View Full Version : Random Number Generator "predicting" events
afree87
15th February 2005, 06:29 AM
http://www.rednova.com/news/display/?id=126649
This has been coming up a lot in the blogosphere. I don't know anything about this guy, can anyone clarify what this project is?
If this really is a random number generator, it would be great for making one-time pads!
MRC_Hans
15th February 2005, 06:58 AM
Yes, it is, presumamby, a true random generator. It works either with a radioactive source, where decay events are detected, or it uses a semiconductor junction that generates quantum noise.
So, how can it "predict" things?
Very simple, really: You have a stream of random numbers. And you have a world full of events. Now, whenever the computer monitoring the random numbers has logged an anomaly, which simple probability guarantees it will do from time to time, you look around in the world for something interesting that happend at the same time, or, lacking that, up to a few days later. Given a world teeming with fighting, loving, dying human beings, several hundred notable natural events per year, etc, this search is bound to show up something.
Final step: You claim that the random numbers predicted that particular event. There is no way it can fail.
Hans
Drooper
15th February 2005, 06:59 AM
I read about this somewhere, but can't remember now. It may even have been on this board somewhere.
I remeber thinking that this loked like an obvoius case of data mining.
alfaniner
15th February 2005, 07:46 AM
I was involved in testing a device of this sort some time back. The creator said a light would go on when the (whatever) field or RNG detected something outside of statistical significance. The thought was that the mind, or "intent" could shift this RNG and make the light come on, and that it was unaffected by space and time.
He concentrated a bit, and the light came on (as it had on occasion while we were talking.) I said, "OK, if it's unaffected by time, how do you know it's not the 'future' you that is causing it to flicker on right now, rather than what you just did?" He did not have a reply for that and said I just didn't understand it. I was tempted to tell him to watch more Star Trek for implications of time-travel...
afree87
15th February 2005, 08:33 AM
Aha, just now I found an entry on skepdic.com.
http://www.skepdic.com/refuge/bunk23.html
edit:
Here's their homepage, with raw data. (http://noosphere.princeton.edu/)
Here's what their sum data looked like on September 10, 2001:
http://noosphere.princeton.edu/data/eggsummary/2002/dev_2002-09-10_comp.gif
Here's what it looked like on September 11:
http://noosphere.princeton.edu/data/eggsummary/2002/dev_2002-09-11_comp.gif
Yeah, I don't know what the article is about now.
jmercer
15th February 2005, 11:52 AM
Originally posted by Drooper
I read about this somewhere, but can't remember now. It may even have been on this board somewhere.
I remeber thinking that this loked like an obvoius case of data mining.
Maybe you're talking about this thread?
http://forums.randi.org/showthread.php?s=&threadid=51271
© 2001-2008, James Randi Educational Foundation. All Rights Reserved.
vBulletin® v3.7.3, Copyright ©2000-2008, Jelsoft Enterprises Ltd.