Facebook一直以来都是一家专注于增长的公司。更多的用户和更多的参与意味着更多的收入。这种一心一意的代价在麻省理工学院技术评论（mittechnologyreview）的这篇精彩故事中得到了清晰的阐述。它详细说明了该公司的人工智能团队使用机器学习解决错误信息的尝试，显然是由于Facebook不愿限制用户参与度而受到阻碍的。

“如果一个模型降低了太多的参与度，它就会被丢弃。否则，它会被部署并持续监控，”Facebook机器学习模型的作者karenhao写道。但这种做法很快就引起了一些问题。最大化参与度的模式也有利于争议、错误信息和极端主义：简单地说，人们喜欢**的东西。”

在Twitter上，郝女士指出，这篇文章不是关于“**分子（做）**的事情”，相反，她说，“这是关于好人真正努力做正确的事情。但他们被困在一个腐朽的体系中，尽最大努力推动不会改变的现状。”

这篇报道也为指责Facebook在唐纳德·特朗普（Donald Trump）担任总统期间安抚保守派的愿望导致其对右翼错误信息视而不见提供了更多证据。这似乎至少部分是由于乔尔·卡普兰（Joel Kaplan）的影响，乔尔·卡普兰曾是乔治·W·布什（George W.Bush）**的一名成员，现在是Facebook全球公共政策副总裁和“最高级别的共和党人”

All Facebook users have some 200 “traits” attached to their profile. These include various dimensi*** submitted by users or estimated by machine-learning models, such as race, political and religious leanings, socioeconomic class, and level of education. Kaplan’s team began using the traits to assemble custom user segments that reflected largely c***ervative interests: users who engaged with c***ervative content, groups, and pages, for example. Then they’d run special ****yses to see how content-moderation decisi*** would affect posts from those segments, according to a former researcher whose work was subject to those reviews.

The Fairness Flow documentation, which the Resp***ible AI team wrote later, includes a case study on how to use the tool in such a situation. When deciding whether a misinformation model is fair with respect to political ideology, the team wrote, “fairness” does not mean the model should affect c***ervative and liberal users equally. If c***ervatives are posting a greater fraction of misinformation, as judged by public c***ensus, then the model should flag a greater fraction of c***ervative content. If liberals are posting more misinformation, it should flag their content more often too.

But members of Kaplan’s team followed exactly the opposite approach: they took “fairness” to mean that these models should not affect c***ervatives more than liberals. When a model did so, they would stop its deployment and demand a change. Once, they blocked a medical-misinformation detector that had noticeably reduced the reach of anti-vaccine campaigns, the former researcher told me. They told the researchers that the model could not be deployed until the team fixed this discrepancy. But that effectively made the model meaningless. “There’s no point, then,” the researcher says. A model modified in that way “would have literally no impact on the actual problem” of misinformation.

报道还说，Facebook的人工智能研究人员在算法偏差问题上的工作已经展开，在这个问题上，机器学习模型无意中歧视了某些用户群体，至少在一定程度上是为了抢先这些反保守情绪的指责，并阻止美国**的潜在监管。但是，把更多的资源投入到偏见中，就意味着忽视了涉及错误信息和仇恨言论的问题。尽管该公司口口声声说要AI公平，但指导原则，郝说，仍然一如既往：增长，增长，增长。

[T]esting algorithms for fairness is still largely optional at Facebook. None of the teams that work directly on Facebook’s news feed, ad service, or other products are required to do it. Pay incentives are still tied to engagement and growth metrics. And while there are guidelines about which fairness definition to use in any given situation, they aren’t enforced.

你可以在麻省理工学院技术评论上阅读郝的完整故事。