<?xml version="1.0" encoding="UTF-8" standalone="yes"?><oembed><version><![CDATA[1.0]]></version><provider_name><![CDATA[The Dish]]></provider_name><provider_url><![CDATA[http://dish.andrewsullivan.com]]></provider_url><author_name><![CDATA[Andrew Sullivan]]></author_name><author_url><![CDATA[https://dish.andrewsullivan.com/author/sullydish/]]></author_url><title><![CDATA[Searching For Prejudice]]></title><type><![CDATA[link]]></type><html><![CDATA[<p>Tom Vanderbilt <a href="http://www.wired.co.uk/magazine/archive/2013/01/features/the-future-of-search" target="_self">highlights</a> an ethical problem for search engines:</p>
<blockquote><p>A few years ago, Google faced controversy when it was revealed a search for the word &#8220;Jew&#8221; returned several anti-Semitic websites. Through brute algorithmic logic, it made sense: the sort of people who use the word &#8220;Jew&#8221; tend to have those sorts of proclivities. Now a search for that word leads in short order to an explanatory page from Google (which states, in part: &#8220;Someone searching for information on Jewish people would be more likely to enter terms like &#8216;Judaism&#8217;, &#8216;Jewish people&#8217; or &#8216;Jews&#8217; than the single word &#8216;Jew&#8217;. In fact, prior to this incident, the word &#8216;Jew&#8217; only appeared about once in every ten million search queries&#8221;). While [<a href="http://singhal.info/">Amit Singhal</a>, a senior vice president at Google] says that &#8220;time and again we decided that Google shouldn&#8217;t intervene in the [search] process,&#8221; it is constantly shaping the world &#8212; for example, it recently struck the peer-sharing site <a href="http://www.wired.co.uk/news/archive/2012-12/20/pirate-bay-proxy-shuts-down">The Pirate Bay</a> from autocomplete &#8212; and the fact that &#8220;Holocaust denial&#8221; yields very different results than &#8220;Holocaust lie&#8221; is as much a social as a search issue.</p></blockquote>
]]></html></oembed>