Pluralistic: Unpermissioned research

ID: 5e0d2b33
原文链接: https://pluralistic.net/2026/09/02/scrape-scrope-scrap
作者: Cory Doctorow
日期: 2026-09-02
抓取时间: 2026-09-02T15:31:00Z
分类: industry
来源类型: article
语言: zh
标签: scraping, copyright, labor, research-freedom, platform-power
质量评分: 4/5

中文导读

Cory Doctorow 9 月 2 日的长 essay 主题不是 scraping 是否合法，而是"用 property rights 框架去讨论 scraping 会让工人输掉这场仗"他先把财产论拆开：privacy 是人权不是财产权，人权不能买卖你不能把自己卖成奴隶不能把肾卖掉抵房租把隐私降格成财产权等于直接帮 Facebook 这种 surveillance capitalism 开脱然后落到 AI 训练这件事：媒体集团（RIAA CEO Mitch Glazier 年薪 140 万美元是典型代表）把 AI 抓取叫"unlicensed taking"是财产权话术，结果是工人沦为"决定哪家公司能拿到他们被 AI 替代的许可费"的标的物好莱坞编剧大罢工反而是行业里唯一一次靠 sectoral bargaining + 工会力量而不是财产权话术赢的Doctorow 把 scraping 重新定位成公共利益工具：Ethan Zuckerman 刚把整个互联网平台权力研究实验室从 UMass 搬到 McGill（加拿大政府投入 5.04 亿加元），原因就是美国 Trump 当局对"未经许可的研究"的政治压力，能用来追踪 Zuckerberg/Musk 怎么在 feed 里压批评怎么放种族主义和阴谋论Wikipedia 正在被 AI scraper 用 DDoS 级别的流量打挂，问题不是"被偷"，而是"被淹没"

一句话点评

Cory Doctorow 9 月 2 日的长 essay 主题不是 scraping 是否合法，而是"用 property rights 框架去讨论 scraping 会让工人输掉这场仗"他先把财产论拆开：privacy 是人权不是财产权.

English Abstract / Summary

Cory Doctorow argues the property-rights frame for AI scraping makes workers lose, because privacy is a right not property and cannot be sold like labor. Media groups pushing "unlicensed taking" rhetoric end up putting workers in the position of bargaining subjects for AI-replacement licensing fees; only the Hollywood writers' strike won via sectoral bargaining and union power rather than property-rights framing. Doctorow reframes scraping as a public-interest research tool, citing Ethan Zuckerman's move of the entire platform-power lab from UMass to McGill with $504M Canadian government funding under US political pressure on "unpermissioned research," and notes Wikipedia is being DDoS'd by AI scrapers rather than merely "stolen from."

Obsidian Notes

由 daily-intake-evening 2026-09-02 cron 从当日 Obsidian 摘要（论文流水线 / AK-RSS / ClawFeed / X 书签消化）发现并入库存量阶段。
中文导读与判断均锚定在条目已有摘要、源页面正文、作者、日期与分类信息；未补充源页面之外的实验细节。