Friday, September 30, 2011

"Moods on Twitter Follow Biological Rhythms, Study Finds"

I doubt that there are any surprises in this story from the New York Times. It's only worth mentioning insofar as it portends future, more sophisticated data mining from sites like Facebook and Twitter.

Ken Pimple, PAIT Project Director

Thursday, September 29, 2011

"Secret memo reveals which telecoms store your data the longest"

Francis Harvey, to whom my thanks, brought this item to my attention. In an article by David Kravets of describes "a newly released Justice Department internal memo that for the first time reveals the data retention policies of America’s largest telecoms." The article includes a link to the one-page memo. Apparently AT&T has the biggest appetite for keeping track of its user's movements:
The biggest difference in retention surrounds so-called cell-site data. That is information detailing a phone’s movement history via its connections to mobile phone towers while it's traveling.

Verizon keeps that data on a one-year rolling basis; T-Mobile for “a year or more;” Sprint up to two years, and AT&T indefinitely, from July 2008.
Also of interest: Verizon keeps "text message content" for 3-5 days, but T-Mobile, AT&T, and Sprint don't keep it at all.

This information could be useful to policy makers, as the article recognizes.
The document release comes two months before the Supreme Court hears a case testing the government’s argument that it may use GPS devices to monitor a suspect’s every movement without a warrant. And the disclosure comes a month ahead of the 25th anniversary of the Electronic Privacy Communications Act, an outdated law that the government has invoked to obtain, without a warrant, the data the Justice Department document describes.
Stay tuned for future developments.

Ken Pimple, PAIT Project Director

Friday, September 16, 2011

"New emotion detector can see when we're lying"

BBC News reports that a research team from universities in England and Wales has developed a new computerized camera system that "successfully discriminates between truth and lies in about two-thirds of cases" when tested with volunteers (reported by Hamish Pritchard, September 13, 2011). The team spokesman, Hassan Ugail of Bradford University, is quoted as speculating that "In a real, high-stress situation, we might get an even higher success rate," even up to 90%, which is reportedly "similar to the performance of the polygraph."

The beauty of such a system for detecting terrorists at airports is obvious - just scan everyone and double-check anyone who seems to be lying. Employers will also be excited at the prospect of using it in job interviews. Private investigators will invest in a portable unit that they can hide in a potted plant at a restaurant where a woman can ask her husband, point blank, whether he's fooling around. The applications are endless. 

The mischief that such a device could create is nearly endless. The last I heard, the polygraph was widely considered ineffective and of dubious worth in criminal cases. I am also given to believe that polygraphs and fingerprints have never been adequately tested for reliability, so if this system is given more rigorous screening, it might prove to be better than I expect it will be. No matter how (in)accurate it turns out to be, people tend to be so credulous about lie-detecting machines that it will probably be taken as infallible.

The kicker comes in the last paragraph: Like the polygraph, the new system detects "emotions, such as distress, fear or distrust, and not the act of lying itself. Fear can sometimes be the fear of not being believed rather than the fear of being caught." Or the fear of flying, or being water boarded.

Thanks to Colin Allen for drawing my attention to this.

Ken Pimple, PAIT Project Director

Monday, September 12, 2011

Mind-controlled robots, the Supreme Court on GPS, Wii without a wand, and computers that write

A small avalanche of PAIT-related articles buried me this weekend; let me see whether I can dig my way out.
  • Disabled Patients Mind-Meld With Robots by Sara Reardon (ScienceNOW, September 6, 2011) - Using Skype and wearing "a cap of tiny electroencephalogram (EEG) electrodes," two people "whose lower bodies were paralyzed and who had been bed bound for 6 or 7 years" controlled the movements of a modified commercial robot - Robotino -100 kilometers away. They used only their brain waves - no moving eyes, no twitching fingers. The paralyzed subjects had been trained for 1 hour a week for 6 weeks. The system had been tested earlier with non-paralyzed people, and the paralyzed subjects "performed just as well as the healthy subjects." Think of what this would mean to paralyzed people and their friends and families - it would be a miracle.

    My enthusiasm is, of course, always tempered by caution. The Hollywood version would have someone hack the system and take control of the robot to frame a paralyzed person for murder. My real concern, though, is accessibility. There's no mention in the article how much this rig would cost, and the manufacturer of Robotino, Festo Didactic, lists all of the prices associated with Robotino as "on request." I fear this is an instance of "if you have to ask the price, you can't afford it."

  • Court Case Asks if ‘Big Brother’ Is Spelled GPS by Adam Liptak (New York Times, September 10, 2011) - At least three federal judges have compared the use of global positioning system (GPS) devices by police to George Orwell's novel, 1984. In November, the Supreme Count "will address a question that has divided the lower courts: Do the police need a warrant to attach a GPS device to a suspect’s car and track its movements for weeks at a time?"

    The answer seems obvious to me (hint: it begins with a "y"), but with the current court I'm not making any bets.

  • Remote Control, With a Wave of a Hand by Anne Eisenberg (New York Times, September 10, 2011) - "Scientists at Microsoft Research and the University of Washington have come up with a new system that uses the human body as an antenna. The technology could one day be used to turn on lights, buy a ticket at a train station kiosk, or interact with a world of other computer applications. And no elaborate instruments would be required." No need for a Wii wand or the Kinect's cameras that track motion.

    Nothing is said about whether this technology could be used to identify and track specific individuals (is your repertoire of everyday gestures as distinctive as your face or fingerprints?).

  • In Case You Wondered, a Real Human Wrote This Column, allegedly by Steve Lohr (New York Times, September 10, 2011) - I had heard some time ago about the effort to make computers write newspaper-style sports articles based solely on the statistics of the game. That such writing would soon be indistinguishable from prose written by a human sports writer I did not doubt. The time seems to be near.

    Somehow I can't get excited by this one.
Ken Pimple, PAIT Project Director