HomeTechGadgetsThe Year the Internet Thought I Was MacKenzie Bezos

The Year the Internet Thought I Was MacKenzie Bezos


I reached out to Google’s press email three times to try to get my problem fixed, starting in November of last year. In February, feeling desperate, I asked if there was anything WIRED could do to make it stop. “Once again I’m experiencing this problem with Mackenzie Bezos, I’ve received probably 100 emails in the last few weeks from Google users who think I am her or work with her,” I wrote. “If there’s anything WIRED can do, please let me know.” I finally heard back from Google this week, after I said I planned to write about my experience.

“In situations where people are searching for something like a phone number that is not readily available online, our systems are understanding these pages (that include those exact keywords plus a phone number) to be the best matches available, even if the phone number is not the correct number for that entity,” Lara Levin, a spokesperson for Google, said in an email. Levin added that the company plans to look at ways it could improve its system “to better recognize” when a phone number should not be resurfaced as a featured snippet.

Until that happens, the only recourse is for WIRED to reconfigure its website to prevent Google from scraping my contact information. I could, of course, take down my email address and phone number altogether, but I want people to know how to reach me in order to do my job. I’m fine with strangers reaching out to me—I just don’t want them to think I’m a billionaire philanthropist.

Is there something you think we should know? You can reach the writer, who is not a billionaire, at louise_matsakis@wired.com or by phone at 347-966-3806.

Google’s stated mission has always been to “organize the world’s information.” For a long time, it mainly did that by ranking search results. Type a question or keyword into Google’s search bar, and it would return a list of websites its algorithms decided had the best chance of containing an answer to your question. You were expected to click through to find what you wanted—Google was a portal, not a destination. Eight years ago, the company fundamentally changed that arrangement when it introduced its “Knowledge Graph.”

Now, Google often tries to answer your query directly, by pulling information from sites like Wikipedia into boxes it calls “featured snippets” or “knowledge panels,” which appear above or alongside traditional search results. If you look up a celebrity’s net worth, for example, Google may scrape the information from somewhere like CelebrityNetWorth.com. Especially on mobile devices, the widgets are convenient for users in search of quick answers, sidestepping the need to dig through information published by multiple sources.

Google featured snippets and knowledge panels are controlled by algorithms, and largely not pre-screened for accuracy. That means the knowledge Google scrapes from the rest of the internet is not always correct. Plenty of errors, some more consequential than others, have made it onto Google this way. In one case, a knowledge panel incorrectly stated entrepreneur and basketball dad LaVar Ball founded the NBA. In a particularly embarrassing instance from 2018, Google’s knowledge panel for the California Republican Party listed “Nazism” among its ideologies. The company attributed the mishap to Wikipedia vandalism that then got pulled into Google, but it still angered Republican lawmakers already concerned the tech giant was biased against conservatives.

As I learned firsthand, people can spend months trying to rectify false information contained in them. In one instance reported by The Wall Street Journal last year, Google incorrectly stated the actor Paul Campbell had passed away, causing his mother to panic. Levin, the Google spokesperson, says that the company encourages people to provide it with feedback, and will “take action on these features in accordance with our policies.” Google’s policies on featured snippets don’t ban false information explicitly, although the company does say that “public interest content—including civic, medical, scientific and historical issues—should not contradict well-established or expert consensus support.”



Source link

NypTechtek
NypTechtek
Media NYC Local Family and National - World News

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Must Read