More evidence Google's not neutral... and seeks to be the supreme arbiter of "truth" on the Internet

Google again has blocked search advertising that promotes political views that Google does not share.

• On the morning of the FCC's net neutrality vote last week, Bret Glass of ExtremeTech.com tried to advertise his white paper, that advocated a light regulatory touch, on Google Adwords only to find that Google blocked his ad as not meeting their "guidelines." (See Mr. Glass' full recounting of this non-neutral content blocking incident at the end of this post.) 
  • This is not the first time Google has blocked content that did not comport with Google's political/policy agenda. For example, Google blocked anti-Moveon.org ads proposed by a U.S. Senator's campaign.

The "relevance" of this evidence of net neutrality violations by Google, to the FCC's just-proposed net neutrality regulations, is that the FCC's clearly stated purposes are: to prevent companies with market power from infringing on free speech and to ensure that those with market power are transparent about their market practices that affect the free flow of information. 

• More than any other Internet player, the facts and evidence show (DOJ) that Google has market power.
• Moreover, advertisers routinely complain that Google's ad auction process is a non-transparent "black box" because it has a "quality score" that Google can alter to arbitrarily bury a website at the bottom of its ranking and ensure that it is not found. TradeComet filed an private antitrust suit alleging this.
• Furthermore, Google admits to providing $270m in free adwords to Google-friendly organizations, but will not disclose who they are because that might show that Google is skirting lobbying, election, and campaign finance disclosure laws by laundering their Adwords market power through astroturf groups supporting Google's public policy agenda. (In a Google keyword auction, if Google or one of its preferred surrogates is "bidding" on a particular keyword, no one can out bid "the house.") 

Much more troubling is Google's public comments indicating their "black box" search engine and auction processes are not neutral, but are affected by Google's political/policy biases.

• In a Google's blogpost in February, Google Senior Vice President Jonathan Rosenberg candidly explained Google's power over finding the world's information: "We won't (and shouldn't) try to stop the faceless scribes of drivel, but we can move them to the back row of the arena." 
• Google CEO Eric Schmidt, on judging political "truth" in an FT interview: 
  • "He forecast that, within five years, “truth predictor” software would “hold politicians to account”. Voters would be able to check the probability that apparently factual statements by politicians were actually correct, using programmes that automatically compared claims with historic data, he said."
  • "Politicians “don’t in general understand the implications” of the internet, Mr Schmidt argued. “One of my messages to them is to think about having every one of your voters online all the time, then inputting ‘is this true or false?’ We [at Google] are not in charge of truth but we might be able to give a probability.”  
• This Google role as arbiter of political truth is still at the forefront of Google CEO Schmidt's agenda because only last week he told the Washington Post in a meeting reported by Mike Musgrove that:

  • "I spend so much time in Washington now because of the work that I've been doing, I deal with all these people who make assertions without fact," he said. Policy people "will hand me some report that they wrote or they'll make some assertion, and I'll say, 'Well, is that true?' -- and they can't prove it." Perhaps that could change some day, he suggested. Technology could help."

In conclusion, the evidence mounts that neither Google's search engine nor its keyword auction system are neutral. It is also clear that Google has long aspired, and still aspires to help its users discern what Google believes is "the truth" in the political/policy arenas.

Given that Google, the Adminstration and the FCC are all on the same page that more transparency is good, why shouldn't Google be more transparent in how its secret algorithms and black box processes determine what information is determined to be "drivel" that Google needs to send to "the back row of the arena" of politics and policy in Washington, as Google's Mr. Rosenberg so eloquently and trenchantly said earlier this year to all Google employees.

• Does Google believe Mr. Glass' white paper is "drivel" because it disagrees with Google's position?

Given that Googleopoly is the world's dominant information gatekeeper, isn't it relevant for the world to know what variables drive what political/policy information the world finds?

• And more importantly what are the variables/biases that Google believes will enable Google's "truth predictor software" to accurately predict the "truth?"

I guess the world would be a simpler place for Google, if people did not have to bother with judgement, critical thinking or learning, because Google could algorithmically provide for them what the "truth" is for any political or policy question.

 ********************************

From: Brett Glass <brett@lariat.net>
Date: October 24, 2009 16:23:04 EDT
To: dave@farber.net, Ip ip <ip@v2.listbox.com>
Subject: Google "disapproves" net neutrality ad

Dave, and everyone:

Several months ago, I noticed that when one typed the phrase "network neutrality" into Google's search engine, the top listed results all advocated Google's regulatory agenda. In fact, Google was contributing free advertising to groups which advocated "network neutrality" regulation (see http://www.google.com/grants/). This gave them an unfair advantage. They could place very high "bids" but not be charged for them, so their ads were guaranteed to show up on Google's result pages whereas paid ads might not.

I therefore created a simple advertisement, using Google's "AdWords" facility, which pointed to a white paper I had written on the issue. This white paper advocated regulation only in instances of anticompetitive practices or market failure, and recommended that content and application providers (including Google) who could serve as gatekeepers be scrutinized for anti-consumer practices as well as ISPs. (You can see the paper at

http://www.brettglass.com/principles.pdf

on my Web site.)

Then, on the morning of the FCC's vote on a Notice of Proposed Rule Making on "network neutrality" regulation, I received the following notice from Google:

Subject: Your Google AdWords Approval Status

 

Hello,

 

Thank you for advertising with Google AdWords. After reviewing your

account, we've found that one or more of your ads or keywords doesn't

meet our guidelines.

I entered Google's Web interface, and discovered that -- during the days before today's FCC meeting -- the ad had received large numbers of clickthroughs. This number dropped to zero, of course, when Google blocked the ad.

I further discovered that the supposed "reason" for blocking the ad was given as "Destination URL not working." So, I checked the Web server and its logs. I discovered that in fact the server had been working perfectly. What's more, according to the log, Google's "bots" had visited the URL of the document twice less than 24 hours before and had issued conditional "GET" requests. Each received a "304" response (meaning that the page was present and had not been recently changed). The log entries looked like this:

crawl-66-249-67-74.googlebot.com - - [21/Oct/2009:10:49:45 -0600] "GET /principles.pdf HTTP/1.1" 304 0 "-" "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)"

rate-limited-proxy-209-85-238-139.google.com - - [21/Oct/2009:21:21:53 -0600] "GET /principles.pdf HTTP/1.1" 304 0 "-" "
AdsBot-Google (+http://www.google.com/adsbot.html)"

Google's statistics for the ad showed that after it had received these POSITIVE responses, the clickthrough rate dropped to zero as Google dropped the ad.

I immediately resubmitted the ad to Google's advertising system, and at first the site said that the ad was "pending review." When I checked later, Google's site said that it had again rejected the ad again because its destination URL was supposedly "unreachable," even though Google's "bots" had made two more successful visits to the document since that time:

crawl-66-249-67-74.googlebot.com - - [22/Oct/2009:16:15:09 -0600] "GET /principles.pdf HTTP/1.1" 200 31639 "-" "Mozilla/
5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)"

rate-limited-proxy-209-85-238-139.google.com - - [23/Oct/2009:00:55:13 -0600] "GET /principles.pdf HTTP/1.1" 200 31639 "
-" "AdsBot-Google (+http://www.google.com/adsbot.html)"

I have resubmitted the ad a third time, and at this writing it is still "pending approval" and is not appearing.

In the meantime, other ads, pointing to other pages on the same group of servers, have continued to be active. Only the one ad pointing to the paper regarding "network neutrality" regulation -- which advocates that content and application providers be scrutinized to the same extent as ISPs for anti-consumer behavior -- is blocked.

It is noteworthy that Google has advocated that "network neutrality" regulation be applied to ISPs such as myself (even though we have not ever censored legal content), but at the same time has advocated that it not be subject to such regulations, even though it serves as a gatekeeper itself. It has even had this change written into the rules proposed in the NPRM released by the FCC on Thursday.

I will leave inferences from these events to the reader.

--Brett Glass