[extropy-chat] I knew this was coming

[Brett Paatsch]

> It is too early to tell whether fMRI can be fooled in the same way as the 
> polygraph, says Faro. However, he says that the results are promising 
> because these characteristic brain patterns may be beyond conscious 
> control, rendering it much more difficult to cheat.

As I understand polygraphs (and I have only a lay understanding which
could easily be wrong) they work on galvanic skin response. A person
convinced that they can't be caught out in a lie is likely to be pretty 
"cool" about lying and not be stressed enough to produce changes in
skin chemistry.  I do get how combining the equipment with a skilled
interviewer could improve the results by getting better baselines for
comparison (by asking questions that would also trigger other sorts
of emotional and stress responses rather than outright lying, and then
comparing), but ultimately that means its not just the polygraph but
the polygraph plus "expert" interviewer that is actually producing the
result. I think polygraph interviewers themselves could probably learn
to fool the polygraph plus other interviewers by knowing how to beat
not the technology but the interviewer. 

I'd have thought the fMRI would work on a quite different principle
whereby what would show up would be activated areas of the brain.

I recently saw a TV show in Australia where it was alledged that 
scientists were claiming to have been able to use fMRI to track single
thoughts. 

Sounds like neat technology but I've grown sceptical of most types
of popular technology reporting where the mechanisms are not
explained. Often the most incremental advances are reported along
with the wildest most optimistic extrapolations of their potential
impact.

I'd be in favour of good and widespread lie detector technology
but I'm scpetical that it will be easy to produce any technology that
will not also come coupled with the need for an accompanying 
human expert to interpret the results. 

Lying with varying degrees of skill and success seems to be too
much an integral part of what people do to succeed in a competitive
world made up of other people to be bypassed by simple tech
solutions.  I suspect (but don't know that) lie detecting tech will
work only on some folks that think it will work on them and are
not particularly interested in beating it.

Brett Paatsch