March 6, 2012

"A free market is a good model for the internet - and markets only do their job if property rights are well-defined" (*). Consistent with its constant editorial line, the Financial Times has now come out publicly in favor of eprivacy.

"The principle must be that data are the private property of the users they are obtained from". Even if the phrase "obtained from" leaves in limbo the majority of confidential data known about users, either "observed" or "inferred", this influential newspaper has endorsed a major tenet of my fillips.

This is hardly a case for loud rejoicing though, rather a quiet solace as misery loves company. For we need a miracle to stop the well known rogues whose fortune is based on appropriating our data. Only last month the talk of the town was that our address books were shared a bit beyond our expectations. For slow learners, the latest news is that our digital photographs share the same fate.

"After a user allows an application on an iPhone, iPad or iPod Touch to have access to location information, the app can copy the user's entire photo library" (**). Nick Bilton had no axe to grind with Apple. Two days later, Brian X. Chen and he added: "Google [...] takes things one step further. By design, Android apps do not need permission to get a user's photos" (***).

If not held to strict fact finding, wouldn't they state any data put in the care of another party will fall victim of benign neglect, if not downright malice? Instead they quote "Ashkan Soltani, a researcher specializing in privacy and security". "Users typically presume some care is given when designing [...] platforms such that their personal data is handled in a consistent way [...] However, this seems to repeatedly be a false assumption".

Actually great care goes to developing those modern platforms, but to favor companies, not users. Read Tanzina Vega on Facebook's recent announcements (****). "Users could soon be faced with a lot more advertising - in their newsfeed, on their mobile devices and even when they log off" and they should expect ads "based on people's expressed interests". Ah, yes, the famous data which should be theirs to control in the first place.

Yet the relationship between users and providers, serfs and masters as I wont to picture them, is not as straightforward as it seems.

If the master is not up to speed, the serf will desert. In the dark ages, I mean twenty years ago, a second was the time ideally allotted to document processing software to display a new page. "Wait a second. No, that's too long" is how Steve Lohr timely updates us (*****). "People will visit a Web site less often if it is slower than a close competitor by more than 250 milliseconds".

Do not get me wrong. The reason is not that users are more impatient today than twenty years ago. It is software speed which has slowed down. "Download times on personal computers average about six seconds worldwide, and about 3.5 seconds on average in the United States".

This illustrates the helicoidal view of history held by these fillips. Computer engineers may seem locked in some Groundhog Day sequel (1) but what used to take place inside an office building is now happening over untold distances. Users are not grown more unreasonable, they have remained only too human. Meanwhile "the major search engines, Google and Microsoft's Bing [...] typically deliver[s] results in less than a second".

When it comes to consumers, it is not hard to see the consequences of a culture of speed and fun. In this context, safety tends to be a killjoy, costing users too much time and thrill and companies too many human resources. What would be the percentage of motorcycle riders who would wear a helmet without helmet laws (2)? What would be the percentage of cars with built-in safety devices without built-in safety mandates?

Steve Lohr introduces "Arvind Jain, a Google engineer who is the company's resident speed maestro". Rather than a conductor, the person in charge of privacy at Google looks like the physician who monitors patients under torture. As long as they do not die, everything is OK. Contrary to more primitive forms of torture though, the more painless it feels, the more information it yields. With History, a turn of the screw may deliver relief.

This unholy alliance between users and companies for speed and fun at the cost of safety looks strong. But not all is well in the Information Age. I have warned the latter suffers from speed addiction and called it a structural flaw. This truth is becoming fast apparent in high-frequency trading. Telis Demos reports that "while regulators acknowledge "algos" as a necessary tool to create liquidity [...], they are increasingly worried about "stupid algos" [...] that may distort prices and disrupt trading" (******).

Google defended its targeting ads to the content of emails by saying machines, not humans, read our exchanges without our freely given permission. If "stupid algos" roam the world though, this defense looks more like an acknowledgement of willfull carelessness. What are apps if not algos?

Speed however is not merely a matter of execution, whether to display a page or trade a security. Users indeed spend far more time in trying to decide what to do than in doing it. According to Tim Bradshaw (*******), Apple offers no less than 550,000 apps on its store. Were none of them "stupid" or plain criminal, the issue remains for users to know which one to pick.

Apple has acquired Chomp, "a search engine start-up launched in 2010 to help people discover new and interesting applications". Not to insist on my helicoidal model, but isn't it Google all over some ten years later?

Remember Google search is but a recommendation engine for the Web. As recommendation systems are a vital part of the Information Age, I was going to write a whole fillip on Chomp. I find more efficient to ask myself: what is new compared to the last time engineers solved this problem?

Not that "developers have turned to underhand means to boost rankings". Google bombs are old hat. Not so much that the Chomp "algorithm learns what apps do, rather than just using their names". Google also looks beyond the primary text. The real difference is that Apple controls the content of its store while Google can only dream of controlling the web. We have gone from an open world to a closed one. This exarcerbates the lack of decentralization which characterizes Google's and now Apple's approach and sacrifices transparency to the needs of large scale success.

Users' time finitude encourages companies to focus on speed over eprivacy. It also increases the controlling power of walled up content publishers.

Philippe Coueignoux

• (*) ............. Privacy property, by editorial column (Financial Times) - Mar 2, 2012
• (**) ........... Photos Vulnerable on Apple Devices, by Nick Bilton (New York Times) - Feb 29, 2012
• (***) ......... Android Phones Can Secretly Copy Photos and Post Them on Web, by Brian X. Chen and Nick Bilton (New York Times) - Mar 2, 2012
• (****) ....... Substantial Growth in Ads Is On the Way to Facebook, by Tanzina Vega (New York Times) - Mar 1, 2012
• (*****) ..... For Impatient Web Users, an Eye Blink Is Just Too Long to Wait, by Steve Lohr (New York Times) - Mar 1, 2012
• (******) ... CFTC seeks to tighten regulation on 'algos', by Telis Demos (Financial Times) - Mar 1, 2012
• (*******) . Apple buys Chomp to cut through the apps, by Tim Bradshaw (Financial Times) - Feb 25, 2012
• (1) see about the movie Groundhog Day in the wikipedia
• (2) see helmet laws in the wikipedia