I will admit it; I am a lazy person at heart. But I’m not the typical lazy where I just prefer to do nothing. I am one of those people who perfers to put 10 hours up front to save 40 hours on the backend. I enjoy researching ways to automate my life. I once bought a robotic lawnmower to avoid going out in the hot summer sun.
When I stumbled across the concept sleep learning, I was immediately excited. Who wouldn’t like to fally asleep to the audio of your favorite Cival War textbook and have instant recollection the next morning?
With the research done by Fox and Robbin, it’s shown that non-consious suggestions do have a fighting chance of being retained, but what good is this if there aren’t any correlations drawn with the rest of your knowledge? Maybe if you’re training for Jeopardy (which isn’t a bad idea) passive retention is OK, but memorizing trivial facts is just that, trivial.
Is there a better way?
While there are older studies outlining passive learning while sleep, would it possible to become to induce a lucid sleep and then administer the desired knowledge to be retained (e.g. audio stream)?
Take a look at these devices…
It’s a build-it-yourself lucid dreaming induction device. There are a couple commercial equivalents out there that are pretty expensive (plus they would have to be adapted for what we need for testing).
Here’s the idea….
The above device would have to be modified so that it could be connected to a PC (e.g. via USB). The user will be required to wear a small earpiece for which audio can be played through (maybe a little radio receiver for the earpiece and the PC audio jack is connected to a small FM transmitter?).
The lucid dreaming induction device would be required to report the amount of eye movement within the given user. Not necessarily just the number of eye movements while in REM sleep but for the very beginning stages of sleep we would want to record this.
As soon as we can detect that the sleep process has been entered, our system will connect to a generalized knowledge site, say Wikipedia.org, and will retrieve the text of a random article (via their RSS feed). The system will then playback part of this article’s text through the through the user’s earpiece (we can find some sort of text-to-speech library to generate the spoken audio).
A different article is played at each different stages of sleep. The full sleep cycle will be allowed to proceed as normal with the only exception being that the system will provide the audio/visual cues in an attempt to induce a lucid dream. The point being that we try to introduce various facts throughout a full spectrum of unconscious states.
Once the sleep cycle has completed and the user awakens, the user immediately returns to the PC for which the device is connected to. Utilizing the various text articles that the system read to the sleeping user, the system will derive simple questions from the article texts and present those to the end user in a quiz format (e.g. what color was the dog? how tall was John?)
The user repeats this process of sleep/quiz for several weeks. Since the system knows what sleep state each set of facts was presented, depending upon where the user gives the most correct answers, the optimal sleep state for unconscious knowledge retention should be revealed.
Which begs the question, if we did find this magical state where we can effortless digest knowledge, could we induce emotional states as well? What would happen if we’re in this vulnerable lucid state and our system plays an audio track with Beethoven’s Moonlight Sonata overlaid with cries of people in panic?
Update: 7/21/2006 10:03pm
I’ve had a few suggestions for upgrading my design.
For one, you may be able to build your own EEG machine for the same price as the REM sleep tracker listed above. Utilizing the OpenEEG framework ( http://openeeg.sourceforge.net/doc/ ), and some low cost electrodes, the EEG may provide a more accurate measure of brain activity and sleep cycles.
A neurophone offers an exciting opportunity in terms of administering the audio queues. If standard audiotory response shuts down during REM, what about the possibilities of ultrasonic sound ( http://www.neurophone.com/home.htm )?