Can Ai Learn …?
Can Ai Learn …? One of the most common questions we get at Rebellion Research is whether or not Artificial Intelligence can learn?
Firstly, we have to address the fact that yes, some Ai can learn while other Ai systems can NOT learn. In addition, what makes a system artificially intelligent in the first place?
To be a “piece of ai” a system needs to be able to follow an intelligent set of commands or tasks. From dishwashers to washing machines, these household appliances are all examples of true Artificial Intelligence.
Now, the next step of Artificial Intelligence that is preoccupied with learning is defined as Machine Learning.
Moreover, the difference between all of the various forms of Machine Learning is immense. Firstly, there are machine learning systems that can give specfic answers vs other Machine Learning systems that in place of a binary answer will give a probability or probabilistic outcome. Furthermore, this form of Machine Learning Artificial Intelligence did not even exist until 1990 when acclaimed UCLA professor Judea Pearl invented Probabilistic Learning.
In addition, Professor Pearl has actually shifted his thinking in recent years away from probabilistic learning. Recently, Professor Pearl has focused on causal learning, or learning focused on causality. The general idea being that multiple semi-relevant pieces of data that create a web of a probability will be more consistent than a Machine Learning that tries to predict without causality in its system.
Causality is still a new family or segment of Machine Learning research so the jury is still out on how effective it is or how wide its applicability will be beyond academia and corporate research.
In conclusion, yes some Ai or Machine Learning Ai can learn!