h a l f b a k e r yNice swing,
no follow-through.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Some experiments in the 1960’s (they’re in a good book called “velocity of honey”) gave me the inspiration for this idea. The first had a group of men look a series of photos of women, one of which had her pupils dilated. Not only did the men judge her to more attractive, their pupils dilated too. A
second similar experiment consisted of a group of hungry people looking at photos of random objects, one of which was a slice of cake, again the volunteers eyes dilated at the sight of it, thus proving that the dilation was a sign of interest. The part of the brain responsible for this was later discovered (a part of the ventral striatum).
Now comes my idea, point a MRI at that part of the brain and convert the level of activity to a variable, the higher the number the stronger the interest in the object. Now put a screen in front of your helpless victim and display a random image. Then show a series of distorted versions of the image, measuring the response to each one. Then take the image that generated the largest response and start showing variants of that one, and so on until you end up with the highest possible response, which will be an image of the thing they want most. E.g., if they where hungry you would get a picture of a pie or an apple or something, if they like someone you get an image of him/her, if they need the loo, you get a picture of that.
Of course, this is probably against every human rights bill on the planet.
the ultimate lie detector?
http://www.webmd.co...timate-lie-detector
[po, Aug 22 2007]
Images from the brain.
http://boingboing.n...ists-extract-i.html
[DrBob, Dec 19 2008]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Destination URL.
E.g., https://www.coffee.com/
Description (displayed with the short name and URL.)
|
| |
Well, I have to say, this is certainly half-baked. |
|
| |
Compared to asking the poor sap you have forced into the MRI scanner against his human rights what he might want the most, it's expensive, impractical, unnecessary, probably mostly ineffective, and generally completely useless. In other words, well done. |
|
| |
Croissant for your first one. |
|
| |
(be aware, however, that the grammar and spelling police will be paying you a visit soon) |
|
| |
Your right, I would get lots of pictures of people murdering me, can’t think why! |
|
| |
<A white rabbit mysteriously forms on the screen> [+] |
|
| |
I like it. I like it alot. Obtain grant moneys and strike ground on an institute, reserch and develope an expirimental system, perfect the system, produce a prototype, a self-contained unit that runs under its own power, produce a smaller, portable version and mount it to a helmet and call me. A home version of this would be ideal. |
|
| |
oh and welcome to the bakery. |
|
| |
Wouldn't context be one of the determinants in the level of interest generated by a given image, especially in marginal changes in level of interest? And wouldn't the context consist largely of your experimental procedure? And wouldn't this be a major confounding factor in your results? |
|
| |
Welcome to the bakery. Answer me my questions three, and put in at least one paragraph break (at //Now comes my idea//), and you might get my vote. |
|
| |
Not if the need is pressing (Loo Food Love Water etc). Otherwise yes, especially the stress of being jammed into some big complex machine. Thanks for the thought. |
|
| |
It wouldn't be the thing they want most; it would just be the thing that your starter image can gradually be turned into, with each step more desirable than the previous, that they want most. (In Operations Research terms, you'd get a "local maximum", not a "global maximum".) |
|
| |
Depending on where you start and how wide you cast your net, it could still be pretty embarrassing. |
|
| |
Yes Jutta, but this could actually be an advantage, for example you could start with a generic human face and go from there, that would allow you to fine tune your search to ensure you get something embarrassing (eg start with a generic human face). Although you would need to take some kind of guess to where to start, this wouldn’t be too hard. |
|
| |
Or you could test out a whole string of generic images of things and start from the best one. |
|
| |
And BrauBeaton, but most likely they don’t know what they want most, or won’t admit it to themselves, or they came to see me fail. Or I use the cattle prodder. Easy! |
|
| |
Captain Jack Sparrow's compass pointed the direction of *what he wanted the most*. |
|
| |
welcome, [deep fry], have a mental image of a bun for your efforts. + |
|
| |
The marketing people will ge a hold of this in no time. Product shapes will start to change and we'll all end up walking around with dilated eyes. The good news is that everyone will start looking more attractive to each other, the bad news is that the population will increase. |
|
| |
I think what I would want most after this long process would be the exit sign. |
|
| |
/determine yes-no answers for locked-in patients/ |
|
| |
This anno gets my bun. What an excellent idea, [longshot]. Not only for locked-in folks (who could read the yes and the no, and presumably dilate appropriately), but for any person who can recognize an image. It could be used for people who have never been able to communicate verbally - for example, mentally handicapped folks. |
|
| |
May I compliment you, Jutta, on your refined understanding of why evolution can not make us perfect, and comes u with so many bizarre solutions? But, I suppose that dealing with so many bakers you have more experience with imperfection than most. |
|
| |
Where did perfection come from? Anyways, anyone with a slight grasp of controlling their emotions will be able to create any image they want. Sure, some folks will generate a two headed sunflower with penis petals and never knew where it came from: lesson learned. Control yourself. |
|
| |
It's only peripherally related, but this idea is at heart an evolutionary process. As such it can only find the local maxima (in biology they are called evolutionary plateaus, presumably because many of us are bad at math and good at geology) that Jutta refers to. |
|
| |
A (probably unnecessary) example from the idea: You have captured James Bond, and loaded him into your machine. Since you are a villian, you have not fed him in the six hours since capture. |
|
| |
You ask, "where do you meet your contact!" and show a blob on the screen. As he is hungry, he sees a slight resemblance to an ice cream cone. His brain heats up (or whatever) as the image develops chocolate sauce, sprinkles, a cherry on top. |
|
| |
The image will never switch to show James Bond giving you the finger from the back of the 4 o'clock train to Innsbruck, the thing he wants most in the world. This is because to do so it would have to go through a period of looking half like an ice cream cone, and half like a spy riding a train; in other words looking like nothing. This would excite no part of his brain, and thus there can be no evolutionarily derived transition to the desired result. |
|
| |
Is where perfection came from. |
|
| |
In evolution, it is the reason your optic nerve goes in front of your retina, blocking your view: The transitionary period (thousands of generations?) to proper placement behind the retina would leave all the generations in the middle blind. It just can't happen. |
|
| |
Why do you scan the brain when you could just measure the retina? |
|
| |
And won't the brain signals change once the retina is fully dilated? |
|
| |
I guess it is possible. But then it is also possible to randomly start out at the absolute maximum, isn't it? And then what would you do? |
|
| |
I think they once tried this with some guys and a more... obvious measure of interest. They were able to associate images not previously of 'interest' so that they caused an 'interested' state. |
|
| |