mediations

Review: The Filter Bubble, by Eli Pariser

The thesis at the heart of The Filter Bubble is simple: We are what we search for.

It seems obvious that the more able a search engine is to deliver results that most closely match our own personal needs the better it must be. Speed and relevance are the two factors that the searcher will most appreciate, but these attributes come at a price: they are enhanced because the search engine is using a whole range of dataset and filters drawn from our previous activity to customise and limit results for speedier delivery.

The downside is that as we express ourselves in terms of Facebook 'likes' or, more subtly, reveal ourselves through Google searches, we are presenting data gold dust to marketeers. 

The Filter Bubble goes beyond privacy concerns arising from internet click trails, and the enormous amount of data routinely being collected and traded by online businesses, to explore more sophisticated scenarios which suggests our very view of the world is being manipulate by those who are trying to sell products and services. From this rich mosaic of potentially highly personal information it is possible to perform what Facebook adviser Dean Eckles calls persuasion profiling; someone who likes this and this, and is friends with that sort of person, is likely to be more susceptible to that framing of a selling message.

Pariser takes these ideas still further. As search becomes yet more identified with the individual it is, for example, possible to see a personalised search based on, say, perceived IQ. In a sentence that might appeal to Edward Bernays, Pariser suggests: "In the wrong hands persuasion profiling gives companies the ability to circumvent your rational decision-making, tap into your psychology, and draw out your compulsions."

At the same time, personalised search can also narrowly focus the breadth of our inquiry and restrict the range of new ideas to which we are exposed. The implications for democracy and pluralism are worrying, with Pariser going so far as to claim: "We are now on the verge of of self-fulfilling identities in which the Internet's distorted picture of us becomes who we really are."

So at the heart of The Filter Bubble there is an interesting paradox. The very urges that that make us want to be promote our individuality though social networks at the same time locate us into ever more tightly drawn market demographics.

As Pariser observes, the conditions that promote 'personalised search' wouldn't be as appealing if it "didn't play to our post-material desire to maximise self-expression."

The more we try to create an individual persona, the more we become integrated into a marketing niche, that is designed to constrain and direct choice: "Algorithmic induction can lead to a level of information determinism in which our past clickstreams entirely determine our future."

The Filter Bubble is a highly readable book which, although stretching some of its main concerns almost to breaking point, raises some very important issues for those concerned with online communication and ethics.

If you are the sort of person who might tap Shirky, Gillmor or Zittrain into Google, don't be surprised to find personalised searching pointing you to Eli Pariser...

• The Filter Bubble, by Eli Pariser, 294pp, Penguin,  £12.99 (Review copy)

First Thoughts on The Filter Bubble, by Eli Pariser

If you are reading this, you are probably a lot like me - interested in (online) PR and cycling. Maybe you follow me on Twitter or use RSS, we might be Facebook friends, or perhaps you just Googled for something like coffee filters or bubble bath.

Whatever the case, it is still a fair bet you you are more likely to be interested in cycling than most other people.

This is partly because we gravitate to people who broadly share our world view, and partly because of the way that Google works.

Most of us would naturally assume that if they search on Google they will get the same results as everyone else.

Not true. December 2009 saw the launch of Google Search which used 57 signals to guess who you are and what you'd like, then tailors results accordingly.

For Eli Pariser, this ushered in "the era of personalisation." Increasingly, we find out about the world through search engines, and increasingly, the view we see is shaped by what computer algorithms think we want to see.

Fears that the internet narrows as well as widens choice are not new, but what Pariser does to great effect is lay bare the technological infrastructure that brings this about. 

And it is all rather scary.

I like buying books, and buy too many of them. One of the reasons I buy too many is that I am quite succumb to buying titles that Amazon recommends based on my previous purchases and the buying habits of people with similar purchases. This is rather good in that I have discovered dozens of books I may well not have read without this function; it is bad in that I am voluntarily surrendering some of my freedom of exploration to a marketing niche occupied by middleaged men with a certain education and view of the world, with a guessable disposable income etc.

I also read the Guardian, not least because it shares my political orientation. But by doing so, I necessarily exclude a lot of framings that don't coincide with my world view. I get around this by looking at other news portals, too, but by personalising my search Google narrows my viewpoint. I still buy a physical newspaper so I flick through the pages and see and read stories that are not directly linked to my interests; it is quite hard to do this online, especially if I am reading news on a my phone.

Scanning stories that I wouldn't normally read helps me break out of the filter bubble that prediction engines have created around me.

The filter bubble is Pariser's term for the unseen lens through which personalised search distorts my online exploration. He claims I am completely alone in my invisible bubble, in that it only surrounds me in a unique way, and there is no way in which I can burst out.

The filter bubble does other things, beyong reducing my opportunity for the chance encounters that bring insight and learning: "Search for 'depression' on Dictionary.com and the site installs 233 tracking cookies so other web sites can target you with anti-depressants."

This information is very valuable. "You behaviour is now a commodity, a tiny piece of market that provides a platform for the personalisation of the whole internet."

Companies like Axciom and Blue Kai track such information in incredible detail. Whereas Google Search used 57 signals, the Axciom data factory apparently uses 1,500 and has a detailed profile of 96 percent of American households, information it interprets for credit card companies and brands from Microsoft to Blockbuster. 

The filter bubble has many far-reaching effects, for media literacy and democracy, for individual privacy and business ethics, and for understanding the changing nature of interpersonal and communication.

Although I am not sure The Filter Bubble has the best of titles but I am sure that it puts Eli Pariser up with Clay Shirky and Jonathan Zittrain as must-read authors in this field.

Full review to follow... (which Axciom no doubt already knew...).

 

 

 

A chaotic challenge to Grunig

Marking does odd things to you, so it is no surprise that Richard Bailey is reacting to reading another slew of essays that reflect the view that Grunigian Excellence theory is the only way of understanding public relations.

In a typically thoughtful piece he suggests that Chaos Theory might help those of us who want to look beyond Grunig but are, quite frankly, floundering in our attempts to construct an alternative. (Actually, that is part of the problem - Grunig isn't wrong, just insufficient...).

I only have the vaguest understanding of chaos theory but I suspect that it's value for PR is in helping explain the implications of fragmentation. Grunig is essentially linear - communication is between an organisation and its publics, and the framing of publics coalescing around issues is a helpful construction for aggregating (stakeholder) sentiment.

In Online Public Relations David Phillips and I suggest that social media has helped shift the vector of communication through 90 degress, putting yet more importance on the conversation surrounding an organisation. It was ever thus, but our claim is that social media gives shape and substance to a myriad of comments, reflections and opinions, bringing them together into something akin to reputation.

I think we are right... but as I was quite possibly the worst candidate ever to attempt A-level Physics, I might struggle to marshall chaos theory to this worthy cause.