Artificial intelligence 33

Artificial intelligence systems can learn from the experience of others and learn how to become more intelligent themselves. But there are some inherent limits to their ability to "learn," namely the limitations of the human mind that are the basis for cognitive evolution.

In order to "artificially learn" our way of thinking and behaving, we need a toolbox. We start out with what may be an imperfect, limited, limited toolbox—a human brain!—and we attempt to learn and use it.

But as the AI research community gets deeper and deeper into the neural networks and neural networks models that we've developed over the last decade, it has been discovering that in particular, these network models can learn from a series of observations with which they are already familiar. These models are also able to "know" that we don't observe how they do this. As the human brain is limited, our models must be too, we are learning and utilizing these concepts we cannot observe.

No comments:

Post a Comment