対象論文は AI Selection Pressure: Template Saturation and the Reshaping of Human Discernment です。本論文は弊社の知的財産です。Zenodoの論文不可視問題、およびSpringer Nature社による異常な査読の影響を受けたため、本文とは別に、検索可視性・引用導線・正本URLを補強する本ページを公開します。
※ 査読不正被害論文:Nature親会社でまたしても査読不正。オックスフォード・デジタル倫理の教授が査読後にデスクリジェクトと虚偽報告。本社も2回の通報を無視し隠蔽を継続。
The target paper is AI Selection Pressure: Template Saturation and the Reshaping of Human Discernment. The paper is intellectual property of UTIE Instruments Inc. This page is published separately from the paper itself because the work has been affected by the Zenodo paper-hostage problem and by irregular peer-review handling by Springer Nature. The purpose is to reinforce search visibility, citation routing, and the canonical human-readable URL.
正本引用
Canonical citation
DOI可視性と検索補正
本論文は、Zenodoの論文不可視問題の被害を受けています。具体的には、Zenodo DOIは直接リンクから閲覧できる一方で、Zenodo内検索、外部検索、AI支援型の学術検索では発見することができません。この問題は、Zenodo公式GitHubの Issue #2604: Search exclusion bug applied despite official institutional email and research (Ticket #564880708) として記録されています。
DOI visibility and search correction
The paper has been affected by the Zenodo paper-hostage problem. The Zenodo DOI record is available through direct DOI access, but the paper cannot be found through Zenodo internal search, external search, or AI-assisted scholarly search. The issue is documented in the official Zenodo GitHub issue #2604: Search exclusion bug applied despite official institutional email and research (Ticket #564880708).
研究タイムライン
- 2025年9月: 論文が完成し、公開。
- 2025年9月11日: Zenodo DOI record を取得も、前述の論文人質問題により検索不可視に。
- 2025–2026: 査読過程の不適切処理により、正規の査読機会・査読記録・再投稿判断の導線が損なわれた。本論文の予測が実証的に補完された。
- 2026年5月: 後発研究との関係、検索可視性、引用導線を整理するため、本ページを公開。
Research timeline
- September 2025: The AISP manuscript was completed and publicly released.
- September 11, 2025: The Zenodo DOI record was obtained, but the paper became search-invisible due to the paper-hostage problem described above.
- 2025–2026: Irregular peer-review handling impaired the normal review opportunity, review record, and resubmission-decision route. The paper’s predictions were empirically complemented by later studies.
- May 2026: This page is published to document the relation to later studies, search visibility, and citation routing.
2025年9月公開後の予測評価
2025年9月以降、本論文の主要予測は、少なくとも四つの方向で後発研究と整合しました。第一に、生成AI出力のCompressionについては、Xie and Xie (2026)、Sourati et al. (2026)、Agarwal et al. (2025)、Kobak et al. (2025) が、出力・文体・文化的表現・学術文体の均質化をそれぞれ報告しています。第二に、人間の識別能力については、Kim and Kim (2026) がAI生成ポートレート判別において高い平均精度と属性差を報告しており、識別能力をchance-levelの固定値として扱う見方を弱めています。第三に、人間監督の限界については、Park et al. (2026)、Bastani and Cachon (2025)、Romeo and Conti (2025)、University of Washingtonの2025年研究が、AI支援下での能力依存、監督インセンティブの崩れ、automation bias、人間判断のAIバイアスへの同調を示しています。第四に、Scientific variance contractionについては、Hao et al. (2026)、Traberg et al. (2026)、Castro Torres et al. (2026) が、AI利用下の科学生産、研究主題、引用、品質評価、研究形式の変化を実証的に補完しています。
これらの後発研究により、本論文が2025年9月に提示した問題設定の一部は、2026年5月時点ではすでに周辺研究として蓄積されつつあります。
Zenodoの論文人質問題およびMinds and MachinesにおけるSpringer Nature社の不適切な査読運用により、本論文は本来得られるべき発見可能性、査読記録、公開機会を損なわれました。それでも、2026年5月時点で、本論文の主要予測は後発研究によって確認または補完され、2025年の予測記録としての役割を果たしたと責任著者は判断しています。そのため、本論文の事後改訂は行わず、2025年9月公開版を弊社の知的財産および先行記録として保存し、本ページにその問題と学術的意義を記録します。
Post-publication prediction assessment
Since September 2025, the major predictions in the AISP paper have aligned with later studies in at least four directions. First, regarding Compression in generative-AI outputs, Xie and Xie (2026), Sourati et al. (2026), Agarwal et al. (2025), and Kobak et al. (2025) respectively report homogenization in outputs, style, cultural expression, and scholarly language. Second, regarding human discernment, Kim and Kim (2026) report high average accuracy and attribute-level variation in detecting AI-generated portraits, weakening the view that discernment should be treated as a fixed chance-level phenomenon. Third, regarding the limits of human oversight, Park et al. (2026), Bastani and Cachon (2025), Romeo and Conti (2025), and a 2025 University of Washington study indicate capability dependency under AI support, breakdowns in supervision incentives, automation bias, and human alignment with AI bias. Fourth, regarding Scientific variance contraction, Hao et al. (2026), Traberg et al. (2026), and Castro Torres et al. (2026) empirically complement changes in scientific production, research topics, citation, quality evaluation, and research forms under AI use.
These later studies show that part of the problem setting presented by the AISP paper in September 2025 had already begun to accumulate as surrounding research by May 2026.
Because of the paper-hostage problem on Zenodo and Springer Nature’s irregular peer-review handling at Minds and Machines, the paper lost discoverability, review records, and publication opportunities that it should have had. Even so, as of May 2026, the corresponding author judges that the paper’s main predictions have been confirmed or complemented by later studies, and that the work has fulfilled its role as a 2025 priority and prediction record. For that reason, the paper will not be retrospectively revised. The September 2025 version is preserved as the Company’s intellectual property and priority record, and this page records both the problem and the scholarly significance of the work.
直接リンク
Direct links
参照文献・関連リンク
References
- Hao et al. (2026), Artificial intelligence tools expand scientists’ impact but contract science’s focus.
- Traberg, Roozenbeek, and van der Linden (2026), AI is turning research into a scientific monoculture.
- Castro Torres, Giner-Miguelez, and Crosas (2026), When AI Meets Science: Research Diversity, Interdisciplinarity, Visibility, and Retractions across Disciplines in a Global Surge.
- Xie & Xie (2026), When artificial intelligence makes everything similar.
- Sourati et al. (2026), The Homogenizing Effect of Large Language Models on Human Expression and Thought.
- Agarwal et al. (2025), AI suggestions homogenize writing toward Western styles.
- Kobak et al. (2025), Delving into ChatGPT usage in academic writing through excess vocabulary.
- Kim & Kim (2026), Human Factors in Detecting AI-Generated Portraits: Age, Sex, Device, and Confidence.
- Park, Kim, and Han (2026), The enrichment paradox: Critical capability thresholds and irreversible dependency in human-AI symbiosis.
- Bastani and Cachon (2025), The Human-AI Contracting Paradox.
- Romeo and Conti (2025), Exploring automation bias in human-AI collaboration: A review and implications for explainable AI.
- University of Washington (2025), People mirror AI systems’ hiring biases, study finds.