82
Google Gone Wild:
The Digital Threat
to Reputation
In an ideal world, anyone searching for information about you
would find a fair description of who you are and what you do, with-
out too much about your personal life being revealed to complete
strangers. Unfortunately, we do not live in an ideal world. Even with-
out any malicious attacks, thousands of reputations are ruined by
negligent mistakes, sloppy programming, bad luck, and the structure
of the Internet itself. These dangers manifest themselves many ways:
Google’s search algorithm unintentionally dredges up false informa-
tion and emphasizes negative information; blog authors negligently
repeat incomplete truths; researchers fail to search deeply enough to
get a full picture; and technology echoes and repeats the most scan-
dalous facts.
CHAPTER
83Google Gone Wild: The Digital Threat to Reputation
Google Gone Wild
Justice is blind, but search engines are
blind, deaf, and dumb.
—FOLK ADAGE ADAPTED TO FIT MODERN REALITY
Almost all popular websites are automated. They are run by com-
puter software that does not need human intervention for most ac-
tions. A bank’s website uses a computer program that automatically
tells you how much money is in your account without a bank em-
ployee ever lifting a finger. An e-commerce website uses a computer
program to tell you what items are in stock and to process your pay-
ment without the need for a cashier to do anything. And a search en-
gine—like Google or Microsoft Bing—uses a computer program to
figure out what websites to return for your search query without a
human editor reviewing the results.
The system usually works because computers are incredibly
smart. When properly programmed, computers can perform literally
billions of calculations in a second. Computers can store more in-
formation on a disk the size of a deck of cards than could ever fit in
books in a public library. And computers can perform the most bor-
ing tasks day in and day out without ever tiring or needing a break.
But the system sometimes fails because computers are also in-
credibly dumb. Computers do exactly what they are told to do, no
more and no less. Computers do not ask whether their instructions
make sense. Computers do not consider other ways to perform the
same task. A computer has no empathy or emotions, nor any sense of
fairness or justice. If a computer is programmed to return the most
popular links in response to a search, then it will return the most
popular links even if they are obviously wrong, unfair, or harmful.
And computers interpret absolutely everything completely literally;
the fictional “Johnny 5” of the movie Short Circuit is perhaps the only
computer in history that “got” a joke.
1
84 Wild West 2.0
Search engines like Google are run entirely by computer pro-
grams. There is no person that reviews each query into a search en-
gine or the result that it returns. And for good reason: there are
millions upon millions of searches each day. It would be impossible
for even a team of thousands of employees to screen results for ob-
viously offensive content, let alone for humans to perform the re-
search required to determine what websites provided the “best” or
most truthful answers. The details of the programs that run Google’s
search engine are secret, but the outlines are known: Google uses one
program, a “spider,” to try to access and analyze every single page on
the Internet. This task is constantly being conducted in the back-
ground, twenty-four hours a day, seven days a week. When a user
sends a query to the Google search engine, a second computer pro-
gram uses the massive database of pages that have been found by the
“spider” to figure out what websites might be relevant. Among other
factors, Google’s computer program considers which pages are the
most “popular” pages related to a search term, on the basis of the
number of other websites that link to it. Other factors, like the fresh-
ness of the content and the number of times the search term appears
in the page, also play a role. For example, imagine searching for in-
formation about the former CEO of Microsoft by entering the
search term “Bill Gates” into Google. The computer program that
runs the Google search engine—which is based on a complex math-
ematical model known as an “algorithm”—would probably calculate
that the entry about Bill Gates in the free encyclopedia at
Wikipedia.org
2
is an extremely popular web page about Bill Gates: at
least 16,000 other websites link to Bill Gates’s biography page on
Wikipedia.
3
On the other hand, a personal blog post that happened
to mention the name “Bill Gates” would be considered relatively
unimportant, in that it would probably be linked by very few other
sites. Such a personal blog post would be unlikely to appear in the
first page of results unless it suddenly attracted a lot of attention and
links from other parts of the Internet.
The key is that Google’s process is completely automated and that
85Google Gone Wild: The Digital Threat to Reputation
it interprets “popularity” as the main measure of the importance of a
page. It does not consider whether a page is truthful, accurate, or fair;
instead, the computer simply asks whether the page is popular. To take
a phrase from linguists, Google is “descriptive” rather than “prescrip-
tive”: the results of a Google search reflect the sites that are popular
rather than Google’s opinion of what sites should be popular. This
moral neutrality can be good at times. Users can trust that the results
returned by Google and other search engines are an accurate reflection
of what exists on the Internet. The policy of neutrality has prevented
partisan censorship and manipulation of search results. For most
users, there are only a few instances of altered search results on Google,
and each alteration is marked by a prominent notice explaining why
the results were altered—usually because of copyright infringement or
child pornography. Many other search engines have followed Google’s
example and similarly limited manual changes to their results. Neu-
trality about controversial issues isn’t necessarily bad: it promotes vig-
orous political debate by exposing viewpoints that might be unpopular
today but that might become popular in the future.
But this Switzerland-like neutrality can also lead to bizarre or
even shocking results. Moral abstention can have profound impacts
on personal reputation. Unfounded attacks may quickly rise to the
top of search results because any controversy will tend to create the
appearance of popularity. Google will not remove demonstrably false,
libelous, or illegal search results without a court order. If a false at-
tack on you, your business, or your family is the first result in a
Google search, Google will tell you that you are out of luck. This pol-
icy has left many people feeling powerless: Google (the company)
will do nothing to control the harmful actions of its search engine,
under the guise of “neutrality.” The slap in the face of victims is even
stronger because of the massive market share of Google: Google
dominates the U.S. search market and many overseas markets to the
point where it is effectively the main way that information is found,
anywhere in the world. The information found in Google shapes
lives and reputations, and false negative information in Google can
86 Wild West 2.0
rend careers, families, and businesses. To victims of online defama-
tion, former U.N. Secretary General Kofi Annan’s words ring true:
“Impartiality does not—and must not—mean neutrality in the face
of evil.”
4
The Risk of Self-Reinforcing Cycles
Rankings based on popularity work well sometimes: the most im-
portant and relevant pages are often the most popular. In the Bill
Gates example, the “popularity” of a page served as a valid heuristic
(there’s that word again) to identify a site with content that is help-
ful to somebody searching for information about Bill Gates. But,
sometimes, things go wrong. The search engine’s reliance on popu-
larity can sometimes cause false, misleading, slanderous, and unfair
pages to show up at the top of a results page. The Google algorithm
has no way to know what information is true or false, fair or unfair.
It performs no original research, and it does not ask for any human’s
opinion before returning results. If a page filled with scurrilous lies
is the most “popular” page for a search term, then Google will list it
as the first result in a search.
Because Google is the most popular search engine, its emphasis
on the popularity of web pages can create a large problem for your
reputation. If a controversial or false page is returned very high in a
Google search, it is likely to be read by many users. And some of
these readers might comment on the page by posting a link on their
own blogs or by using a discussion site like Digg or Reddit. Google’s
spider will then scan these new pages commenting on the original
page and take them as evidence of the original link’s popularity, even
if the new discussion is about how controversial or false the original page
is. This leads to a self-reinforcing cycle of popularity: Lots of people
see the page because it is at the top of a Google search, so lots of peo-
ple comment on and link to the page, which makes it seem even more
popular to a Google search, which promotes the page even higher up
in Google’s rankings . . . the cycle continues. The risk of this self-
perpetuating cycle is enhanced even further because Google tends to
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.