Personalization should be a choice!

I was wrong. There, I admitted it. I was wrong about the threats posed by Google Social Search. I was wrong about the threat posed by personalization. I was wrong to say that motivated reasoning was a pressing threat when there is no solid evidence to support such a claim:

A recent study by Matthew Gentzkow and Jesse M. Shapiro of the University of Chicago Booth School of Business [finds] “no evidence that the Internet is becoming more segregated over time” or leading to increased polarization…Instead, their findings show that the Net has encouraged more ideological integration and is actually driving us to experience new, unanticipated viewpoints.

I was wrong, but I wasn’t alone. Many observers have lamented the rise of personalization, suggesting, for example, that it’s a threat to democracy. Others have praised personalization as a natural evolution that promises to make the Web more useful.

We we’re all wrong because we all missed the larger point. Whether personalization is good or bad isn’t the real issue here. The issue is that personalization, as it exists today, is being provided without user consent.

Universality, openness, and transparency are, in my view, principles that underpin the Web that I know and love. Personalization without user consent runs counter to these principles. It’s isn’t universal because it shows different things to different people. It isn’t open because it’s not a choice. And it isn’t transparent because it’s not obvious that personalization is taking place.

I get it though. I get it that all Google and Facebook and Yahoo! are trying to do is to provide a better service. I also understand that they have a right to provide whatever service they want. I’m not accusing them of anything; their users have chosen to use their service and, and some point, accepted their terms and conditions.

I also understand that the majority of users would opt for personalization if given the choice and that asking for permission might consequently seem like a mute point. But this is where my grief lies, because not asking shouldn’t be an option. There are consumer protection laws against opt-out schemes and fine print – neither of which provides a perfect parable but that does go some ways toward explaining why this isn’t how things are suppose to get done.

Search engines started out as relatively straight forward services: you typed in your query and the algorithm fed back X because it was deemed more relevant than Y – a conclusion reached by analyzing information about X and Y. It wasn’t until later that information about the user was factored into this analysis and it has, since then, remained below the radar of most users.

To be fair, Google et al doesn’t hide the fact that personal data is being made use of. They’re more than happy to tell anyone willing to listed that “Web History uses [personal information] to give you a more personalized search experience”. But what use are such statements when they don’t stand in contrast to something? How different is non-personalized search? Maybe I won’t like it. Maybe I will. Maybe I’ll prefer personalization all day everyday. I just don’t know. Chances are I’m not aware that my search is personalized. And even if I was, I wouldn’t know what difference it actually made.

I can guarantee that the vast majority of Internet users have as a mental image the pre-personalization version of search. They don’t think of it as “personalized” simply because they’ve never been given a good reason to think of it that way. In my last post, I argued against Eli Pariser. But here, I must concede that he’s doing us a favor by raising the issue, albeit for the wrong reasons.

Personalization isn’t a threat to democracy. Not yet, anyway. The real issue is that the services we’ve grown accustomed to has changed without our knowledge. Google has even said their service now includes some form of political filter which, I guess, should make Pariser happy. But not me. As I made clear in my last post, I find politizied search immensly ill adviced.

Again, Google et al have all the right in the world to make these types of changes. It’s their service and they can do with it what they like. But that doens’t automatically make it right and it definitely doesn’t mean we, as users, have to sit idly by and say nothing.

Whether you think personalization is problematic or not, I hope you agree that personalization should be a choice. I also hope you agree that usage of such data should require explicit users consent; that it should be based on an opt-in rather than opt-out scheme. Lastly, I hope you agree that we, as users, have a right to know what and when our personal information is being used.

Do you? Do you agree personalization should be a choice?

This has been an unscheduled rant. I’ll get back to the promised discussion on the merits of personalization next time. I should, by that time, also be able to present some ideas to get us back on track in terms of universaility, openness, and transparency.