I agree with some of your points—well, all of them if we’re discussing control systems in general—but a couple of them don’t quite apply to brains, as the cortical systems of brains in general (not just in humans) do use predictive models in order to implement both perception and behavior. Humans at least can also run those models forward and backward for planning and behavior generation.
The other point, about actions determining perceptions, is “sorta” true of brains, in that eye saccades are a good example of that concept. However, not all perception is like that; frogs for example don’t move their eyes, but rely on external object movement for most of their sight.
So I think it’d be more accurate to say that where brains and nervous systems are concerned, there’s a continuous feedback loop between actions, perceptions, and models. That is, models drive actions, actions generate raw data that’s filtered through a model to become a perception, that may update one or more models.
Apart from that though, I’d say that your other three points apply to people and animals quite well.
I agree with some of your points—well, all of them if we’re discussing control systems in general—but a couple of them don’t quite apply to brains, as the cortical systems of brains in general (not just in humans) do use predictive models in order to implement both perception and behavior. Humans at least can also run those models forward and backward for planning and behavior generation.
The other point, about actions determining perceptions, is “sorta” true of brains, in that eye saccades are a good example of that concept. However, not all perception is like that; frogs for example don’t move their eyes, but rely on external object movement for most of their sight.
So I think it’d be more accurate to say that where brains and nervous systems are concerned, there’s a continuous feedback loop between actions, perceptions, and models. That is, models drive actions, actions generate raw data that’s filtered through a model to become a perception, that may update one or more models.
Apart from that though, I’d say that your other three points apply to people and animals quite well.