Slashdot

Just a quick reminder of the facts:

Brain scanner can tell if you are going to buy a product or not:
http://www.boingboing.net/2007/01/11/brain-scans-predict-.html [boingboing.net]

Brain scaner can tell what you are looking at:
http://science.slashdot.org/article.pl?sid=08/03/06/0435226 [slashdot.org]

Brain scanners are so easy to do that now they are in game controllers:
http://hardware.slashdot.org/article.pl?sid=08/02/20/1314254 [slashdot.org]

And better than a tinfoil hat, we will need something able to filter what you let or do not let through, as was done with the rfid firewall:
http://www.rfidguardian.org/index.php/Main_Page [rfidguardian.org]

It's even crappier. We already know know about the normal polygraphs that they don't really work. They just mention someone's reaction to stress, and from there it's a leap of faith that "lying emotional stress". The latter just isn't so.

1. As you mention, what do you do about people who genuinely believe something bogus?

As a milder example, human memory isn't photographic, ever. It seems to store more like the description of a scene, and just ad-lib the details that it forgot. Over time you'd forget that, say, the guy was wearing a blue shirt, or maybe that detail never even made it into permanent memory in the first place. But if you try too hard to remember it, it will just give you some best guess. Like that he was wearing a black shirt.

2. We know that people can train to not feel much emotion about lying, and to psychopaths it even comes naturally. So even measuring their pulse and blood pressure and everything directly, you just can't tell that they're lying.

Basically we're relying there on the false idea that everyone was educated that it's not nice to lie, and everyone therefore has a hard time telling one and is feeling severely guilty about it. Which is false from start to finish. E.g., speaking of education, we know that some people's upbringing just taught them that it's perfectly _normal_ and indeed _logical_ to tell a lie, if the alternative is a savage beating by your father. They won't feel any guilt extrapolating from there to lying to save their arse from jail.

3. That emotional stress someone is feeling, can be for a bazillion other causes.

E.g., because the topic is painful to them for other reasons. A rape victim being the witness in someone else's rape trial might experience severe stress just thinking about it, whether they tell the truth or not. A PTSD [wikipedia.org] sufferer will be in a disproportionate amount of stress when recounting the event that caused it, or anything that reminds them of it. So, you know, some grandpa who fought in Vietnam and still wakes up in cold sweat after dreaming of it, would register as shamelessly lying when they tell you about the atrocities of war. Etc.

E.g., particularly bad cases of repressed memories and/or the results of some particularly hard to justify cognitive dissonance, can cause a disproportionate emotional responses when you're forced to think or talk about something which challenges them. You see that not only in polygraph tests. A lot of people who are rabidly against something are really just against you challenging their already decided model of the world. The less of an actual justification they have to support that position, other than "but my daddy said so", actually the harder it can be to get them to think logically about it.

Etc.

Basically let's just say there are good reasons why that test can't be demanded in court.

So now we have something that promises to test one parameter from a distance, instead of several measured directly, and which must correlate in certain ways to be considered a "yep, he's lying" proof. It's basically adding one more indirection step to that already weak inference chain. But even if the correlation between skin pores and all those parameters were that infallible, you're back to "stress he's lying", which is already known to be false even measured up close with electrodes.