打字猴:1.702427086e+09
1702427086
1702427087 销声匿迹:数字化工作的真正未来 [:1702424986]
1702427088 第一章 环路中的人类
1702427089
1702427090 1.M. Six Silberman, “Human-Centered Computing and the Future of Work: Lessons from Mechanical Turk and Turkopticon, 2008–2015”(PhD diss., University of California, Irvine, 2015).
1702427091
1702427092 2.Brad Stone, The Everything Store: Jeff Bezos and the Age of Amazon (New York: Little, Brown, 2013).
1702427093
1702427094 3.Ibid.
1702427095
1702427096 4.Ibid.
1702427097
1702427098 5.据说,MTurk是杰夫·贝佐斯最喜欢的项目,他把MTurk当成一个工具,不仅可以内部使用,还可以作为一个新的市场给商家创造收入。商家也需要清理自己的清单。参阅Stone,The Everything Store。
1702427099
1702427100 6.Daniel W. Barowy et al., “VoxPL: Programming with the Wisdom of the Crowd,”in CHI ’17: Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems (New York: ACM, 2017), 2347–58, https://doi.org/10.1145/3025453; Suri, “Computing with the Crowd,”101.
1702427101
1702427102 7.很多人批评亚马逊给工作平台取名过于随便,但我们还不清楚最初亚马逊是否打算公布MTurk。从2005年发布以来,MTurk一直处于测试状态。
1702427103
1702427104 8.Kevin P. Murphy, Machine Learning: A Probabilistic Perspective (Cambridge, MA: MIT Press, 2012).
1702427105
1702427106 9.Fei-Fei Li, “ImageNet: Where Have We Been? Where Are We Going?,”ACM Learning Webinar, https://learning.am.org/, accessed September 21, 2017; Deng et al., “ImageNet: A Large-Scale Hierarchical Image Database,”in 2009 IEEE Conference on Computer Vision and Pattern Recognition (Piscataway, NJ: IEEE), 248–55.
1702427107
1702427108 10.从2010年到2016年,精度从72%提高到97%。
1702427109
1702427110 11.事实上,亚历山大·威斯纳–格罗斯指出:“限制人工智能发展的因素可能是数据集,而非算法。”参阅Alexander Wissner-Gross,“2016:What Do You Consider the Most Interesting Recent[Scientific]News?What Makes It Important?”Edge,https://www.edge.org/response-detail/26587,访问于2018年10月21日。
1702427111
1702427112 12.为了激励研究人员使用这些数据集,李飞飞和她的同事每年组织一场挑战赛,从世界各地的研究团队中选出最佳的图像识别算法。科学家朝着这一目标所取得的进展是惊人的。过去8年,在一年一度的ImageNet挑战赛中,错误减少到约1/10,识别图像的精度提高了约3倍。最终,视觉算法的错误率比人工还要低。科学家在8年的挑战赛中取得的算法和工程上的进步,在很大程度上推动了最近神经网络的成功,即所谓的“深度学习革命”,它将影响多个领域和多种问题。
1702427113
1702427114 13.Djellel Difallah, Elena Filatova, and Panos Ipeirotis, “Demographics and Dynamics of Mechanical Turk Workers,”in Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining (New York: ACM, 2018), 135–43, https://doi.org/10.1145/3159652.3159661.
1702427115
1702427116 14.Ibid.
1702427117
1702427118 15.Panos Ipeirotis, “How Many Mechanical Turk Workers Are There?,”A Computer Scientist in Business School (blog), January 29, 2018, http://www.behind-the-enemy-lines.com/.
1702427119
1702427120 16.我们有意让工人自己报告自己的位置,这样他们报告的精确度就可以根据他们的意愿而定。这张地图允许工人在需要的时候搜索、缩放和平移,所以他们可以自由地在他们的房子、社区、区县或者城市等坐标上钉上大头针。放置一个大头针并点击“保存”后,工人就会看到一张必应世界地图,上面显示了之前50名做这个任务的工人的大头针。为了保护工人的隐私,每个大头针都随机移动了一点,并且在页面上提醒了他们。完成这项任务非常容易,只需要不到一分钟,我们为每个任务支付25美分。
1702427121
1702427122 17.Hara et al., “A Data-Driven Analysis of Workers’ Earnings on Amazon Mechanical Turk,”2018 CHI Conference on Human Factors in Computing Systems, Paper No. 449, 2018.
1702427123
1702427124 18.浏览器插件Turkopticon等工具帮助工人通过评论和评级系统共享请求者的信息,评级内容包括请求者的沟通、定价、公平性以及付款速度和对工人询问的响应时间。工人也有精心设计的、自己主持的论坛,可以在收费墙后面讨论更多细节。请求者的口碑评级和工人论坛都不能直接在MTurk上使用。参阅Lilly C. Irani and M. Six Silberman,“Turkopticon:Interrupting Worker Invisibility in Amazon Mechanical Turk,”in CHI ’13:Proceedings of the SIGCHI Conference on Human Factors in Computing Systems(New York:ACM,2013)。
1702427125
1702427126 19.参阅Siou Chew Kuek,Cecilia Paradi-Guilford,Toks Fayomi,Saori Imaizumi,Panos Ipeirotis,Patricia Pina,and Manpreet Singh,“The Global Opportunity in Online Outsourcing,”World Bank Group,June 2015,and Panos Ipeirotis,“How Big Is Mechanical Turk?,”A Computer Scientist in Business School(blog),November 8,2012,http://www.behind-the-enemy-lines.com/2012/11/is-mechanical-turk-10-billion-dollar.html。
1702427127
1702427128 20.如果任务需要分发给10个以上的工人,亚马逊收取20%的额外费用。同一个任务,每个工人只能被分配一次。
1702427129
1702427130 21.讽刺的是,我们作为微软公司的雇员,和那些按需工人受制于相同的保密协议。因此,我们能分享的关于UHRS的运营细节十分有限。
1702427131
1702427132 22.Annalee Newitz, “The Secret Lives of Google Raters,”Ars Technica, April 27, 2017, https://arstechnica.com/features/2017/04/the-secret-lives-of-google-raters/.
1702427133
1702427134 23.Gillespie, Custodians of the Internet; Jeremias Prassl, Humans as a Service: The Promise and Perils of Work in the Gig Economy (Oxford, England: Oxford University Press, 2018); Sarah T. Roberts, “Digital Detritus: ‘Error’ and the Logic of Opacity in Social Media Content Moderation,”First Monday 23, no. 3 (March 1, 2018), http://firstmonday.org/ojs/index.php/fm/article/view/8283; Sarah T. Roberts, “Social Media’s Silent Filter,”The Atlantic, March 8, 2017, https://www.theatlantic.com/technology/archive/2017/03/commercial-content-moderation/518796/.
1702427135
[ 上一页 ]  [ :1.702427086e+09 ]  [ 下一页 ]