1704605412
1704605413
1704605414
1704605415
1704605416
图20 2016年美国各州最常见的职业
1704605417
1704605418
1704605419
1704605420
来源： S. Ruggles et al., 2018, IPUMS USA, version 8.0 (dataset), https://usa.ipums.org/usa/。
1704605421
1704605422
美国中西部的一名卡车司机不大可能成为硅谷的一名软件工程师。他可能找到一份看门的工作，也可能从事场地维护的工作，保持公园、房屋和公司的干净整洁。我们估计，这些工作在下一次自动化浪潮中都不会受影响。如果他成了看门人，等于是将一份年收入41,340美元（2016年的年收入中位数）的卡车司机的工作换成了一份年收入24,190美元的工作。如果他成了场地维护工人，他每年能挣26,830美元。再或者，他可能成为一名年收入46,890美元的社工，但那样的话他就需要获得一个大学学位。
1704605423
1704605424
列昂惕夫曾开玩笑说，如果马获得了投票权，它们就不太可能从农场消失了。虽然美国中产阶级几乎不可能遭遇农场马匹那样的命运，但我们也不指望美国人会轻易接受工资的下降。如果自动化只会暂时降低人们的收入，他们可能愿意接受自动化。但如果他们的收入在数年甚至数十年内似乎都不可能恢复，他们就更有可能抵制自动化。事实上，如果个人对市场的裁决感到不满，他们要么会试图抵制技术，要么会通过非市场的机制和政治激进主义来寻求更多的再分配。我们在第三章中就讨论过，卢德主义者和其他群体曾激烈反抗那些威胁他们生计的机器的推广。除了发动暴乱，他们也曾向议会请愿，呼吁政府限制劳动取代技术的引进。但由于缺少政治影响力，他们没有成功的希望。如今的劳动者不仅对政府必须提供的东西有更高的期望，他们也拥有了政治权利。
1704605425
1704605426
注释
1704605427
1704605428
1 E. Brynjolfsson and A. McAfee, 2017, Machine, Platform, Crowd: Harnessing Our Digital Future (New York: Norton), 71–73.
1704605429
1704605430
2 C. E. Shannon, 1950, “Programming a Computer for Playing Chess,” Philosophical Magazine 41 (314): 256–75.
1704605431
1704605432
3 C. Koch, 2016, “How the Computer Beat the Go Master,” Scientific American 27 (4): 20.
1704605433
1704605434
4 F. Levy and R. J. Murnane, 2004, The New Division of Labor: How Computers Are Creating the Next Job Market (Princeton, NJ: Princeton University Press).
1704605435
1704605436
5 E. Brynjolfsson and A. McAfee, 2014, The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies (New York: W. W. Norton), chapter 3, Kindle.
1704605437
1704605438
6 Koch, 2016, “How the Computer Beat the Go Master,” 20.
1704605439
1704605440
7 M. Fortunato et al. 2017, “Noisy Networks for Exploration,” preprint, submitted, https://arxiv.org/abs/1706.10295.
1704605441
1704605442
8 Cisco, 2018, “Cisco Visual Networking Index: Forecast and Trends, 2017–2022,” (San Jose, CA: Cisco), https://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/complete-white-paper-c11-481360.html.
1704605443
1704605444
9 P. Lyman and H. R. Varian, 2003, “How Much Information?,” berkeley.edu/research/projects/how-much-info-2003.
1704605445
1704605446
10 A. Tanner, 2007. “Google Seeks World of Instant Translations,” Reuters, March 27.
1704605447
1704605448
11 Y. Wu et al., 2016, “Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation,” preprint, submitted October 8, https://arxiv.org/pdf/1609.08144.pdf.
1704605449
1704605450
12 I. M. Cockburn, R. Henderson, and S. Stern, 2018, “The Impact of Artificial Intelligence on Innovation (Working Paper 24449, National Bureau of Economic Research, Cambridge, MA).
1704605451
1704605452
13 E. Brynjolfsson, D. Rock, and C. Syverson, forthcoming, “Artificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics,” in The Economics of Artificial Intelligence: An Agenda, ed. Ajay K. Agrawal, Joshua Gans, and Avi Goldfarb (Chicago: University of Chicago Press), figure 1.
1704605453
1704605454
14 “Germany Starts Facial Recognition Tests at Rail Station,” 2017, New York Post, December 17.
1704605455
1704605456
15 N. Coudray et al., 2018, “Classification and Mutation Prediction from Non–Small Cell Lung Cancer Histopathology Images Using Deep Learning,” Nature Medicine 24 (10): 1559–1567.
1704605457
1704605458
16 A. Esteva et al., 2017, “Dermatologist-Level Classification of Skin Cancer with Deep Neural Networks,” Nature 542 (7639): 115.
1704605459
1704605460
17 W. Xiong et al., 2017, “The Microsoft 2017 Conversational Speech Recognition System,” Microsoft AI and Research Technical Report MSR-TR-2017-39, August, https://www.microsoft.com/en-us/research/wp-content/uploads/2017/08/ms_swbd17-2.pdf.
1704605461