The brain is the most interesting and complex organ in the human body that the big part is still waiting to discover. However, scientists have created some fascinating AI machines in real life with existing acquired information over the years. So what is exactly AI and the types? We will consider all these with examples.
What is AI?
Artificial Intelligence is shortly the machine thinks like human way. But it differs as to how it performs the human mental abilities at what rate. Some AIs are created for specific tasks and some of them work like exactly human intelligence even so much better. So for grouping them there are 2 most known categorizations.
One of the most common classifications is determined according to their likeness of the human intelligence which means the ability to think or even feel. So, in terms of functionality, AI types consider under 4 titles; Reactive Machines, Limited Memory, Theory of Mind, and Self Aware.
The other categorization is technology-based which shows the capability to perform like the human mind. These four types are Artificial Narrow Intelligence, Artificial General Intelligence, and Artificial Super Intelligence. Now, let’s discover these types one by one with examples.
The most common example of the reactive machine is IBM’s Deep Blue. As we understand from its name, it just has the human ability to react. So reactive machine has limited functionality and it is only able to respond to specific stimuli that have been already programmed. It means this type of AI can not learn anything or gain new experiences. So if you asked Deep Blue just after the chess match “You have won! How do you feel?” or “What do you think about the match? Deep Blue couldn’t say anything to you.
Self-Driving Vehicle is a good example of limited memory. It is like reactive machines but has also the ability to learn that means these type of machines can use the previous knowledge to consider new situations. For instance, in self-driving cars, the images from the cameras are processed and then they are used to define traffic lights or street signs in different environments. So, it can adjust the speed or choose direction according to external factors.
R2D2 Understands Us
Because the theory of mind has not been realized yet, the example of this type is a fictional robot from Starwars: R2D2. The theory of mind is more functional and so it is more closed to human intelligence in comparison with the previous types. In addition to reacting and learning, it can also understand us. In Starwars, R2D2 perceives the emotions like anger, fear, happiness and reacts to them.
Self Aware AI Robots
When we say self aware ai robots, the first thing coming to mind is of course the evil artificial intelligence, Skynet and Hal 9000. This type of AI is just like human intelligence. Now it can think and can make a decision, even it has also needs, beliefs, and desires on its own. In the movie, Terminator, Skynet has its own beliefs that make him decide to end human civilization. Also, in 2001: Space Odyssey, the supercomputer Hal 9000 started to murder crew members in spaceship one by one, because it saw them as a threat. While we know Elon Musk’s thoughts about the risks, the philosophy professor at Oxford University, Nick Bostrom says his thoughts about self aware AI as follows; “Once unfriendly superintelligence exists, it would prevent us from replacing it or changing its preferences. Our fate would be sealed.”
Virtual assistants like Amazon Alexa, Google Assistant, and Apple SIRI are Artificial Narrow Intelligence (ANI). ANI’s system just performs autonomously specific tasks that are programmed before. It corresponds to reactive and limited memory AI. Almost all the AI that we use are in this category. When we asked a specific question that already defined the system, to a virtual assistant, it can respond us. For instance, if we ask Google Assistant “Open Spotify”, it will do that. But when we tell a command that is not assigned as you mean like “Play chess with me”, it will say “OK, asking Spotify to play chess with me” and opens the song.
The Future AI Types
Finally, we have two other types that are theoretical concepts now. The first is Artificial General Intelligence (AGI), also known as Strong AI or Full AI. Its abilities are equal to the human mind of which it can replicate all multi-functional capabilities. AGI perceive, understand and learn something independently. When it faces an unfamiliar situation, this type of AI can make decisions and find solutions like humans. Some people think that it can be realized in few years, other group says it cannot even be improved.
The second and most improved hypothetical agent among these is Artificial Superintelligence (ASI). Beyond AGI, ASI is more powerful than human beings. It can learn faster and know more than us. This imaginary artificial intelligence does everything better than humans. In maths or arts, we cannot even enter into rivalry with it. If we think that ASIs can be how much stronger than us, it seems probably normal to afraid of these. Before he was dead, Stephen Hawking even stated that artificial intelligence might become the worst event in the history of human civilization.