In the 1950's - same era as Watson and Crick, Nicholas Rashevsky [1] invented Relational Biology in which he was looking for the mathematics of life itself. His student Robert Rosen [2] went on to pen a book "Life itself..." in 1989, but not until after he had penned a different book "Anticipatory Systems".
While looking for the mathematics of life (living things, not machines), several points came to life, one of which is that while machines have the luxury of plan s and makers,living things don't. How to resolve the difference led to the thesis that living things are anticipatory - they are driven by predictive "models".
The low level paramecium is able to swim towards "food" by using chemical sensors and measuring gradients; of course, that creature doesn't have anything like a human brain (the topic of this thread), so its models are more innate internal structures; humans have those too, but over time, the adult brain emerges as the supreme modelling agent. It is anticipatory.
What does anticipatory mean?
A child lets go of a heavy object and it smashes the child's toe. Next time, the child won't let go of heavy objects because it learned an important correlation (causality isn't yet in play). Correlation is sufficient to teach that child a new anticipatory rule; simplest form: dropping heavy stuff --> pain.
That applies to reading and human prediction machines. Just try learning a foreign language by reading children's books - the pictures really help. Over time, you don't need them anymore. @hunta20097 on this page says this with different words, but the game is the same: model building and refinement.
There are books 'out there' which try to explain all of this in Bayesian terms, that we are always "updating our priors" as we encounter new experiences. It's way above my pay grade to know if that's right, but it still offers satisfying explanations.
>Just try learning a foreign language by reading children's books
People often trot this strategy out like it's a good idea because children's books are "simple". Have you read children's books? The language in them is actually typically very weird and playful, the situations presented are often strange and non-sensical, and to that end they're not actually all that useful to a foreign language learner as a good starting point at all.
Perhaps. But, as evidence, I have a Chinese friend who bootstrapped his way to learn German when he became an exchange student in Germany, then, when he was doing his residency after graduating med school, he did the same with Spanish so he could treat his Hispanic patients in their language. He used children's books to kickstart, then graduated to advanced books, always with a dictionary at hand.
While looking for the mathematics of life (living things, not machines), several points came to life, one of which is that while machines have the luxury of plan s and makers,living things don't. How to resolve the difference led to the thesis that living things are anticipatory - they are driven by predictive "models".
The low level paramecium is able to swim towards "food" by using chemical sensors and measuring gradients; of course, that creature doesn't have anything like a human brain (the topic of this thread), so its models are more innate internal structures; humans have those too, but over time, the adult brain emerges as the supreme modelling agent. It is anticipatory.
What does anticipatory mean? A child lets go of a heavy object and it smashes the child's toe. Next time, the child won't let go of heavy objects because it learned an important correlation (causality isn't yet in play). Correlation is sufficient to teach that child a new anticipatory rule; simplest form: dropping heavy stuff --> pain.
That applies to reading and human prediction machines. Just try learning a foreign language by reading children's books - the pictures really help. Over time, you don't need them anymore. @hunta20097 on this page says this with different words, but the game is the same: model building and refinement.
There are books 'out there' which try to explain all of this in Bayesian terms, that we are always "updating our priors" as we encounter new experiences. It's way above my pay grade to know if that's right, but it still offers satisfying explanations.
[1] https://en.wikipedia.org/wiki/Nicolas_Rashevsky [2] https://en.wikipedia.org/wiki/Robert_Rosen_(biologist)