Brain activity can identify preferences

[INRM]

Arthwollipot,

What does the physical machinery look like?


INRM
BTW: I checked google awhile back looking for this -- couldn't find anything useful

 

[Holler Hoojer]

Marcel Just is also doing something along these lines at CMU, using FMRI, and claims he will be able to "read" minds in 3 - 5 years.

 

[INRM]

Holler Joller,

He'll be able to "read minds" with fMRI's in just 3-5 years?

Don't you think this is cause for ethical concern?


INRM

 

[INRM]

Arthwollipot,

I found out what the device looks like. It looks like a type of head-band. NASA also developed a similar device which looks like a helmet.


Holler Joller,

CMU does a lot of work with the government right? Well with that said, it's highly likely that this technology will be used by intelligence agencies in the future...

Unless something is done, even the most basic of privacy (which is already jeopardized in many ways) will go the way of the dinosaur.


INRM

 

[arthwollipot]

Arthwollipot,

What does the physical machinery look like?


INRM
BTW: I checked google awhile back looking for this -- couldn't find anything useful

Arthwollipot,

I found out what the device looks like. It looks like a type of head-band. NASA also developed a similar device which looks like a helmet.

I'm not sure why you were asking me what it looks like. I have neither knowledge nor interest.

In exactly what way would the use of this device constitute an invasion of pricvacy, INRM? Especially considering it could not possibly be used covertly?

 

[The Man]

Ok, I finished looking over the paper (some time ago but have been unable to post) and although I find the paper interesting the claim of 80% accuracy is dubious at best. They had to do a considerable amount of data mining just to hit their mother load of an 80% data bias. As specific detectors and emitter combinations for most of the features were unique for each individual subject. Only one emitter detector combination was “Feature 1” for two individuals and another as “Feature 2” for another two individuals. At least three combinations that were “Feature 1” for one individual were “Feature 2” for some other. As one might expect since we are all not exactly the same. Then very specific time intervals for each unique detector emitter feature for each subject were applied. For at least three subjects the time intervals considered do not even include the presentation of the second drink choice. Of course some people make up their minds right away and do not change them. No detector emitter combination and time interval is the same for any two individuals even as “Feature 1” or “Feature 2”. The only thing this study tells me is that it is going to be very difficult and require unique calibration or training to possibly get some useful computer interface for specific choice selection using NIS, and that is when you have a subject capable of expressing that choice by some other means so that calibration and training can be verified. However as the authors stated intent is for the development of brain computer interfaces for those incapable of direct interaction, I find that goal highly unlikely given the unique requirements of each individual demonstrated by this published test. To their credit the authors do note the intended target group (totally locked in state individuals) would be “extremely challenging”. Unfortunately that is also their folly “extremely challenging” becomes down right improbable without alternate confirmation from the individual about their preference, but that’s ok because if the system is not operating correctly the TLS individuals probably can’t complain that they are getting the wrong selection anyway.