Slashdot Mirror
Indian Woman Convicted of Murder By Brain Scan
Kaseijin writes "Neuroscientist Champadi Raman Mukundan claims his Brain Electrical Oscillations Signature test is so accurate, it can tell whether a person committed or only witnessed an act. In June, an Indian judge agreed, using BEOS to find a woman guilty of killing her former fiancé. Scientific experts are calling the decision 'ridiculous' and 'unconscionable,' protesting that Mukundan's work has not even been peer reviewed. How reliable should a test have to be, when eyewitnesses are notoriously fallible? Does a person have a right to privacy over their own memories, or should society's interest in holding criminals accountable come first?"
It is possible to do a brain scan to detect that a statement is untrue or unsettling in some way, but that doesn't mean that the person is guilty of a specific crime.
It takes a long time of interrogation to be able to measure what's normal and what's not. And even if you get an abnormal reading it may not be caused by guilt - it may be because the subject is unsettling.
If builders built buildings the way programmers wrote programs, then the first woodpecker would destroy civilization.
My guess is they would argue that brain scans are of the same family of evidence as DNA; e.g. it doesn't "testify against you", but is rather physically relevant to the case.
I guess it is a grey area (no pun intended!), but really we shouldn't even need to have that conversation. The study hasn't been peer reviewed, it's a new and relatively untested technology, what the hell are they doing admitting it at all, as testimony or as evidence?
Hell, the last time I saw MRI-based lie detection it was on Mythbusters, and even there it failed outright on one of the three people they tested it on.
In Red Dwarf Arnold Rimmer has to undergo a mind scan after which he is found guilty of the 1st degree murder of the whole crew of the Red Dwarf. Kryton is able to get Rimmer aquitted by pointing out that the radiation leak was caused by Rimmer being an incompetent half wit anf the mid scan confused the guilt he felt with culpability, in his own mind he tried and convicted himself... How would this mind probe deal this?
In the not too distant future, next Sunday A.D.
Actually, I would imagine it would be fairly easy to distinguish a lie from the truth by EEG or fMRI. The pathways for recollection as opposed to creativity (lying), cause activity in different parts of the brain.
'Where were you last Tuesday at 3:00pm?' - If the person tells the truth, they're recalling the events. If they're lying, they're constructing a scenario in their head. The two would be very distinguishable.
That said, it's not without issues: First of all, if I pre-construct a scenario and run it through my head enough, it becomes a recollection and not a creation, I believe. Also, I'm not entirely sure that there's been enough actual studies of using fMRIs and/or EEGs for detecting lies vs. truth, nor how beatable the system is. Until these things have been studied and documented, they certainly shouldn't be used by courts.
There are companies in the U.S. trying to get fMRIs used for precisely this purpose. One example is the company, No Lie MRI.
If such systems can be proven reliable, then I'm all for using them in courts. Not so much to convict people, so much as to keep the innocent from being convicted, which happens plenty in the U.S.
I'm not so sure that it is good to convict someone of a crime, but it is pretty accurate. It is simple to do with a brain scan too.
1) Hook person up to a brain scanner.
2) Show the person random images of places they never seen until their brain doesn't care anymore.
3) Show the person an image of a place they've seen, and it will trigger thoughts.
It is helpful for interrogation. It is a bit spooky to use for crimes.
God spoke to me.