[governance] 'search neutrality' to go with net neutrality

Ian Peter ian.peter at ianpeter.com
Mon Jan 4 19:00:35 EST 2010


One thing to realise in this debate is that there is an inherent potential
conflict of interest involved when a search provider is also an advertising
and content provider. Particularly as internet names become more irrelevant
and search continues to expand as the major discovery mechanism.

And when the worlds biggest search provider is also the biggest content
owner and biggest advertising revenue source on line, there is a recipe for
problems and potential monopolistic behaviour.

And here's another bit for the puzzle! I have just been the victim of an
internet fraud - I realised in time (I hope!) that I was the subject of a
scam, but to escape it I had to cancel a credit card.

Where Google comes in here is that the search result that led me to the
fraudulent site was a number one hit on Google (I was looking for a
particular piece of Mac software and was drawn to a file sharing site that
wanted a small fee - but as I found out later has the habit of using credit
card numbers obtained from many similar sites for all sorts of other
charges). Quite sophisticated sites.

I cant blame a search algorithm for directing me to a site which happens to
be fraudulent. But I am not going to argue for entirely neutral algorithms
either - in time and as we become more mature as regards cybercrime, search
algorithms should be rejecting fraudulent sites where possible (yes this is
difficult I know).

So we probably don't want neutral search entirely. The Halal search engine
discussion here a few months ago raised similar issues. But what I do want
is clear disclosure and some clear overall policies regarding search
behaviour.

This is a critical issue, and especially for governance. It isnt going away
and there is no logical home for addressing these issues holistically at
present.

Ian Peter
 



> From: Lee W McKnight <lmcknigh at syr.edu>
> Reply-To: <governance at lists.cpsr.org>, Lee W McKnight <lmcknigh at syr.edu>
> Date: Sun, 3 Jan 2010 22:19:24 -0500
> To: "governance at lists.cpsr.org" <governance at lists.cpsr.org>, Lee W McKnight
> <lmcknigh at syr.edu>, McTim <dogwallah at gmail.com>, Thomas Lowenhaupt
> <toml at communisphere.com>
> Subject: RE: [governance] 'search neutrality' to go with net neutrality
> 
> And to be a bit more blunt and to keep it real simple: Google is the largest
> Internet operator in the US now, and largest source of Internet traffic
> worldwide.
> 
> Traditional 'telcos' like AT & T &  Verizon are still in top 10, but just
> barely.
> 
> But never mind the real world, we'll ignore Google's network and focus lots of
> effort developing net neutrality regs focused on....someone else. Brilliant.
> 
> Lee
> 
> Source: see Arbor Networks study Oct. 2009; or readwriteweb article excerpted
> below.
> 
> Google Accounts for 6% of All Internet Traffic
> Written by Sarah Perez / October 13, 2009 6:38 AM / 8 Comments
> « Prior Post Next Post »
> 
> Five years ago, Internet traffic was, for the most part, managed by tier 1
> providers like AT&T, Verizon, Level 3 Communications and Global Crossing, all
> of which connected to thousands of tier 2 networks and regional providers.
> Today, that has changed. Now, instead of traffic being distributed among tens
> of thousands of networks, only 150 networks control some 50% of all online
> traffic. Among these new Internet superpowers, it's no surprise to find Google
> listed. In fact, the search giant accounts for the largest source (6%) of all
> Internet traffic worldwide.
> 
> This data comes from a new report put out by Arbor Networks, who has just
> completed a two-year study of 256 exabytes of Internet traffic data, the
> largest study of global traffic since the start of the commercial Internet in
> the mid-1990's.
> "Hyper Giants" Rule Today's Internet
> 
> The biggest trend to come out of Arbor Networks' report is clearly that of the
> Internet's consolidation. Today's Internet is "flatter" and "more densely
> connected" than ever before, reveals Arbor Networks' Chief Research Officer
> Danny McPherson. Not only is Google the largest traffic source, there are only
> 30 large companies in addition to Google and including sites like Facebook,
> Microsoft and YouTube which now account for a disproportionate 30% of all
> Internet traffic.
> 
> According to Craig Labovitz, chief scientist at Arbor Networks, this shift
> represents the Internet's move into a second phase where it's no longer "all
> about contacting websites." Rather, "over the past two years larger
> organizations have been buying up the smaller websites and by July 2009, 30
> per cent of the internet was owned by a few large sites." The acquisitions,
> the result of billions of dollars spent by large companies snapping up smaller
> ones, has created a new Internet core of "hyper giants," a coin termed by the
> report.
> 
> The other companies making the list of Internet giants include names like
> Akamai, Limelight, BitGravity, Highwinds, and Gravity - hardly household
> names, and certainly not big telco providers. Instead, these content delivery
> networks (CDNs), are the new Internet backbone that help move large amounts of
> data across the web.
> 
> 
> ________________________________________
> From: Lee W McKnight [lmcknigh at syr.edu]
> Sent: Sunday, January 03, 2010 9:53 PM
> To: governance at lists.cpsr.org; McTim; Thomas Lowenhaupt
> Subject: RE: [governance] 'search neutrality' to go with net neutrality
> 
> McTim, everyone,
> 
> 1st a nit: Google's patents are a matter of public record, like everyone's
> (when issued).
> 
> Of course really people including Parminder and Thomas are more
> curious/concerned about Google's latest 'secret sauce' tweaked algorithms
> which are trade secrets and McTim is right they have no obligation nor
> motivation to share secrets. As has also been noted, the whole idea is not to
> be neutral but to bias search  in one way or another ie, either for financial
> reasons of Google or for usability  of their customers. Fine. They're a
> business, we would assume they are trying to make money from their customers.
> So forget search neutrality, ain't happening.
> 
> Next, and main point: as Ralf pointed out, 'networks' exist at all levels of
> the stack and of human society (and amongst consenting devices).
> 
> So 'net neutrality' is....amongst ISPs and their customers. But wait, not
> including mobile ISPs. Or them too? But not amongst IAPs (application
> providers) like Google right. So if you're operating in the cloud providing
> applications as a service across the Internet...what are you exactly from a
> neutral net point of view?
> 
> Anyway, I don't mean to confuse folks, except to point out that if search
> neutrality is an oxymoron....well draw your own conclusions re net neutrality.
> IMHO.
> 
> Lee
> ________________________________________
> From: McTim [dogwallah at gmail.com]
> Sent: Sunday, January 03, 2010 5:26 PM
> To: Thomas Lowenhaupt
> Cc: governance at lists.cpsr.org
> Subject: Re: [governance] 'search neutrality' to go with net neutrality
> 
> On Sun, Jan 3, 2010 at 1:47 AM, Thomas Lowenhaupt
> <toml at communisphere.com> wrote:
>> 
>> McTim,
> 
>  <snip>
> 
>> 
>> Here¹s my concern and the problem: How are the neighbors to know what's not
>> there?
>> 
>> 
>> 
>> If Google hand-wires the results to suite its needs, overriding its secret
>> proprietary sauce, how are the atomized neighbors to know there are others
>> opposed to the development? How are they to know to go to Bing? I suspect
>> many would conclude, ³Gee, no one else cares. Maybe I¹m wrong?² And Google
>> wins.
> 
> Why would Google "hand-wire" their results in a case like that?  They
> make money showing ads next to search results.  If such an
> intervention became public, the resulting negative publicity would be
> not worth whatever gain they would get from NOT listing Google haters
> in search results.
> 
> Try googling "google haters", there are millions of results.
> 
> If you were in a struggle vs Google re: a local real estate deal,
> would you really use Google Search to find allies??
> 
> Whats missing in all this is this (easily found using Google "how does
> pagerank work" I might add):
> 
> PR(A) = (1-d) + d(PR(t1)/C(t1) + ... + PR(tn)/C(tn))
> 
> While this may not be the current algorithm, it was the early version
> of it.  In other words, with a few seconds effort, one can easily find
> out the information asked for in this thread.
> 
>> 
>> 
>> 
>> Of course education is the answer. And we've plans to develop curriculum that
>> begins in the 3rd grade and to educates the public at all levels as to civic
>> ills that might arise by putting too much trust in one search engine. But
>> this is likely to take a decade or so to permeate society.
> 
> 
> key phrase there is "might arise".  Its brand loyalty, that's all.   I
> would hope that we as a group have far bigger fish to fry than this.
> 
> 
>> 
>> 
>> 
>> (Before I move to my second point of disagreement let me slip in another
>> example. Imagine we¹re a few years down the road and Google ³winner$² begin
>> running for public office. How are we to trust its opaque search algorithm
>> during the rough and tumble of an election campaign?
> 
> same as during a non election year.  If it is useful to you, use it.
> if not, then find something that is useful.
> 
>> Then we¹ll clearly see the relationship between link and ballot voting!
> 
> I cant parse this.
> 
>> And even if Google didn¹t hand-wire, opponents would surely charge that it
>> did, poisoning the system.)
> 
> 
> What you are doing is putting any search engine into a no win system.
> 
> I, for one, appreciate what google has done around US election information.
> 
>> 
>> 
>> 
>> Second, when you say,
>> 
>> 
>> 
>> "We, as IGC (or even CS as a whole) can't expect to seriously ask Google to
>> show us their patented IP, can we?"
>> 
>> 
>> 
>> I disagree. Given the importance of search in the development of civic
>> attitudes - like the newspapers and TV of old - I think it¹s vital that we
>> address the issue.
>> 
>> 
>> 
>> Here¹s a path. Initially we make the importance of ³search transparency²
>> known to Google and encourage them to provide their secret sauce¹s recipe. (I
>> prefer ³search transparency² to ³search neutrality² as it is a somewhat
>> easier to devise a metric.)
>> 
>> 
> 
> I think they already know that people want to know how they know what they
> know.
> 
> 
>> 
>> Google has capitalized their search lead and integrated it into a plethora of
>> other services and should be able to keep their lead for the foreseeable
>> future, and might be prescient to see the poisoning possibility and be
>> agreeable to the need to move toward transparency.  Perhaps they might
>> initially agree to a trusted outsider initially, a Moody¹s-like entity to
>> judge all search engines. And if Google doesn¹t see the light, perhaps Bing
>> might take a lead in offering transparent search.
> 
> 
> perhaps, but i wouldn't bet on it.
>> 
>> 
>> 
>> And if the search industry doesn¹t see the necessity ­ no one steps in - it¹s
>> incumbent on civil society to educate the public and decision makers about
>> the impact of search opacity and encourage the development of a transparent
>> search engine.
>> 
> 
> That has been tried, http://en.wikipedia.org/wiki/Wikia_Search
> 
> You are welcome to do it again if you feel the need.
> 
> --
> Cheers,
> 
> McTim
> "A name indicates what we seek. An address indicates where it is. A
> route indicates how we get there."  Jon Postel
> ____________________________________________________________
> You received this message as a subscriber on the list:
>      governance at lists.cpsr.org
> To be removed from the list, send any message to:
>      governance-unsubscribe at lists.cpsr.org
> 
> For all list information and functions, see:
>      http://lists.cpsr.org/lists/info/governance
> ____________________________________________________________
> You received this message as a subscriber on the list:
>      governance at lists.cpsr.org
> To be removed from the list, send any message to:
>      governance-unsubscribe at lists.cpsr.org
> 
> For all list information and functions, see:
>      http://lists.cpsr.org/lists/info/governance
> ____________________________________________________________
> You received this message as a subscriber on the list:
>      governance at lists.cpsr.org
> To be removed from the list, send any message to:
>      governance-unsubscribe at lists.cpsr.org
> 
> For all list information and functions, see:
>      http://lists.cpsr.org/lists/info/governance


____________________________________________________________
You received this message as a subscriber on the list:
     governance at lists.cpsr.org
To be removed from the list, send any message to:
     governance-unsubscribe at lists.cpsr.org

For all list information and functions, see:
     http://lists.cpsr.org/lists/info/governance



More information about the Governance mailing list