100% accurate lie detectors?

My flatmate asks me a lot about if there will be 100% accurate lie detectors in the future and how much it would cost to use them… any ideas? Thanks.

Well, it seems likely that a good enough understanding of the brain should allow some sort of scanning technology to come close. Although the results would probably be less binary than Truth or Lie, since people so often shade and distort or just are selective about the facts rather than just flatly lying. As for how expensive it would be there’s no way to know really, given that the technology doesn’t even exist yet.

Thanks, btw I did a little research…

http://noliemri.com/

http://noliemri.com/products/Overview.htm
“Current accuracy is over 90% and is estimated to be 99% once product development is complete.”

Yes.

$47.50, adjusted for inflation.

You wouldn’t necessarily need an understanding of how the brain works. Artificial neural networks and similar machine learning methods are pretty good at teasing out fuzzy concepts without any actual understanding of how it works.
Of course, the hard part is gathering accurate data to train the computer. A researcher saying “my nose is five metres long” probably results in different brain activity than when an actual suspect in a criminal case says “she was dead when I got there”.

I’m not sure I get the part of the question about how much it would cost to use a lie detector. I’m pretty sure that I’ve never seen a coin-operated lie detector at Walgreens that’s for general public use. Wouldn’t such a thing normally be a one time purchase by a police department? In that case the cost for a single use would be pennies for paper and ink.

I think they’re lying. Seriously.

Is this the MRI lie detector that Grant from Mythbusters was successful in fooling?

I was watching some show on daytime TV the other day at the laundry mat where the schtick is apparently putting people on lie detectors and having them confront family or friends about their “lies”.

One guy was adamant that the lie detector was wrong (said he bought a prostitute or something), and the host kept saying “the detector doesn’t lie, you do.” or something similar.

To the extent that some questions may not have simple yes/no answers, or that the subject may be convinced, rightly or wrongly, that the truth is “gray” (remember “depends on the meaning of ‘is’”, or the definition of “sex”?), then a 100% accurate lie detector is fundamentally impossible, whatever the technology.

But how do you distinguish the actual truth from what someone believes to be the truth?

As George “Seinfeld” Costanza said “It’s not a lie, if you believe it.” :slight_smile:

Unless the machine can also detect the particular form of delusion/error the subject is falling for, all you can do is check the facts. It’s a lie detector, not a Truth detector.

What I have seen on television was not so much a lie detector as it was a machine that could detect whether a crime scene (as depicted in a photograph) was familiar to a suspect. There were certain involuntary reflexes in the brain or in the eyes that could be measured.

Excuse me if I remain extreme skeptical.

Exactly why does he want to know?:dubious:

There are no lie detectors at present, only stress detectors.
They don’t work well and most countries won’t allow them to be used in court.
I think you could do just as well by hooking nervous suspects up to an impressive looking fake machine and announcing triumphantly “You’re lying!”

If you would prefer a perpetual motion machine, here’s one you can invest in.

Apparently due to his mental illness he has false memories that never happened… I guess he wants to prove he isn’t lying about them (though he’d be falsely believing it)

There’s at least one already; I call her Mrs. Extra_Sharp.

Polygraph machines are already 100% accurate. They measure every breath, every skin response, every heartbeat with total accuracy.

The reason it’s not reliable is because of human operator bias, poorly worded questions, and human interpretation of the results. Even if they come up with a machine that measures different physical responses than a polygraph, these problems will still remain.

It’s not just a problem with operators. Although the machine may perfectly record its data, those data do not perfectly correlate with lying or truthtelling - not by a long shot. Even if you know precisely what the body is doing, and even if the process is performed perfectly well, you don’t know exactly what that means. At best, it’s an educated guess.