Reactive machine
The most basic but yet quite useful artificial intelligence, as its name suggests, is termed reactive AI. An example of this type of machine is Deep Blue, IBM's chess-play supercomputer, which in the late 1990's beat international Grandmaster Garry Kasparov. Reactive AI works how the input is programmed with a predictable output. These reactive devices will always react the same to identical situations.
Limited Memory
The next most advanced AI is known as AI limited memory. It is widely being perfected in the present day. It operated based on the experiences, in the same way as human newtons are fixed. It is based on its experience the same as the way the human neurons link. This is characterized by the capacity to absorb learning data and develop it over time. It can be useful for establishing trustworthy AI systems.
Theory of mind
Let’s just stop here to differentiate between the machines we have already created and what we are about to build in the future. The next, more advanced class of machines will not only represent the world but also other agents and entities throughout the planet. This is called the "theory of mind" in psychology—the notion that persons, beings, and items in the world can have thinking and emotional impacts on their behavior.
Self awareness
The last step in AI development is to design systems that can be self-representing. The AI researchers will create machines that not only need to understand consciousness but will also have to construct computers to do so. For example, “I want a medicine” is different from “I know I want a medicine.” Humans are aware of their consciousness. But machines are yet to be. Well, this is something time will tell us. We have to wait till then.
