A conversation with the Director of the Information Innovation Policy Research Center at the National University of Singapore

Question: What problems does so-called "perfect digital memory" pose?

Viktor Mayer-Schönberger: It seems to me there are two kinds of problems that comprehensive digital memory poses to us.  The first one, I’d like to label "power problems," or power challenges.  It usually has to do with the fact that when we share information with somebody else, that person has some informational power over us.  That person can use in order to sell us some stuff, to market us some stuff, and that information can be aggregated, can be shared with others so there’s a certain informational profile can arise about us. 

This is what has been the focal point of informational privacy research for a very long period of time.  But I believe this kind of diatic relationship, if you want, between a person, a data subject that gives information and a process that receives information, is only part of the picture.  We need to go beyond this relatively small-scale view, particularly when we look at the importance of memory. 

Let me give you an example.  For a very long period of time, we thought that sharing stuff on the Internet is somewhat dangerous because it’s not just our friends that have access to it, but it’s everybody around the world that has access to it.  Hundreds of millions of people in thousands of jurisdictions, and somebody somewhere might actually be offended by what we have to say and might get a court to agree and then we find ourselves having to defend our actions.  Much like the Google managers had to do a few weeks back when they were successfully sued and convicted in an Italian court for what they did or didn’t do on YouTube. 

And so that’s been around... that view has been around for a couple of years.  It’s called the panopticon.  The panopticon basically is an idea that has been with us for a hundred and some years, invented by Jeremy Bentham in Britain, the idea is it would be great to create a prison in which the prisoners don’t know when the prison guards watched them.  They have to assume that they’re always watched and therefore they have to behave all the time.  Similarly, because we don’t know who is watching us online, we have to behave all the time to the lowest common denominator and therefore our behavior is skewed as constrained.  And many people have been writing about that. 

But I believe there is with digital memory something much more problematic coming up and that has to do with the longevity of digital memory.  So, we have to face the fact that what we say and do online today will not only be viewed by the hundreds of millions of people that are online today, but might be viewed and interpreted differently by people and institutions 10 years, 20 years, 30 years down the road when we are no longer young and we might not be as engaged in public discourse and protests any more, but we might want to apply for a well-paid investment banker’s job and then they might just Google us and find out that 15 years earlier we said something that wasn’t so complimentary to the banks, or so.  So, therefore, comprehensive digital memory creates what I call a temporal panopticon.  It creates a situation in which we have to fear that we are not only watched today, but we are watched by future generations, by our future.  And that really dramatically constrains what we ought to do online and pushes us toward self-censorship, exactly where we don’t need self-censorship because we need robust public debate online, that’s what online discourse is all about.

Question: What kinds of new problems could arise as our lives become archived?

Viktor Mayer-Schönberger: Well let me give you an example of what happens already today.  So, there are people who lose their jobs who have relationships blow up basically because of what they said online earlier on.  But there are also cases that are quite shocking in a way.  I had a woman call in at one of the radio shows that I did and she told me about her case.  She had been convicted of a crime in the State of California many years ago, when she was 17 years.  Served in prison and then was released.  Started a new life, went to a place, found a husband, had kids, started a family.  Basically everything was in order.  She even found God, and so hey, this sounds like a happy end story.  Except, one day a colleague of her son in school Googled her and found a website of mug shots of all prison inmates in California over the last 20 or 25 years.  And she was right in there.  And suddenly the community, her small community that she lived in knew that she actually was an ex-convict and immediately ostracized her.  And her new life unraveled.  And she called in and said what can I do?  And I said nothing.  This is the unfortunate side effect of a digital memory that doesn’t forget while our society forgets.  We forgive even ex-convicts.  But the digital memory doesn’t do that anymore.

Question: How might we set an expiration date on information?

Viktor Mayer-Schönberger: I advocate a revival of forgetting.  And there’s a number of ways to do that.  One is expiration dates.  Expiration dates would be very simple to implement.  It’s just another form of meta-data, much like you have file names, date of creation, date of modification, exact location of a file on your hard disk and so forth.  And expiration dates would just add another category of meta-data to the file system.  We would be able to select the expiration date in any form or shape we want, be able to change it after the fact, of course, but once the expiration date has been hit, the file would be deleted by the system. 

The importance is that by entering or having to enter an expiration whenever we store something, we are reminded – we are reminded of the importance that information is not timeless, but it is connected to a particular context in time and loses it’s value over time.  Most information does and so by setting expiration date, we really link time with information, something that biologically we cannot do. 

If I may, I’d like to interject something here and that is, we started off and I said there’s two kinds of dimensions, the power dimension, and did I mention that there is a second dimension that gets overlooked quite frequently, which I call the time dimension.  And that has to do with the fact that we humans are biologically programmed to forget.  We forget most of what we experience every day.  That’s a way by which we can abstract and generalize and evolve and grow and rid ourselves of stuff that is no longer relevant to us. 

What is interesting is that if we can’t do that, then we become burdened by the details of our past to the point that it makes us indecisive and it shapes the way we decide.  We know a little bit about that because there’s a small number of people who cannot forget.  They have a biological difficulty of forgetting.  So if you ask them about a day 30 years ago, they can tell you when they got up, who called, what was on television in the morning, what they had for breakfast, and so forth for every single day in the last 30 years.  It would be great, I thought because they would never forget where they parked their car on the mall parking lot, but the problem is, they hate that.  Many of those people who have difficulties forgetting hate the ability to not get rid of the old.  They remember all of the mistaken decisions of their past all the time and that troubles them a great deal and it inhibits their ability to decide and act in the present and to think in the future. 

And so comprehensive digital memory might actually create that for us.  It might give us a sense of not forgetting anymore and thereby preventing us humans from generalizing, abstracting, evolving, growing, and also accepting others to change over time, to evolve and to grow.  And without forgetting, we don’t have an ability to forgive.  [00:12:30.05] 

Question: Who would decide what information to set a date on?

Viktor Mayer-Schönberger: I believe that the expiration date is a way by which we humans, we individuals, have a chance to reflect and to choose.  So, it’s us that decide.  It’s not government; it’s not the processors, the online servicers that decide how long they want to keep information.  It’s us.  And whenever we share information, we also add an expiration date to it.  And then the other side can choose whether to accept the information with the expiration date or not; whether to actually cut the deal and transact or not.  In fact, many vendors, particularly online vendors who have a very close relationship with their customers, as consumers actually would probably prefer, or enjoy expiration dates. 

Think of Amazon.com.  Amazon now has a lot of transactional information about my books and other purchases that I did in the past.  But what is it good for Amazon to know what I shopped for nine months ago, or 12 months ago, or 15 months ago if I’m no longer interested in what I shopped then.  It would be great for Amazon to know, not what my preferences were 15 months ago, but what my preferences are today.  And with the expiration date, we help the vendors as well to limit the amount of data that they have, that is irrelevant and to focus more on the still relevant information, the still relevant preferences and values that they can then use to make recommendations to us.
 
Question: How can we keep from destroying information that might be relevant later?

Viktor Mayer-Schönberger: You know, the bottom line is that for all of human history, we forgot most of what we experienced and we remembered only those things that we thought were really important.  Sometimes we were right and we remembered the right things.  Sometimes we were wrong and we remembered idiotic things.  The importance is that the remembering was the exception and forgetting was the default, was the rule.  And today, this has become reversed.  Remembering today is the default and all of our digital tools and artifacts we use and forgetting, deleting, is costly and time consuming. 

The fundamental problem is that I want to right this shift again and to bring back forgetting into our society.  But I also appreciate and value the fact that for certain kinds of information, we need to be protective.  We need to protect the information, keep it recorded and archive it; public information, governmental information, court records, information that the media publishes.  These are all incredibly important sources of societal memory and societal history that we ought to preserve.  But these are the exceptions.  The rule still should be for most of that we still can forget and ought to.

Question: Do you think we’ll see increased storing of sensitive information off line?

Viktor Mayer-Schönberger: I don’t know.  I don’t know.  My sense is that at this point in time, if you have a present convenience and a danger that is far into the future, a lot of people opt for the present convenience.  And so I don’t think that a lot of people today would choose to be careful.  Except, of course if they’ve been burned.  And many, many people are already suffering from what they said on Facebook, or what they Tweeted.  And more will suffer.  The fact that Twitter Tweets are now archived and recorded in the Library of Congress brings that point to the forefront.  The fact is that Twitter for a very long period of time has let other companies look at the Twitter feeds.  It’s not just the Library of Congress that has it.  It’s many other commercial companies that have Twitter feeds and are actually doing stuff with it.  And that might come to haunt us.  The more people that are getting affected by it, the more will change the behavior. 

I am troubled by the fact that many of those people will change their behavior toward self-censorship and I am troubled by it because that’s not what we need in society and that is denying the web tools that we have today, the value that is inherent in them in sharing knowledge and experiences and so forth.

Question: How have advances in information technology transformed government?

Viktor Mayer-Schönberger: For a very long period of time, starting in the 1960s and then into the 1970s, computers were seen as a way information technology in general was seen as a way of making government bureaucracy more efficient and effective.  So, the drive was always productivity, efficiency, doing more with less.  Maybe in the 1990s or so, we had another dimension come into this picture and that is to be more consumer friendly, or citizen friendly in the services that the government provides.  But these are just two very specific ways of looking at how government interacts with society because it looks at the citizens as consumers as transactional partners with the government.  They go for efficiency to lower the cost, or you go for consumer friendliness or transactional partner friendliness in order to provide some service quality. 

I believe that electronic government information or technology in government must go much further than that.  We as a society have a right to know better what the government is doing, to engage with the government and to have a government in place that is willing and able to use the technological tools available to engage us citizens.  That’s currently not happening or not happening at a sufficiently high level.  We are still in this old fashioned mode of thinking about transaction efficiency and user friendliness.

Question: Is the U.S. government lagging in information technology?

Viktor Mayer-Schönberger: The U.S. government lags behind in electronic government quite a bit large because whenever you need to create an electronic government transaction, software, you need to involve a number of different government stakeholders.  And they just despise each other usually, and they hate when the other side has access to their information and data.  Data silos and information silos in government are one way of defining power and influence.  And so obliterating those silos really reduces the power of individual departments and agencies and that’s why they are not keen on doing that. 

And let me give you what I thought a wonderful example is, if I may.  There was a Web site on moving called “Moving” and it helped people who were moving house to do the change of address form, to move the electricity and the gas and the utilities over the telephone, to also hire perhaps a moving van and so forth.  It was a one-stop shop for all of this.  And it was a public-private partnership that made this all possible.  And it was a huge success.  People really loved this website. 

Government agencies realized that and they thought, gosh, this means there’s real traffic to be generated, and immediately what they did was to leave that umbrella, that one stop shop and created their own little shops in order to drive traffic to their own little shops.  Of course, the sun was really more – the sun was more than just the parts put together and once you take these parts apart and everybody has this little shop the consumers were really frustrated again and don’t want to go to 15 or 16 different websites in order to change their addresses.  The problem therefore is that when you breakdown the silos, you create value, when you resurrect the silos, you destroy value.  That’s something that government agencies theoretically understand, but have practical difficulties realizing and putting in place. 

But I am really more interested in another aspect.  I am really interested in the way by which government in the United States has now decided to make public a lot of the information that it collects.  My colleague and friend, Beth Novak, and others have really pushed this very hard in the Obama Administration and I think that’s exactly right.  What we need is more information that the government collects, not personal information, but general information to be shared with the public at large, with NGO’s, with the society at large so that they can then look at that data and really see what is happening in our society.  We now have the software tools to do that, we now have the interest in the public to do that, that’s why we need to move ahead and make more and more of that government information freely available.

Question: Does the Obama Administration approach information technology differently than its predecessors?

Viktor Mayer-Schönberger: Oh yes.  There’s a huge difference.  If you look at the website that the Obama Administration set up to be able to trace the money of the stimulus package going into the economy and you can see to which communities, which counties, to what companies and institutions, organizations, the money goes.  How much as been dispensed, what was the impact on employment and so forth.  This data is fabulously well designed and presented through a geographic interface and it is simply superb; very, very different from the previous administration.  In fact, the outgoing Clinton-Gore Administration put in place, for example, a website with the Environmental Protection Agency where the Environmental Protection Agency made accessible a self-reporting inventory of toxic waste and linked it to a geographic information system so that people could actually look at their neighborhood and see whether there were any toxic waste repositories in their neighborhood. 

When that went online, a lot of people started pressuring the companies in their neighborhood that had toxic waste to clean up and to get their act together because land value would adjust.  And so there was almost a wonderful market for this as well as a democratic force behind it.  Of course, the George W. Bush Administration, after 911, took the website down, ostensibly because of a security risk.  Terrorists could access the toxic waste **** and find out where the toxic waste was I this country.  I not so sure the security risk was really was that high.  I think a lot of people in the – I suspect a lot of people in the Bush Administration connected to some of the toxic wasters were quite happy to put that database offline.

Question: How might the eGovernment revolution affect international intelligence?

Viktor Mayer-Schönberger: Well, first of all, if we permit people to have access to government information, we don’t necessarily become more vulnerable.  The cyber security task is mostly one focused on keeping the infrastructure up and running.  And we are very vulnerable there.  They’re very vulnerable not just on the internet information structure, but on the energy and electrical infrastructure as well.  And they’re vulnerable because for many years, even decades, we have increased the efficiency of the infrastructure, but limited the investment and the redundancy and robustness of the infrastructure.  And that makes all these infrastructures prime targets for hackers and terrorists around the world, whether they are organized by nation/state or whether they are non-combatant terrorists, Al Qaeda's ilk or anything else.  I think we need to spend more money on that.  I think we need to spend more money on securing the infrastructure, on knowing what the other side is doing.  The federal government does not have yet its act together on cyber-security.  That’s still a big problem, it will require all the major stakeholders to come together and it will require some legislative change as well and it will require a different mindset of the people. 

The missiles of tomorrow are not going to going to come through the sky; they’re coming through the fiber optic networks.
 
Question: Will the European Union emerge as a global superpower to rival the U.S. in the next century?

Viktor Mayer-Schönberger: I think the European Union is superbly positioned to become a major player in the 21st century.  And it has done quite well.  The common currency, the centralization, necessary centralization, coordination through European law, European courts have done a great deal of good to the continent that was divided so long.  I think Europe faces two fundamental challenges that are very hard to overcome.  The first challenge is the demographic challenge.  A graying continent.  Europe is graying much faster than the United States is.  And so, Europe has to deal with that problem.  That means that not only there are more people who want to retire and get a state pension, that means that every year you have about 2 million people that are missing in the workforce because they retire and you need to replenish them.  Two million.  That means 2 million immigrants from somewhere.  There isn’t a country close to Europe large enough to supply two million immigrants a year, let alone the question of cohesion of integrating these people in to the European culture.  So, Europe faces a huge challenge, much more so than the United States. 

The other challenge of course is, and that’s the beauty of the United States, the beauty of the United States is that it pushes itself to the brink, but it never falls off.  Once it’s at the brink, even extremely close, it then comes back.  And it retains its unique spirit of entrepreneurship and innovativeness.  Europe needs to begin to harness that and begin to believe in its own ability to be entrepreneurial, creative, and innovative and to believe in itself.  So far, it hasn’t done that, it’s still always eying towards the United States.  I think a little bit more self-confidence would do it good.

Recorded April 22, 2010
Interviewed by Austin Allen