Chapter 24 — The Second Watershed

10 references

#	Reference	Links	Status
1	Sida Peng, Eirini Kalliamvakou, Peter Cihon, and Mert Demirer, "The Impact of AI on Developer Productivity: Evidence from GitHub Copilot," arXiv preprint 2302.06590, February 2023, https://arxiv.org/abs/2302.06590. The controlled trial reported 55.8% faster completion of an HTTP-server task by the Copilot-assisted group versus the unassisted control. arxiv.org	Source	🟡 No archive
2	Erik Brynjolfsson, Danielle Li, and Lindsey R. Raymond, "Generative AI at Work," Quarterly Journal of Economics 140, no. 2 (2025); originally NBER Working Paper 31161, April 2023, https://www.nber.org/papers/w31161. n = 5,179 customer-support agents at a Fortune 500 software firm; 14% mean increase in issues resolved per hour, 35% for the bottom-skill quintile. For the visible boundary on customer-service automation, see also Klarna's 2024 AI-agent rollout and its 2025 partial reversal, reported in the Financial Times and Bloomberg, May 2025. nber.org	Source	🟡 No archive
3	Fabrizio Dell'Acqua, Edward McFowland III, Ethan Mollick, Hila Lifshitz-Assaf, Katherine Kellogg, Saran Rajendran, Lisa Krayer, François Candelon, and Karim Lakhani, "Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality," Harvard Business School Working Paper 24-013, September 2023, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4573321. n ≈ 758 BCG consultants using GPT-4; roughly 25% faster task completion, ~12% more tasks completed, and roughly 40% higher quality on tasks inside the model's competence frontier. papers.ssrn.com	Source	🟡 No archive
4	Tooling, for example, cannot be shortened: cutting metal or CVD (Chemical Vapor Deposition) of chip layers require set times.	—	— Cross-reference
5	Eaton Corporation, "About Us," https://www.eaton.com/us/en-us/company/about-us.html. eaton.com	Source	🟡 No archive
6	"Eaton's Generative AI Cuts Product Design Time by 87%," aPriori case study, 2024, https://www.apriori.com/resources/case-study/eatons-generative-ai-cuts-product-design-time-by-87/. apriori.com	Source	🟡 No archive
7	Augury machine-health customer case studies. Colgate-Palmolive – "Superior Machine Insights Help Colgate-Palmolive Reimagine a Healthier Future for All," https://www.augury.com/success-stories/colgate-palmolive-optimizes-maintenance-using-ai-insights/. PepsiCo – "PepsiCo's AI Innovations Boost Manufacturing Efficiency," https://www.augury.com/success-stories/pepsicos-manufacturing-innovation-leads-to-tangible-roi/. augury.com	Source 1 Source 2	🟡 No archive
8	The √N rule is a standard result in signal processing: averaging N independent measurements of a common signal raises the signal-to-noise ratio in proportion to √N, because the coherent signal grows with N while uncorrelated noise grows only with √N. See "Signal averaging," Wikipedia, https://en.wikipedia.org/wiki/Signal_averaging; for a reference treatment, Julius S. Bendat and Allan G. Piersol, Random Data: Analysis and Measurement Procedures, 4th ed. (Wiley, 2010). en.wikipedia.org	Source 1 Source 2	🟡 No archive
9	Testing many candidate patterns inflates false positives, so the detection threshold must rise to compensate – but only with the square root of the logarithm of the number of hypotheses tested (\propto \sqrt{2\ln M}). On controlling false discoveries across many tests, see Yoav Benjamini and Yosef Hochberg, "Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing," Journal of the Royal Statistical Society, Series B 57, no. 1 (1995): 289–300, https://doi.org/10.1111/j.2517-6161.1995.tb02031.x. The same search penalty appears in physics as the look-elsewhere effect: Eilam Gross and Ofer Vitells, "Trial Factors for the Look Elsewhere Effect in High Energy Physics," European Physical Journal C 70 (2010): 525–530, https://doi.org/10.1140/epjc/s10052-010-1470-8. doi.org	Source 1 Source 2	🟡 No archive
10	On detecting departures from learned-normal behavior with no predefined fault signature, see Varun Chandola, Arindam Banerjee, and Vipin Kumar, "Anomaly Detection: A Survey," ACM Computing Surveys 41, no. 3 (2009): article 15, https://doi.org/10.1145/1541880.1541882. doi.org	Source	🟡 No archive

Found a broken link or outdated source?

Report on GitHub • Email report