We are currently at a wonderful juncture of time where we are implementing data based learning in our machines, leveraging our understanding on how humans learn and undertake different tasks. Probably, there is a thing or two we humans can remind ourself as well from the data based learning paradigm.
With the retention of knowledge in the form of books and language, the last 1000 years have remarkably expedited the human learning evolution. It is believed that a person is born with the knowledge of human evolution worth billions of years in a concise representation worth 50 harry potter novel books. It is a miracle, how so much information is stored in our body in such a concise format in the form of genetics. Though one can argue that majority of tasks like pattern recognition, eating, walking etc. a new born learns from trying out different actions, however, there is a 'big intuition' that a new born is born with, from which it knows about the action of crying for food, following a particular breathing pattern, and above all to be be a reasonable agent to express itself so that other humans can identify if its a red flag or a green flag situation. These axiomatic tasks, are probably extracted from the genetics carried by the baby, transferred from the parents to the baby.
Here let us assume that our brain consist a logical block that we call candidly as a model and given a particular situation, the model tells us what would be the best action possible so as to get a desired outcome. I believe there is a collection of models that we are born with. Each model corresponds to different topics (activities) and we end up tuning these collections of models based on our lifetime experiences. The starting models would be based on your family/ancestral experiences and are transferred from generations based on their importance in the latest times. The broad classes of different topics may correspond to reasoning, physical activity, relationships etc. Though these classes may seem independent but updating the model of one class could after all result in updating models in other classes. We tend to use the primitive model that we inherit if we have not encountered a type of situation before. Problem with that is given the limited number of experiences and ever evolving social and physical life on this planet, one may come up with an action that is fatal given the current setting. For example, if you do not know what to do when you encounter a tiger in the jungle (maybe run for your life ?), it might cost you your life. A social setting example would be, if you have not lived a particular kind of relationship, some early actions which maybe biased by the primitive model may prove fatal for the relationship.
Loosely speaking we grow wiser with age based on the assumption of more experiences to learn from, for each of the model topics. However, a young person can outperform the elder person if he/she has more exposure in the corresponding situation. With lack of sample points to learn from, there is the risk of using a primitive model which could be catastrophic for a human. However, luckily we as humans have the power to tune our models right ! Given the four dimensions of space and time, we can explore the dimension of space or travel to collect more sample points to perform satisfyingly well in different facets of life. We can additionally download/add models collected from communication and knowledge resources of books and other media. Most importantly, even with little data points corresponding to a particular task, our attention/focus to the right samples (maybe using techniques like meditation or procrastination ) could help us retune our models to give the intended results based on the situation. Theoretically speaking, given the right weights, even one successful experience (and other failures) can help you to tune the model right , and even one mistake can tune your model to avoid it in the future. Thus, we can be the best version of ourselves focussing on our right past learning samples. To continue to be the most evolved species of life known to us, it is essential to tune our models constantly, and consciously choose the right weights for the different experiences to tune our models right.