Electronically-assisted telepathy?

Jan 02, 2008 13:58

Scientists conducted an experiment where they asked people in an MRI brain scanner to look at ten pictures of different tools and houses. Meanwhile, the researchers fed the data into a computer algorithm so that the program would learn to recognize the unique signature of electrical patterns produced by the objects in each subject's brain. ( Read more... )

psychology, information science, intelligence, biology, science, computers

Leave a comment

anonymous January 3 2008, 08:42:56 UTC
But it seems we're more alike than that. Why? I can't think of any way we could have evolved a specific neural net for recognizing a screwdriver.

Hm. Perhaps laziness?

Assume that it is possible to create a computer simulation of the world as we know it (n.b. that this doesn't require that there is some external world corresponding to our own; the simulators could be BEMs) that has sufficient processing capability that the simulated creatures on the world are as sapient as we are. Further assume that computing resources are cheap enough for the simulators that they could spawn off any number of such simulations. Finally assume that some simulators able to do this are willing to do so (e.g. no ethical qualms for Teller-oid-7). Sheer numbers would indicate that we're more likely simulated than (potential) simulator. Bostrom has written about this.

And the programmer was a bit lazy with the human code, and just reused the same mental maps, with only minor random variations, or whatever.

Also works well to explain the Fermi paradox; the simulators are interested in us, or something involving us, or something about the world as we know it (e.g. 'Let's see what would've happened if we dinosaurs were wiped out' or 'I bet you couldn't have sapient organic life on a planet with lots of water') and didn't worry about modeling too much of the rest of the universe. The anthropic principle also seems to come up.

Also see:
http://angryflower.com/benefi.gif
http://angryflower.com/alight.gif
http://angryflower.com/anaddi.gif

And yet, the intelligent design folks probably wouldn't care for the idea! Go figure.

--josh

Reply

tevarin January 4 2008, 02:52:22 UTC
Intelligent design + re-use of code sounds like a viable explanation, whether the designer is deific, alien, extra-universal, or whatever.

There may be other possibilities, similarly outlandish: Maybe our neural circuits get their templates from a Jung-style collective unconscious?

It could be more localized. If a child's caregivers somehow imprint their own brain patterns onto the kid, then a whole culture could end up with similar patterns. The imprinting could be subconscious telepathy, or somehow mediated through speech, body language, etc. You could test this theory by using the brain scanner on patients from very different cultural backgrounds, New Guinea tribesmen and whatnot.

I'm not sure I buy the other arguments for us being a simulation. I'm inclined towards the first and second options in Bostrom's tripartition--

At least one part of Bostrom's tripartition must be true:

1. Almost no civilization will reach a technological level capable of producing simulated realities.
2. Almost no civilization reaching aforementioned technological status will produce a simulated reality, for any of a number of reasons, such as diversion of computational processing power for other tasks, ethical considerations of holding entities captive in simulated realities, etc.
3. Almost all entities with our general set of experiences are living in a simulation.

http://en.wikipedia.org/wiki/Nick_Bostrom

--The requirements for simulating the full sensory experience of a human being are considerably beyond our current tech. Ditto for providing the computer to run the simulated being's mental program. Maybe nobody (or almost nobody) can do it, and any who can would have to devote a large fraction of their power to maintaining relatively few simulated realities. The odds of our being simulated go down accordingly. I realize this is a pessimistic assumption, that Moore's law won't go on forever, that computing power won't keep getting exponentially cheaper. If we actually do go through a Singularity, I'm willing to admit I was wrong :)

Reply


Leave a comment

Up