It's one of the things that seems extremely weird if you consider humans to be what they say they are ... and dead normal if you realize what humans actually ar

It's one of the things that seems extremely weird if you consider humans to be what they say they are ... and dead normal if you realize what humans actually are. How memory works, according to humanity : we store all data input in our mind, where it stays. Every now and again some data disappears or is overwritten. 5 seconds of thinking about how a mind looks and what it does will leave you with 100 very hard questions about how you could possibly make that happen in a neural network. How memory works, the real version : you "learn" memory. When a person is presented with a pointer to a specific point in time or a specific event, his mind then reconstructs the memory by repeatedly asking the question "what happened next ?". All the data produced comes from the predictor, not from actual memory. So what is actually stored in your memory ? Well useful things. Not an index of dates combined with what happened. But "I start my day, then wash my teeth, then make a piece of toast, then ..." "What did you do yesterday ?" ("conversation" in someone's mind starts happening : "well obviously I woke up. Then what ? I washed my teeth. Ok, then what ? interrogator interrupts You ate a biscuit your mind Ok that sounds reasonable, and I'm not all that sure at all. He may be right. I'll go with it) Memory changed. Next time you remember it will be a lot easier to go with the changed memory. When you remember something, the memory is not "recalled" like on a computer, but recreated from scratch. You start with a specific event (that itself may or may not have happened, but it must be quite normal for it to happen so the predictor doesn't see it as too unlikely). Because external persons can both contribute and interfere with the recreation process it can be used to change memories (your mind doesn't separate multiple inputs, so if a person is talking when you're recalling, it influences the recall process). You want your memory to be accurate, spend a while with extremely limited inputs first. Because of how the predictor works, it "synchronizes" with other people's predictors (nothing magical, it simply predicts what the other person's mind is doing and takes that into account). But that results in that if you have extreme uncertainty about an event and the person you're talking to feels very strongly about it, your mind will be quite likely to follow the other person's convictions. Keep doing this, and you can convince someone they're a monster (note: requires that the interrogator believes that to be true, builds trust and avoids firing the "bullshit detector" and even then large changes require quite a bit of time) There are a few more of these things that are quite normal, given that our minds copy others' behaviors, but people find extremely hard to believe. The one I have refused to believe for the longest time was that people who get abused, whether by a military, simply be economic situation, or even when kidnapped, initially want to get out. But if the situation persists, the no longer want the abuse to stop. They want to become the abuser (because that person is the person their mind most often gives attention to). For a behavior copy "machine", this is expected and normal. In psychology this simple fact has been known for centuries. Many other counterintuitive facts become easy to explain (e.g. why commercials work).