mYfAVE'StOYOU: The Ethical Issues of Personalisation Online

sábado, 2 de julio de 2011

The Ethical Issues of Personalisation Online

It feels kind of negative to start a post with a warning, but you’re all busy people and I’d hate for you to waste your time – so here goes: Today I’ll be talking about the ethics of personalisation online, and the potential issues which might arise as a result. As such I won’t be offended if you decide this isn’t quite your cup of tea and decide not to read on.

Still with me?

Thanks for sticking around, you look lovely by the way...

So this post was prompted by a seminar I attended this week where I was lucky enough to see Eli Pariser present on precisely this topic. He’s very passionate about what he calls ‘the Filter Bubble’ (i.e. the effects of online personalisation) so much so he’s written a book on the subject. So, what’s the deal with personalisation? Well, as Cyrus highlighted in his recent Personalised Search post:

“Google’s personalised search means nearly every result returned within a browser is altered one way or another. It’s rare that two different people on Earth ever see the exact same set of search results.”

What’s wrong with that? Google personalise results so people see the things which are most relevant to them. Normally we’d see relevancy as a good thing... Personalisation is the future – improved user experience – hurrah!

But wait - Eric Schmit says:

“It will be very hard for people to watch or consume something that has not been tailored for them.”

That sounds a little sinister... It’s going to be hard for me to see stuff that hasn’t been tailored for me? Why? I thought your algorithms were being used to show the ‘best’ results...

Well, it turns out that they are, it’s just that ‘best’ is subjective – ‘best’ is based on your point of view.

“Increasingly online, it’s becoming impossible to escape your own point of view.” - Eli Pariser

Fortunately my point of view is 100% correct. I am never wrong. Ever. I really like being right because I get a natural high thanks to the dopamine which is released by my very clever brain. It’s awesome when you see your own views reflected right back at you via the magic of the interweb.

But what if I’m not right? What if I’m never again exposed to a contrary point of view?

And what happens when other sites get in on the act? Oh hai Facebook!

“A squirrel dying in front of your house may be more relevant to your interests right now than people dying in Africa.” – Mark Zuckerburg.

Uh oh – it’s beginning to feel like you’ve got me wrong... I want to hear about important stuff. I might not ‘like’ hearing about stories like that, but that doesn’t mean I don’t think that they’re important.

Pariser refers to this as ‘the psychological equivalent of obesity’. The personalisation filters which are currently being used may bias the content which you see – for example, if you mainly spend your time online seeking entertaining content which doesn’t seek to challenge you in any way (drunk party pictures on Facebook *like*) then that might soon be all that you see.

Now clearly, if people are going to head to Facebook for their daily dose of real-world news then mankind has got itself some massive problems. But if people are going to Google for news and are still only seeing a one-sided view of the world which just so happens to coincide with what they already believe is that really the ‘best’ result? On the face of it, it might seem to be a ‘good’ result from the user’s perspective – but is it for the greater good?

Here comes the ethics question – Is it ethical for search results to be personalised in this way?

In my view, it’s really not ok. Most people (i.e. those who work outside of search) don’t realise that their search results are personalised. Plus, as Cyrus quite rightly highlighted “...Google made sure it’s darned hard to turn off.” Much as it might comfort me to only view results which sit comfortably with my own politics, is it good for me? We’re handing over control to an algorithm that by design does not strive to provide a balanced mix of results.

This scares me.

But, is it reasonable to expect a corporate entity to act for the greater good? Particularly if providing users with a more balanced SERP results in them high-tailing it straight into the warm embrace of the competition? In any case – wasn’t it always this way? Before the internet people consumed news only via whichever media sat comfortably with their own political affiliation. Plus of course, even if a more balanced mix of results are shown, you can’t *make* people click through to read something they don’t want to.

So, what do you think? Should we be afraid of personalisation? Should we push for easier ways to turn it off? Should there be more ‘balanced’ results for certain types of queries? Should I get myself a tin foil hat, cancel my broadband, flush my smartphone and hide under my desk?

Over to you mozzers, I’d really like to hear your views – especially if they don’t coincide with my own.

Image credit: Lucky Lager - http://www.flickr.com/photos/laughingsquid/5281358464/sizes/m/in/photostream/

Do you like this post? Yes No