Reading Terrorists' Minds About Imminent Attack

An anonymous reader writes "Imagine technology that allows you to get inside the mind of a terrorist to know how, when, and where the next attack will occur. In the Northwestern study, when researchers knew in advance specifics of the planned attacks by the make-believe 'terrorists,' they were able to correlate P300 brain waves to guilty knowledge with 100 percent accuracy in the lab, said J. Peter Rosenfeld, professor of psychology in Northwestern's Weinberg College of Arts and Sciences."

Comments

Lab Accuracy != Real World Accuracy

[Da Cheez]

"they were able to correlate P300 brain waves to guilty knowledge with 100 percent accuracy in the lab"

Bet the accuracy wouldn't be so good in a non-controlled, non-laboratory environment. Of course, that wouldn't necessarily stop such a technology from being used, now would it?

Re:Lab Accuracy != Real World Accuracy

[camperdave]

"Without any prior knowledge of the planned crime in our mock terrorism scenarios, we were able to identify 10 out of 12 terrorists and, among them, 20 out of 30 crime- related details,"

Yeah, 10 out of 12 is 100%. We need to give these guys more money so they can upgrade from their first generation pentiums.

Re:Lab Accuracy != Real World Accuracy

[slashqwerty]

It is really easy to achieve a 100% true positive rate. Just accuse everyone.

The article didn't mention false positives. It would not surprise me at all if this technology would have at least two orders of magnitude more false positives than true positives in the real world. You can't get away from the fact that terrorists are rare so they will be lost in the noise of all the people who are not terrorists.

Let's say the police go through 50 suspects, none of whom are terrorists. With an 83% accuracy rate the odds of all suspects correctly identifying no targets is 0.83^50 = 8.99 x 10^-5 = 0.00899%. In other words, with just 50 suspects there is better than a 99.99% chance law enforcement would be acting on bogus information. It takes only four suspects before there is better than a 50% chance of acting on bogus information.

Real world use would likely see results worse than the 83% achieved in the lab.

Terrorists schmerrorists

[line-bundle]

Why is everything legitimized by putting the word terrorist in it? What does this have to do with terrorism?

As someone said here on /., terrorism is one of the magic keys, the other being child porn.

Re:so PRE crime starts now and how do they jury tr

[KDR_11k]

As long as public indecency is illegal we all have something to hide.

Re:Define 'guilty'

[maxwell demon]

and if the device lights up to say that you recognize Bob, then I know you just lied to me.

No. You have an indication that he lied. Maybe his brother knows Bob, and he has seen him once with his brother but didn't know who he was. Then he was 100% right when he said that he didn't know Bob, but he nevertheless recognized the person on the picture, although he didn't recognize him as Bob, but as the person his brother was talking to. Or maybe he was earlier shown a photo of Bob by another policeman who forgot to tell you about that detail, and he recognized the photo as the same one the policeman had showed him a week ago. Or maybe Bob looks quite similar to John, and he momentarily mis-identified the man on the picture as John, maybe not even long enough for this recognition to get into his consciousness, but long enough for his brain to cause the characteristic pattern of recognizing.