核心發現:原始論文因繪圖軟體臭蟲,導致最高破壞性分數的論文未被納入長條圖;更關鍵的是,早期 Web of Science 資料庫中許多論文參考文獻欄位空白,被錯誤歸類為「最大破壞性」[4][5]。
隨著資料庫品質提升,這類錯誤的「最大分數」越來越少,於是產生了破壞力持續下降的假象[5]。
What study by researchers from Vrije Universiteit Brussel and KU Leuven, published as a rebuttal in Nature in August 2025, found that a wideAI-generated editorial hero image for What study by researchers from Vrije Universiteit Brussel and KU Leuven, published as a rebuttal in Nature in August 2025, found that a wide.
AI 提示詞
Create a landscape editorial hero image for this Studio Global article: What study by researchers from Vrije Universiteit Brussel and KU Leuven, published as a rebuttal in Nature in August 2025, found that a wide. Article summary: The rebuttal study is Holst et al., "Dataset artefacts can partially drive the measured decline in disruption," published in Nature in August 2026 (DOI: 10.1038/s41586 026 10787 y) [4][5].. Topic tags: general web, workflow, code, regulation, benchmarks. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, click
openai.com
反駁研究為 Holst 等人,〈Dataset artefacts can partially drive the measured decline in disruption〉,2026年8月刊登於《Nature》(DOI: 10.1038/s41586-026-10787-y)。作者包括 Vincent Holst、Andres Algaba、Floriano Tori、Sylvia Wenmackers 以及 Vincent Ginis,分別來自布魯塞爾自由大學(VUB)與荷語魯汶大學(KU Leuven)。這篇論文直接挑戰2023年同樣刊登於《Nature》、由 Park、Leahey 與 Funk 所發表、宣稱科學與技術的「破壞性創新」(disruptiveness)數十年來持續走低的轟動研究。
核心發現:原始論文因繪圖軟體臭蟲,導致最高破壞性分數的論文未被納入長條圖;更關鍵的是,早期 Web of Science 資料庫中許多論文參考文獻欄位空白,被錯誤歸類為「最大破壞性」[4][5]。
隨著資料庫品質提升,這類錯誤的「最大分數」越來越少,於是產生了破壞力持續下降的假象[5]。
What study by researchers from Vrije Universiteit Brussel and KU Leuven, published as a rebuttal in Nature in August 2025, found that a wideAI-generated editorial hero image for What study by researchers from Vrije Universiteit Brussel and KU Leuven, published as a rebuttal in Nature in August 2025, found that a wide.
AI 提示詞
Create a landscape editorial hero image for this Studio Global article: What study by researchers from Vrije Universiteit Brussel and KU Leuven, published as a rebuttal in Nature in August 2025, found that a wide. Article summary: The rebuttal study is Holst et al., "Dataset artefacts can partially drive the measured decline in disruption," published in Nature in August 2026 (DOI: 10.1038/s41586 026 10787 y) [4][5].. Topic tags: general web, workflow, code, regulation, benchmarks. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, click
openai.com
反駁研究為 Holst 等人,〈Dataset artefacts can partially drive the measured decline in disruption〉,2026年8月刊登於《Nature》(DOI: 10.1038/s41586-026-10787-y)。作者包括 Vincent Holst、Andres Algaba、Floriano Tori、Sylvia Wenmackers 以及 Vincent Ginis,分別來自布魯塞爾自由大學(VUB)與荷語魯汶大學(KU Leuven)。這篇論文直接挑戰2023年同樣刊登於《Nature》、由 Park、Leahey 與 Funk 所發表、宣稱科學與技術的「破壞性創新」(disruptiveness)數十年來持續走低的轟動研究。
錯誤根源: 原始作者所使用的視覺化軟體存在一個臭蟲,導致破壞性分數最高的論文在已發表的長條圖中被自動隱藏,且未發出任何警告。更嚴重的問題在於,早期的 Web of Science 資料庫記錄經常缺少完整的參考文獻後設資料——那些被列為「零篇參考文獻」的論文,被 CD 指數方法錯誤地歸類為「最具破壞性」。隨著資料庫品質逐年改善,這類虛假的最高分數越來越少,進而營造出破壞性創新持續衰退的趨勢。
錯誤根源: 原始作者所使用的視覺化軟體存在一個臭蟲,導致破壞性分數最高的論文在已發表的長條圖中被自動隱藏,且未發出任何警告。更嚴重的問題在於,早期的 Web of Science 資料庫記錄經常缺少完整的參考文獻後設資料——那些被列為「零篇參考文獻」的論文,被 CD 指數方法錯誤地歸類為「最具破壞性」。隨著資料庫品質逐年改善,這類虛假的最高分數越來越少,進而營造出破壞性創新持續衰退的趨勢。