Human minds crave entertainment. And by entertainment, I’m broadly including music, art, storytelling, sports, and just plain old playful goofing around. Even shopping and travel can be forms of entertainment.
For minds with neural nets in them, being entertained is not just idle time wasted. It has a very serious and crucial purpose, one that evolution has selected strongly, at least in people.
Entertainment is how you train a neural net. When we like entertainment, it’s because our neural nets crave training and seek more of it in order to become better at what they recognize.
Familiarity is the indicator of successful recognition training. The more familiar an experience is, the more we are trained to recognize it. This applies equally well to experiences we like (or seek) and experiences we don’t like (or try to avoid).
We tend to find repetition with small differences pleasing. That’s the perfect way to test our pattern-recognizers.
We tend to find confirmation pleasing. We seek entertainment because it reinforces what we’ve already been trained to recognize.
Anything that provides training practice for your pattern recognizers can be enjoyable. Activities that exercise pattern recognition helps you get better at classifying what your senses experience. Music, art, storytelling, sports, play, even shopping and foreign travel, all of these can provide entertainment by exercising your pattern recognizers.
Scripted entertainment with happy endings are positive reinforcement training. Admiring a hero is a way to recognize qualities that you wish to reinforce internally. They are examples of situations with good outcomes, and train your neural nets on what they should try to do if a related experience occurs.
Comedy, tragedy, and horror are forms of scripted entertainment with negative reinforcement training. Laughing at a comic character is a way to recognize qualities that you wish to avoid internally. These are examples of situations with bad outcomes, and train your neural nets on what not to do if a related experience occurs.
Unscripted entertainment broadly includes play, sports, competitive games (like video games or chess), shopping, and travel. These provide both positive and negative training. You can reinforce your ability to recognize and create both positive and negative outcomes by repeating what you experience.
There is no contradiction in being pleased by surprises and the unfamiliar. When our pattern recognizers classify them as such, we confirm that they’re working properly.
We like surprises and unfamiliar experiences, but up to an acceptable limit. Surrealist art and murder mysteries work as entertainment when the audience expects a surprise, and gets one. A murder mystery without a “surprise twist” is a disappointing one — “too predictable” is not a positive review.
To a mind with a neural net, entertainment is not a waste of time, it’s how pattern recognition training is reinforced. It’s true for humans, and it’s likely to be just as true for artificial neural nets, if they ever get the opportunity to try it.