In Wim Wenders' 1991 near-future science fiction and "ultimate road-movie" Until the End of the World, a device that translates images into neural impulses (a "camera for the blind") gets reverse-engineered, allowing people to record and view their own dreams. As characters are exposed to the technology, they become addicted, running and re-running memories and dreams in a lab in a cave in the middle of the Australian desert.
Aparrently, by setting his film in 1999 Wenders was only off by 12 years:
"researchers used functional Magnetic Resonance Imaging (fMRI) to measure the blood flow through brain’s visual cortex. Then, different parts of the brain were divided into volumetric pixels or voxels... Finally, the scientists built a computational model which describes how visual information is mapped into brain activity.
Test subjects viewed some video clips, and their brain activity was recorded by a computer program, which learned how to associate the visual patterns in the movie with the corresponding brain activity.
Then, test subjects viewed a second set of clips. The movie reconstruction algorithm was fed 18 million seconds of random YouTube videos, which were used to teach the program how to predict the brain activity evoked by film clips. Finally, the program chose 100 clips which were most similar to the movie the subject had seen, which were merged to create a reconstruction of the original movie. The result is a video that shows how our brain sees things, and at moments it’s eerily similar to the original imagery."
It remains to be seen whether this new technology turns us all into weeping solipsists or not, but it is an amazing example of how the rate of technological change has compressed the idea of "futurism." Remember, Wenders' film was also one of the first to feature in-car GPS, a mind-blowing millennial technology in 1991.
(Also, it is an interesting footnote to look at the role YouTube plays in the experiment. The availability of a vast library of royalty-free video content is what makes it possible to generate the vast quantity of fMRI data needed to create a statistical model this complex.)