It seems to me that labs at Berkeley always gets a pass when it comes to vastly overstating the significance of their work. This is so far from actually reading the visual data from the brain that even mentioning the future of reading memories seems ludicrous.