This is a really interesting perspective on cognition, and it all kind of makes sense if you consider the brain to be a black box pattern recognition machine with various built-in biases.
New data is always added to the model, but not in an entirely rational fashion. The updated model is likely to slightly overfit new data ("compiled at the time they happen"), and particularly salient bits and pieces of old data (see https://en.wikipedia.org/wiki/List_of_cognitive_biases) are disproportionately weighted.
New data is always added to the model, but not in an entirely rational fashion. The updated model is likely to slightly overfit new data ("compiled at the time they happen"), and particularly salient bits and pieces of old data (see https://en.wikipedia.org/wiki/List_of_cognitive_biases) are disproportionately weighted.