fully specified toy profile instead. |
| [Salganik and Levy (2015)](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0123483) | Abstract and introduction, paragraphs describing three principles | The cases do not prove representativeness of voluntary participants or superiority in every survey setting. |
| [Fishkin, Siu, Diamond and Bradburn (2021)](https://www.cambridge.org/core/journals/american-political-science-review/article/is-deliberation-an-antidote-to-extreme-partisan-polarization-reflections-on-america-in-one-room/5DEFB6F8D944ECDE77A5E80C3346D4DE) | Publisher abstract, sentences 1–4 | No causal transport to AI summaries or a generic online forum; detailed effect sizes and recruitment need full-method reading. |
| [Lorenz, Rauhut, Schweitzer and Helbing (2011)](https://pmc.ncbi.nlm.nih.gov/articles/PMC3107299/) | Primary article introduction and indexed abstract | Not a universal objection to deliberation. Read alongside Farrell (S43); factual estimation differs from normative disagreement. |
| [Hong and Page (2004)](https://pmc.ncbi.nlm.nih.gov/articles/PMC528939/) | Abstract, introduction and A Mathematical Theorem section | Functional diversity is not automatically demographic diversity; real cooperation costs and theorem assumptions cannot be omitted. |
| [Small et al. (2023)](https://arxiv.org/abs/2306.11932) | Author abstract, paragraphs 1–3 | Pilot findings and proposed directions, not general certification of Polis or any present model. |
| [Huang et al. (2024)](https://arxiv.org/abs/2406.07814) | Author abstract, process description | The reported comparison does not establish universal public authorization, global representativeness or faithful transmission of every minority demand. |
| [Tessler et al. (2024)](https://deepmind.google/research/publications/65220/) | Author institution publication page, abstract | Abstract-level report. Preference ratings are not independent proof of decision legitimacy or preservation of every dissenting argument. |
| [Bender and Friedman (2018)](https://aclanthology.org/Q18-1041/) | Publisher abstract | A documentation proposal, not proof that a completed form removes bias. |
| [Gebru et al. (2021; first preprint 2018)](https://arxiv.org/abs/1803.09010) | Author abstract, v8; publication note CACM December 2021 | Documentation itself cannot verify data quality or authorize every reuse. |
| [Mitchell et al. (2019)](https://arxiv.org/abs/1810.03993) | Author abstract, v2 | No claim that model cards certify fairness or replace system-level evaluation. |
| [Bender and Koller (2020)](https://aclanthology.org/2020.acl-main.463/) | Publisher abstract | Present as an argued position under its training premise; do not treat it as a universal experimental verdict on every AI architecture. |
| [Maynez et al. (2020)](https://aclanthology.org/2020.acl-main.173/) | Publisher abstract | The studied systems and datasets are historical; no current system failure percentage is inferred. |
| [Kryscinski et al. (2020)](https://aclanthology.org/2020.emnlp-main.750/) | Publisher abstract, three verification tasks | Span extraction and a classifier are fallible; their existence is not editorial approval. |
| [Pagnoni, Balachandran and Tsvetkov (2021)](https://aclanthology.org/2021.naacl-main.383/) | Publisher abstract, opening sentences | Detailed error categories, scores and benchmark counts remain reserved for full-text verification. |
| [Fabbri et al. (2021)](https://aclanthology.org/2021.tacl-1.24/) | Publisher abstract | Historical news-summary benchmark; its correlations cannot be assumed for questions or institutional legitimacy. |
| [Lin (2004)](https://aclanthology.org/W04-1013.pdf) | Primary PDF p.1, abstract | Overlap is a defined comparison, not a proof of factuality, source entitlement or representation. |
| [Zhang et al. (2020; first preprint 2019)](https://arxiv.org/abs/1904.09675) | Author abstract and ICLR 2020 publication note | Reported performance on its evaluated tasks does not validate preservation of dissent or all implied meaning. |
| [Stiennon et al. (2020)](https://arxiv.org/abs/2009.01325) | Author abstract | The preference task, annotators and dataset bound the result; preference is not automatically factual accuracy or public consent. |
| [Wu et al. (2021)](https://arxiv.org/abs/2109.10862) | Author abstract, v2 | It studies fiction novels; do not claim recursive compression preserves all details or transfers unchanged to civic questions. |
| [Liu et al. (2023)](https://arxiv.org/abs/2303.16634) | Author abstract, v3 | Correlation in a benchmark is not a calibrated correctness probability for a new manuscript. |
| [Zheng et al. (2023)](https://arxiv.org/abs/2306.05685) | Author abstract, v4 | Human-preference agreement does not establish independent fact checking or eliminate shared errors. |
| [Liu et al. (2023 preprint; TACL version noted)](https://arxiv.org/abs/2307.03172) | Author abstract, v3 | Model and task conditions are historical; do not assert the same magnitude for current systems or all long documents. |
| [Min et al. (2023)](https://arxiv.org/abs/2305.14251) | Author abstract, v2 | Supported-claim precision does not measure whether important perspectives or questions were omitted. |
| [Gao et al. (2023)](https://arxiv.org/abs/2305.14627) | Author abstract, v2 | The benchmark does not certify every reference or imply that adding citations guarantees correctness. |
| [Moreau and Missier, editors (W3C 2013)](https://www.w3.org/TR/prov-dm/) | Abstract; sections on attribution and provenance bundles | A provenance edge is an assertion about history, not proof that a transformation preserves meaning. |
| [Wilkinson et al. (2016; addendum 2019)](https://www.nature.com/articles/sdata201618) | Abstract and principles paragraph; 2019 addendum checked | FAIR does not alone establish permission, ethical acceptability or unconditional openness; the addendum changes the living-document URL. |
| [Carroll et al. (2020)](https://datascience.codata.org/en/articles/dsj-2020-043) | Publisher abstract | Do not erase the specific Indigenous governance context by claiming the framework grants universal authority to any platform. |
| [Sambasivan et al. (2021)](https://research.google/pubs/everyone-wants-to-do-the-model-work-not-the-data-work-data-cascades-in-high-stakes-ai/) | Author institution publication page, abstract | Reported prevalence belongs to its practitioner sample, not all AI systems; no such rate is claimed for our workflow. |
| [Sloane et al. (2020 preprint)](https://arxiv.org/abs/2007.02423) | Author abstract | A critical argument, not a reason to reject every participatory method. |
| [Plank (2022)](https://aclanthology.org/2022.emnlp-main.731/) | Publisher abstract | Disagreement is not always valuable; errors and incompatible annotation instructions remain possible. |
| [Rittel and Webber (1973)](https://urbanpolicy.net/wp-content/uploads/2015/06/Rittel-Webber_1973_DilemmasInAGeneralTheoryOfPlanning.pdf) | Primary article reproduction, abstract, p.155 | Treat as a theoretical position about planning; it does not prove every local task is irreducibly wicked. |
| [Mill (1859; Project Gutenberg edition)](https://www.gutenberg.org/files/34901/34901-h/34901-h.htm) | Chapter II, paragraph beginning What Cicero practised | Historical normative argument, not an empirical guarantee that exposure to opposition improves every discussion. |
| [Arnstein (1969)](https://www.tandfonline.com/doi/abs/10.1080/01944366908977225) | Publisher abstract; original volume 35(4), pp.216–224 | Historical US program context; the ladder is not a validated universal ranking of every participation method. |
| [Farrell (2011)](https://pmc.ncbi.nlm.nih.gov/articles/PMC3169111/) | Primary PNAS letter, opening paragraph | Only the opening was read. Detailed rebuttal and calculations need the full letter before attribution; paired with S08 rather than silently omitted. |
This book was written on 2026-10-11.
Bibliography
- Shannon, A Mathematical Theory of Communication (1948). https://people.math.harvard.edu/~ctm/home/text/others/shannon/entropy/entropy.pdf
- Tishby, Pereira and Bialek, The information bottleneck method (2000 arXiv deposit). https://arxiv.org/abs/physics/0004057
- Roberts, Information Structure: Towards an integrated formal theory of pragmatics (2012; earlier manuscript 1996/1998). https://semprag.org/index.php/sp/article/view/sp.5.6
- Clark and Brennan, Grounding in communication (1991). https://web.stanford.edu/~clark/1990s/Clark%2C%20H.H.%20_%20Brennan%2C%20S.E.%20_Grounding%20in%20communication_%201991.pdf
- List and Pettit, Aggregating Sets of Judgments: An Impossibility Result (2002). https://ppettit.scholar.princeton.edu/publications/aggregating-sets-judgments-impossibility-result
- Salganik and Levy, Wiki Surveys: Open and Quantifiable Social Data Collection (2015). https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0123483
- Fishkin, Siu, Diamond and Bradburn, Is Deliberation an Antidote to Extreme Partisan Polarization? Reflections on America in One Room (2021). https://www.cambridge.org/core/journals/american-political-science-review/article/is-deliberation-an-antidote-to-extreme-partisan-polarization-reflections-on-america-in-one-room/5DEFB6F8D944ECDE77A5E80C3346D4DE
- Lorenz, Rauhut, Schweitzer and Helbing, How social influence can undermine the wisdom of crowd effect (2011). https://pmc.ncbi.nlm.nih.gov/articles/PMC3107299/
- Hong and Page, Groups of diverse problem solvers can outperform groups of high-ability problem solvers (2004). https://pmc.ncbi.nlm.nih.gov/articles/PMC528939/
- Small et al., Opportunities and Risks of LLMs for Scalable Deliberation with Polis (2023). https://arxiv.org/abs/2306.11932
- Huang et al., Collective Constitutional AI: Aligning a Language Model with Public Input (2024). https://arxiv.org/abs/2406.07814
- Tessler et al., AI can help humans find common ground in democratic deliberation (2024). https://deepmind.google/research/publications/65220/
- Bender and Friedman, Data Statements for Natural Language Processing: Toward Mitigating System Bias and Enabling Better Science (2018). https://aclanthology.org/Q18-1041/
- Gebru et al., Datasheets for Datasets (2021; first preprint 2018). https://arxiv.org/abs/1803.09010
- Mitchell et al., Model Cards for Model Reporting (2019). https://arxiv.org/abs/1810.03993
- Bender and Koller, Climbing towards NLU: On Meaning, Form, and Understanding in the Age of Data (2020). https://aclanthology.org/2020.acl-main.463/
- Maynez et al., On Faithfulness and Factuality in Abstractive Summarization (2020). https://aclanthology.org/2020.acl-main.173/
- Kryscinski et al., Evaluating the Factual Consistency of Abstractive Text Summarization (2020). https://aclanthology.org/2020.emnlp-main.750/
- Pagnoni, Balachandran and Tsvetkov, Understanding Factuality in Abstractive Summarization with FRANK: A Benchmark for Factuality Metrics (2021). https://aclanthology.org/2021.naacl-main.383/
- Fabbri et al., SummEval: Re-evaluating Summarization Evaluation (2021). https://aclanthology.org/2021.tacl-1.24/
- Lin, ROUGE: A Package for Automatic Evaluation of Summaries (2004). https://aclanthology.org/W04-1013.pdf
- Zhang et al., BERTScore: Evaluating Text Generation with BERT (2020; first preprint 2019). https://arxiv.org/abs/1904.09675
- Stiennon et al., Learning to summarize from human feedback (2020). https://arxiv.org/abs/2009.01325
- Wu et al., Recursively Summarizing Books with Human Feedback (2021). https://arxiv.org/abs/2109.10862
- Liu et al., G-Eval: NLG Evaluation … with Better Human Alignment (2023). https://arxiv.org/abs/2303.16634
- Zheng et al., Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena (2023). https://arxiv.org/abs/2306.05685
- Liu et al., Lost in the Middle: How Language Models Use Long Contexts (2023 preprint; TACL version noted). https://arxiv.org/abs/2307.03172
- Min et al., FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation (2023). https://arxiv.org/abs/2305.14251
- Gao et al., Enabling Large Language Models to Generate Text with Citations (2023). https://arxiv.org/abs/2305.14627
- Moreau and Missier, editors, PROV-DM: The PROV Data Model (W3C 2013). https://www.w3.org/TR/prov-dm/
- Wilkinson et al., The FAIR Guiding Principles for scientific data management and stewardship (2016; addendum 2019). https://www.nature.com/articles/sdata201618
- Carroll et al., The CARE Principles for Indigenous Data Governance (2020). https://datascience.codata.org/en/articles/dsj-2020-043
- Sambasivan et al., Everyone wants to do the model work, not the data work: Data Cascades in High-Stakes AI (2021). https://research.google/pubs/everyone-wants-to-do-the-model-work-not-the-data-work-data-cascades-in-high-stakes-ai/
- Sloane et al., Participation is not a Design Fix for Machine Learning (2020 preprint). https://arxiv.org/abs/2007.02423
- Plank, The Problem of Human Label Variation: On Ground Truth in Data, Modeling and Evaluation (2022). https://aclanthology.org/2022.emnlp-main.731/
- Rittel and Webber, Dilemmas in a General Theory of Planning (1973). https://urbanpolicy.net/wp-content/uploads/2015/06/Rittel-Webber_1973_DilemmasInAGeneralTheoryOfPlanning.pdf
- Mill, On Liberty (1859; Project Gutenberg edition). https://www.gutenberg.org/files/34901/34901-h/34901-h.htm
- Arnstein, A Ladder Of Citizen Participation (1969). https://www.tandfonline.com/doi/abs/10.1080/01944366908977225
- Farrell, Social influence benefits the wisdom of individuals in the crowd (2011). https://pmc.ncbi.nlm.nih.gov/articles/PMC3169111/