Fascinating report involving the use of machine learning in dream decoding! The algorithm was trained by watching brain activity while hundreds of images were shown to subjects. Then brain activity during dreaming was input for the same subjects. It’s still VERY primitive, but the computer predict that the subject had dreamed about a car, not a chair. It only worked if the computer was given two choices, but still that’s rather impressive.
Two things I get from this, we are one step closer to ‘Minority Report’ and I need to finally get around to watching ‘Inception’.