Mind reading machine goes online

[catmoon]

Using sophisticated MRI machines and advanced decoding algorithms, it is now possible to view images as they form inside the brain and display them on a video screen. The technology is at about the same stage as the very earliest photography; the images are blurry, noisy and sometimes unrecognizable. In thirty years, it must be assumed that the privacy of your mind will be invaded. In a century, handheld devices should be capable of crisp clear image extraction, possibly from a considerable distance.

[plwk]

Catmoon's

[catmoon]

[maybay]

TV's have been around for ages, no?

[catmoon]

Yeah, but you can't use one to see what the guy next to you is thinking...

[Kare]

Online mind reading has been around for a while.

Just check it out:

http://www.caveofmagic.com/

[Nemo]

If it works you will never hear about it again.

[Pero]

Online mind reading has been around for a while.

Just check it out:

http://www.caveofmagic.com/

Hehe, took me a couple tries to figure it out.

[Adumbra]

Good thing I meditate a lot. I can blank my mind out at any time.

I think this is promising development. An open society is a less judgmental one. Who am I to condemn you for your racy thoughts if I'm addicted to gambling, you dig?

[tomamundsen]

Yeah, but you can't use one to see what the guy next to you is thinking...

This machine doesn't either. Having some background in machine learning and computer vision, I can make a couple guesses as to how this works. And, while it is impressive, it's nothing close to actually reading minds.

Basically, they take a snapshot of a video and a person's brain simultaneously. They identify features in the movie image and in the brain image. Then you can use probability to determine if this feature in the video image occurs, then with some probability, this feature will be seen in the brain. Using this kind of model, if you only have images of the brain, you can attempt to reconstruct the corresponding video image. So, as you can see, it doesn't really involve a deep understanding of the brain itself. You just have to be able to identify _some kind of features_ in the images of a brain scan. You don't have to understand them. This means it would take a lot of work to use this same technique for any other kind of mental phenomena. Each phenomena would have its own entirely different model.

[catmoon]

true enough,but I noticed something disturbing in the images. There were persistent bars of light, rectangular in shape in some of them. I believe that they are a sort of after image of bits of text that had previously appeared on the screen, and they stayed there because the subject was still thinking about what he had read.

So although the bulk of the signal is just as you say, there are some odd things in them that should not be there if we are just looking at what the subject's eyes are seeing. Stay tuned, nobody has yet tried the same technique on the cognitive centers of the brain.