

AI as a technology is projected to move in three phases. This first one is called Artificial Narrow Intelligence. This article presents the concept of this stage and explains what you can expect of the technology at this level.
The fact that AI in the first stage is “Narrow” in its intelligence definition means that technology has mastered few intelligence dimensions.
For example, your GPS is better than you to navigate. Your calculator is better at math than you are, etc. However, AI in this type of Artificial Intelligence class does not master all the dimensions of intelligence that a human does.
The limitations of the technology
It is this limitation that will keep the technology from being able to take over and automate many types of human jobs and tasks in principle. Because the technology is too narrow in its intellectual capacity, but even here, there will be some niches that technology will master better than humans.
The more dimensions of intelligence that we include – or the more capabilities that are equal to human capacity, the closer the machines come to be able to perform tasks that would otherwise only be performed by humans.
Layers of human-like traits
It is not enough that AI is good at counting; AI must also be able to function in a social context, be able to understand emotions etc. In other words, build layers of human-like traits in the way the AI works.
If you succeed in developing AI in that direction, then AI will at some point catch up on humans.
The consequence will be that we will develop an AI that will be able to match human intelligence. This stage is called Artificial General Intelligence.

The fact that AI in the first stage is “Narrow” in its intelligence definition means that technology has mastered few intelligence dimensions.
For example, your GPS is better than you to navigate. Your calculator is better at math than you are, etc. However, AI in this type of Artificial Intelligence class does not master all the dimensions of intelligence that a human does.
It is this limitation that will keep the technology from being able to take over and automate many types of human jobs and tasks. Although the technology is too narrow in its intellectual capacity, there are some niches that technology will master better than humans.
Take the example of the Google booking bot that could book meetings. It is a case where Artificial Narrow Intelligence outperforms people, processes, and IT systems in a particular task.
The more dimensions of intelligence that we include – or the more capabilities that are equal to human capacity- the closer the machines come to be able to perform tasks that would otherwise only be performed by humans.
It is not enough that AI is good at counting; AI must also be able to function in a social context, be able to understand emotions, etc. In other words, build layers of human-like traits in the way the AI works.
If you succeed in developing AI in that direction, then AI will at some point catch up with humans.
The consequence will be that we will develop an AI that will be able to match human intelligence. This stage is called Artificial General Intelligence.
The three concepts of intelligence are explained by Max Tegmark in the book Life 3.0

More from the AI Wiki

