Far West Consulting Far West Consulting
EN 繁中

網站上的每一條引文，都集中在這一頁。

我們對廠商的生產力宣稱採取折扣態度。我們不發表沒有來源的量化結果。下方這頁就是這項規則的結構性承諾——Far West Consulting 網站上引用的每一條研究文獻，依編號排列，附公開的驗證狀態。

標記為 上線時已驗證 的引文，於上線前已對照原始出版物確認。隨網站演進，這套上線前驗證結構持續適用於每一條未來新增的引文；如某條引文在原始來源確認之前就被加入，會在下方 待驗證 區塊呈現，且生產部署會被檔住，直到處理完成。

我們如何在自己的工作中使用 AI——具名工具、具名產出、具名流程，請見 我們如何使用 AI。

  1. 註腳 1 · 上線時已驗證 · 2026-05-04

    Gartner press release, June 25, 2025: "Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027." (Source: gartner.com newsroom.)

    來源： gartner.com newsroom

  2. 註腳 2 · 上線時已驗證 · 2026-05-04

    Microsoft 2025 Work Trend Index, APAC release (April 30, 2025): "APAC emerges as global AI frontrunner."

    來源： Microsoft Work Trend Index 2025 APAC release

  3. 註腳 3 · 上線時已驗證 · 2026-05-04

    MIT NANDA / State of AI in Business 2025.

    來源： MIT NANDA State of AI in Business 2025

  4. 註腳 4 · 上線時已驗證 · 2026-05-04

    Grant Thornton AI Governance Readiness Survey, April 2026.

    來源： Grant Thornton 2026 survey

  5. 註腳 5 · 上線時已驗證 · 2026-05-05

    KPMG, The importance of value streams in the age of AI (2024). Author: Adrian Clamp. Source: kpmg.com/xx/en/our-insights/ai-and-technology/the-importance-of-value-streams-in-the-age-of-ai.html.

    來源： kpmg.com/xx/en/our-insights/ai-and-technology/the-importance-of-value-streams-in-the-age-of-ai.html

  6. 註腳 6 · 上線時已驗證 · 2026-05-05

    Microsoft Research, GitHub Copilot randomized controlled trial — n=200+ engineers, randomized. Telemetry showed no measurable productivity improvement, despite engineers self-reporting time savings (2024).

    來源： Microsoft Research GitHub Copilot RCT 2024

  7. 註腳 7 · 上線時已驗證 · 2026-05-05

    Brynjolfsson et al., Quarterly Journal of Economics 2025 — field experiment, n=5,172 customer service agents. AI assistance produced +30% productivity for low-skill workers, ~0% for high-skill workers, and a measurable quality decline at the top of the skill distribution.

    來源： Brynjolfsson Quarterly Journal of Economics 2025

  8. 註腳 8 · 上線時已驗證 · 2026-05-12

    Acemoglu, The Simple Macroeconomics of AI, NBER Working Paper 32487 (2024) — total factor productivity contribution at most 0.66% over 10 years (≈0.06%/year).

    來源： NBER Working Paper 32487

  9. 註腳 9 · 上線時已驗證 · 2026-05-14

    Nielsen, 4 Metaphors for Working with AI: Intern, Coworker, Teacher, Coach (UX Tigers).

    來源： UX Tigers / Nielsen

  10. 註腳 10 · 上線時已驗證 · 2026-05-05

    Regulation (EU) 2024/1689 of the European Parliament and of the Council (the EU AI Act), Article 4 — AI literacy. Imposes on providers and deployers of AI systems the obligation to ensure a sufficient level of AI literacy of their staff. Adopted June 2024; AI-literacy obligation in force as of 2 February 2025. Source: eur-lex.europa.eu/eli/reg/2024/1689.

    來源： eur-lex.europa.eu/eli/reg/2024/1689

  11. 註腳 11 · 上線時已驗證 · 2026-05-14

    US Department of Labor, AI Literacy Framework — five core competency areas (February 2026).

    來源： US DOL AI Literacy Framework

  12. 註腳 12 · 上線時已驗證 · 2026-05-05

    Reif, J.A., Larrick, R.P., & Soll, J.B. (2025). Evidence of a social evaluation penalty for using AI. Proceedings of the National Academy of Sciences. n>4,400 across four experiments. DOI: 10.1073/pnas.2426766122.

    來源： doi.org/10.1073/pnas.2426766122

  13. 註腳 13 · 上線時已驗證 · 2026-05-13

    The Impostor Phenomenon of Workplace Artificial Intelligence Augmentation. Academy of Management Proceedings, Vol. 2024 (148bp). Source: journals.aom.org/doi/abs/10.5465/AMPROC.2024.148bp.

    來源： journals.aom.org/doi/abs/10.5465/AMPROC.2024.148bp

  14. 註腳 14 · 上線時已驗證 · 2026-05-05

    Slack Workforce Lab. The Fall 2024 Workforce Index. November 12, 2024. Source: slack.com/blog/news/the-fall-2024-workforce-index-shows-executives-and-employees-investing-in-ai-but-uncertainty-holding-back-adoption.

    來源： slack.com/blog/news/the-fall-2024-workforce-index-shows-executives-and-employees-investing-in-ai-but-uncertainty-holding-back-adoption

  15. 註腳 15 · 上線時已驗證 · 2026-05-13

    Gillespie, N. & Lockey, S. Trust, Attitudes and Use of Artificial Intelligence: A Global Study 2025. University of Melbourne / KPMG. April 2025. Source: kpmg.com/xx/en/media/press-releases/2025/04/trust-of-ai-remains-a-critical-challenge.html.

    來源： kpmg.com/xx/en/media/press-releases/2025/04/trust-of-ai-remains-a-critical-challenge.html

  16. 註腳 16 · 上線時已驗證 · 2026-05-14

    Kusumegi, K. et al. Scientific production in the era of large language models. Science, Vol. 390, Issue 6779. December 24, 2025. DOI: 10.1126/science.adw3000. Source: science.org/doi/10.1126/science.adw3000.

    來源： science.org/doi/10.1126/science.adw3000

  17. 註腳 17 · 上線時已驗證 · 2026-05-14

    Axios HQ. 2025 State of Internal Communications. Survey of 457 business executives and 813 knowledge workers across U.S. industries. Source: axioshq.com/research/state-of-internal-communications-report.

    來源： axioshq.com/research/state-of-internal-communications-report

  18. 註腳 18 · 上線時已驗證 · 2026-05-14

    Model AI Governance Framework for Generative AI, published by the AI Verify Foundation and Singapore's Infocomm Media Development Authority (IMDA), May 2024. Builds on the traditional Model AI Governance Framework (Second Edition, Personal Data Protection Commission, January 2020). A principle-based AI governance reference. Source: aiverifyfoundation.sg.

    來源： aiverifyfoundation.sg

  19. 註腳 19 · 上線時已驗證 · 2026-05-14

    William Bridges, Transition Model. Three phases: Ending / Losing / Letting Go, the Neutral Zone, and the New Beginning. Distinguishes external change (the situational event) from internal transition (the psychological process people move through). Canonical reference: William Bridges, Managing Transitions: Making the Most of Change (1991; current edition with Susan Bridges). Source: wmbridges.com.

    來源： wmbridges.com/about/what-is-transition/

  20. 註腳 20 · 上線時已驗證 · 2026-05-14

    Cynefin Framework, created by David J. Snowden in 1999. A sense-making framework with four primary domains: Clear, Complicated, Complex, and Chaotic — each calling for a different decision posture, plus a central state of Disorder. Foundational article: Snowden, D.J. & Boone, M.E., "A Leader's Framework for Decision Making," Harvard Business Review, November 2007. Current stewardship: The Cynefin Company (formerly Cognitive Edge). Source: thecynefin.co.

    來源： thecynefin.co/about-us/about-cynefin-framework/

我們如何使用來源

有原始來源時優先。同行評審的期刊論文與公開揭露方法的機構研究。我們不把廠商發表的生產力宣稱當作證據——那些可作為產業訊號，但不能成為服務提案的基礎。

公開揭露方法。當一項研究的樣本量、時間範圍或方法影響到該如何應用其發現時，我們在引文中呈現這些細節，而不是把它們埋在一句宣稱裡。讀者不必把論文翻出來就能判斷份量。

上線前驗證。網站上的每一條引文都會經過建置時檢查（scripts/check-verifications.mjs）。生產部署在每一個註腳完成驗證前會被檔住。上方這頁是這項內部紀律的對外呈現。

廠商宣稱折扣規則也適用我們自己。Far West Consulting 不會在沒有同步揭露方法、樣本與時間範圍的情況下，發表自己合作的量化結果宣稱。你正在閱讀的這頁就是這項結構性承諾。