Peter Bihr

Archive
Tag "privacy"

The New York Times on the New Art of Flickr

The New Scientist touched on a fascinating concept: Pre-Social Networks that would foster serendipity by matching people based on their interests and their current (or even future) location.

Imagine heading to a café, and your phone recognizes where you are going and lets both you and someone at that café know that you have certain interests in common. It doesn’t matter if it’s a hobby, a former employer, a common friend, or even an ex girlfriend for that matter.

It’s the chance of massively increasing an element of serendipity that’s incredibly powerful.

There are huge potential implications here, with privacy only being a small chunk. We’re talking changing social dynamics, ways of meeting peers with less friction, less awkwardness. Mapping where people of certain interests hang out. Etc, etc, etc.

Thinking about it like that, maybe privacy isn’t just a small chunk of this after all.

And yet. All this by basically matching the data we already codified online: Facebook social graph, Twitter social graph, interests based on links we share, location by phone GPS and Foursquare checkins. Maybe throw a few extras for more richness, more flavor, like Last.fm music preferences and pull from Tripit which places you like to travel to, both being strong social connectors.

Not sure which of these data sets we’d actually want to match, and who we’d want to match them. It’s a strong, powerful notion, though, and one we’d better think about sooner than later.

Image by Thomas Hawk, some rights reserved (CC by-nc)

[permalink]

A few weeks ago, I tried out Gigalocal. It’s a platform that lets users announce jobs they would like to see done, and how much they’re willing to pay for it. (“Clean my apartment for 20 Euros”-style jobs.) I signed up so a journalist friend of mine could try out the process in a controlled environment, and without having to clean someone else’s apartment. Yet, there was a bit of a problem.

The minute I put up a test job offer for my friend (“I’d like a cold soft drink, now, delivered”), the service tweeted the job. Makes sense, I guess, as it makes it easier to track jobs out and about waiting for jobs. (If the users of said service have smartphones, that is.) But they didn’t keep it to the job description.

Gigalocal tweeted my full address, down to the house number.

That’s right. They didn’t restrict the location info to the neighborhood (close enough to figure out if that job’s a good fit for you), or next subway stop, or street level, or a 500m radius. No, they tweeted the full address.

I canceled my account and mentioned in the cancellation form that I find publicly tweeting addresses quite unacceptable, as I hadn’t been aware before that the company might do that.

Here’s the reply (translation below):

Grundsätzlich kann jeder User seine Daten selbst schützen. Niemand ist gezwungen seinen Wohnort anzugeben. Ein User kann einen Gig überall erstellen, seine aktuelle Position wird vom GPS Modul (Smartphone) oder durch die IP (Website) vorgegeben. Diese Ortung hat man schon wenn man auf Google Maps geht und dem Browser erlaubt den Standort zu erkennen. Jedem User steht es frei wo er seinen Gig erstellt, er kann ihn also gerne 4 Straßen weiter erstellen und sich dort mit dem Gig Erfüller treffen.

Translation: Generally, every user can protect their own data. Nobody is forced to input their home address. A user can submit a gig wherever they like, their current position is read through the GPS module (smart phone) or IP address (website). This triangulation takes place even if you just go to Google Maps and allow the browser to read your location.

Every user is free to set up their gig wherever they like, so they can set it up 4 streets down and meet up with the job fulfiller there.

I was shocked. Shocked at this level of ignorance in building a user service that requires granular privacy. (Home addresses!) Shocked at how the staff didn’t even seem to consider they might have made a grave mistake. And shocked at myself for being even surprised by the two aforementioned failures.

Gigalocal, here’s a hint: You want your “job fulfillers” to know if a job is close enough to make it worth the trip. You don’t want to show the world people’s home and office addresses. And you never, ever want to tweet personal information without asking permission first.

Now go back to the drawing board and don’t come back before you know what you’re doing.

[permalink]

Goopymart

Facebook has done it again: The company enabled a new feature that uses facial recognition to prompt your Facebook connections to “tag” you in photos they are shown. In other words: It recognizes user faces in photos, then shows them to their friends, encouraging them to identify the user by putting a name to the image.

Sounds useful? Yeah, right.

Consider this: A user does not get the option to pre-approve of photos of themselves being published. As you might know, I usually go for a share-all approach. This case, though, is another notch in Facebook’s bedpost of privacy violations. As so many times before, Facebook defaulted to share personal information instead of to protecting it. Again, they went for an opt-out model (where the user has to become actively involved to protect their privacy) and – to top things off – decided to hide the option to disable this “feature” way down in the privacy settings.

This is beyond bad style. Here’s how you disable facial recognition on Facebook.

I can’t wait for a truly privacy-conscious social networking service.

mystery cat / goopy mart / CC BY-NC-SA 2.0

[permalink]

Diaspora has launched its consumer-facing alpha (as opposed to the last release that was for developers and tinkerers only). The privacy-conscious social network was off to a bit of a rocky start since it was profiled (in the media, on the web) enthusiastically as The Facebook Killer – a level of expectation that led to huge crowd-funding on Kickstarter as well as completely overwhelming expectations no one could possible live up to.

Fast forward half a year to now. The dust has settled, the first release is out. The “alpha” isn’t in the name to look more cute, it actually is a very early release with likely a lot of bugs and certainly only very basic functionality.

However, it does seem to work, and after the first few pokes at the service it looks quite good to me. A few screenshots:

Diaspora

The blog, just because I kinda like the logo.

Diaspora

The Diaspora dashboard is clean and minimalistic. Works fine for me, but it’ll only really become clear how usable it is once more contacts are linked to my profile.

Diaspora

To handle privacy and granular sharing, Diaspora uses the metaphor of “aspects” of your identity. An aspect could be your friends, your family, your work life: you can choose granularly which of these groups sees what you post. In Diaspora’s own words:

Diaspora lets you create “aspects,” which are personal lists that let you group people according to the roles they play in your life. We think that aspects are a simple, straightforward, lightweight way to make it really clear who is receiving your posts and who you are receiving posts from. It isn’t perfect, but the best way to improve is to get it into your hands and listen closely to your response.

At a glance this makes a lot of sense. Again, time will tell if it holds up.

Diaspora

On your dashboard you can also always see with whom you shared what kind of information.

Diaspora

Status updates and photos can also easily shared with external services. So far (ironically) this is limited to Twitter and Facebook. You cross-post by simply ticking the “make public” box.

Diaspora

User profiles are very minimalistic as of yet – for example you can’t put in a link to an external website. The age indicator is one of the less charming ones – never before have I actually felt old using a social network ;)

Since Diaspora is positioned as a more responsible social network than Facebook, data export and deleting your account is a simple enough task:

Diaspora

It’ll take a little while to test it all in full, and to gather a bit of a crowd on Diaspora to check out all the interactions. But at a first glance, despite this being very clearly alpha ware, it looks very promising. Another half year, maybe, and this may be a F… no. I’m kidding. This has nothing to do with Facebook, or being a Facebook killer – but it really doesn’t have to. This looks great by itself.

[permalink]

The other day I visited Friedrich-Ebert-Stiftung (FES) for a day to give a keynote speech and a workshop for FES stipends at the summer academy. (Full disclosure: paid gig.) For completeness’ sake I’m posting the slides below. In order for them to make sense I’d recommend downloading the file from Slideshare so you can see the notes.

On a side note, I have to say I really enjoyed particularly the discussions with these students. We talked a lot about privacy on social networks and the implications of using these online services. I was surprised on more than one occasion: Not a lot of the participants use smartphones, which may be a budget thing given they’re all still studying. The crowd was much more critical of online social networking than I expected. (There was a strong split in the group, with those seeing chances rather than risks on one side and those highly critical of social networks on the other.)

Two things became very clear, though: (1) Just like German society overall this group had a significant part of online critics (with varying degrees of informed argumentation). (2) All of them are acutely – almost painfully – aware of the role of privacy and how it’s being affected by voluntary participation in online sharing behavior (social networking, Twitter etc), involuntary sharing (government involvement) and commercialization (all major actors are international corporations).

While I wished the overall discourse (on a societal level) about the complex issues of privacy/ownership/control of data online was based on a more informed basis, it’s very clear that we’ll be having this discussion for awhile to come. And that’s good: Keep thinking, discussing, debating. Just please make sure to stay away from panic and fear driven rhetoric as well as hyperbole. And if you happen to encounter such arguments, feel free to drop in some facts and see the fear go away.

[permalink]

Collaboratory ReportIt’s done! Over the course of the last few months I’ve been part of the German Google Collaboratory “Internet & Society”. The expert group is part of Google’s effort to reach out to multiple stakeholder groups and discuss with experts the challenges and opportunities our society faces. We discussed a whole range of topics from privacy to media literacy to democracy, and conducted a survey among the German web community about all this. We crunched the numbers and interpreted the input we got.

Today, the final report was published. Download the report here (PDF, 1.9MB, language: .de). (Update: find all reports here.)

Together with Henning Lesch of eco and I focused mainly on privacy. You can find the results and essence of the research under the headline “Datenschutz & Privatsphäre”.

What really struck me – in the expert group as well as in the survey and in the discussions we had with politicians – is how hard it is to come to any clear solutions in this field. It almost seems like we, as a society, need to have some in-depth discussion on which exact questions to even ask, and how to define all the buzz words. Do we need anonymity online? How much? Absolute or relative anonymity? What exactly does anonymity entail? It’s really not all that simple.

Although I will say that personally I strongly favor an internet that allows for anonymity over one where all action can be tracked. In fact, I believe that the opportunity to discuss open and free and without fear of repercussion (in other words: in anonymity) is an important basis of a strong and free democracy. (Luckily, many, many of the participants of our survey hold a similar view.)

Long story short: I tremendously enjoyed working with all these smart folks. And I hope this report can help foster a good discussion about where we’re headed and how we’d best get there.

As it is today, there is much to discuss, and not all of the folks in a position to make the relevant decisions seem to be up to speed. (And you can’t blame them, there’s just too much on everyone’s plate!) In the US, Code For America exists solely to support the decision makers in DC with external expertise in all questions related to the web. If there’s demand for something similar in Germany, sign me up. We as the web community shouldn’t stand back and complain about politicians, but instead offer a hand to work with them. There’s no other way forward.

Full disclosure: I’m a member of the Google Internet & Society Collaboratory and I worked with Google before.

[permalink]

So after seemingly endless debates, moving launch dates and massive protests by privacy groups and – worth noticing – Ilse Aigner, the minister of of Food, Agriculture and Consumer Protection, it looks like Google just pretty silently launched Streetview in Germany. Kind of.

Kind of because it’s not the real Streetview, but a slightly different take on the original idea, based not on Google’s own photo material but on user-generated photos. Let’s have a look.

But first, let me put it in perspective briefly. While sharing many privacy concerns regarding all kinds of online services, I find Streetview to be very much not a problem at all. All faces are pixelated, and it’s the public space. I’ll side with Jeff Jarvis who says:

I argue that what is public belongs to us, the public, and efforts to reduce what’s public steals from us. Journalists should be particularly protective of what is public; so should we all.

He goes on to state that once we start prohibiting private companies from taking pictures there’s a threat we’ll start doing the same with citizens, then journalists. You get the drift. Full ack. What’s public is public. We live in a state where politicians openly demand more video surveillance and even does the occasional biometric video surveillance pilot project – in my book, that’s much, much worse.

But back to the point. Below you see a screenshot of Streetview in New York. Please note the high-quality of the pictures, and perspective of the environment as seen from the street. You can “move” along the street. It’s useful. It’s not super pretty, but it does the job, which is help you get oriented.

Google Streetview NYC Screenshot

Then look at the German flavor of Streetview (screenshot below). It’s based on user-generated photos (by ways of Panoramio), which isn’t necessarily bad, but doesn’t quite seem to work here. It’s not consistent, it’s all different perspectives, and it’s not all current photo material. Some of the photos submitted are clearly from the 80s or early 90s. Interesting, sure. Useful, not really.

Google Streeview Germany Screenshot

What’s more, since the photos are user-generated and – I’m assuming – not automatically analyzed and manipulated, the faces aren’t pixelated like the ones taken by Google themselves. In other words: Privacy is much worse, as is the overall usefulness. We get the worst of both worlds.

I’m not sure if I should applaud Google for the pluck to go with this hack; if I should congratulate privacy groups for a success (if you want to call it that) in standing up to a large corporation; or if I should lament that our minister for consumer protection just delivered the worst possible result in this conflict, which means that neither user rights (privacy) nor user needs (useful navigation services) are met.

In a way, this whole play really summarizes all that’s skewed in the odd love-hate relationship that Germany has with the web.

Full disclaimer: I’ve worked with Google before and I’m a member of the Google Internet & Society Collaboratory.

[permalink]