What are the differences between artificial intelligence and ordinary software? How do intelligent robots work and how exceeds human intelligence? Humans are the smartest creatures we know and artificial intelligence imitates human intelligence.
However, in turn, artificial intelligence (AI) is a large area of research within computer science. The goal of the AI area is to create intelligent systems that operate independently of human beings.
WHAT DOES ARTIFICIAL INTELLIGENCE IMITATE?
For example, human beings communicate with language, which means that the speech recognition area in AI. In addition, since artificial intelligence generally does this statistically, we call speech recognition as statistical learning. In addition, human beings can write and read in a specific language. To emulate it comes into the field of NLP, that is, natural language processing.
Of course, human beings are not limited to cognitive abilities. For example, they see the world with their eyes and make sense of what they see in the brain. And that’s in the field of computer vision. Computers use computer vision to symbolically process visual data. However, in recent years a deep learning technique has been developed to do this.
ENVIRONMENT AND OBJECT RECOGNITION
Human beings can see the environment in their eyes and envision the world in their minds. This field is called as image processing. Ultimately, image processing is a must for computer vision, and biggest sites on the planet such as Facebook and Amazon use a very advanced version of this technique: Object recognition.
With the object recognition feature, machines can understand that the group in a photograph is a family of mothers, fathers, children, and cats; they realize which is female and which is male. They measure their height, take their size and realize what color they are wearing.
They can understand what they are wearing, as well as distinguish the forested area behind. In short, computer vision and image processing are also used to understand which products customers buy and review.
PERCEIVING THE WORLD
Human beings can move around easily, for example, without stumbling or hitting objects. Even if we don’t think of it at first glance, this power enters the field of robotics.
Robots are machines that imitate human movements, and robots don’t have to be smart. Except for Boston Dynamics’ Spot Mini robot dog or Atlas robot athlete, the robotic arms used in automotive production lines are not smart.
Human beings can also recognize and classify objects. They can distinguish apples from pears and red balls from blue balls; they can even understand letters, words, words, and meanings. In the AI we call it pattern recognition.
Machines are much better at recognizing patterns than human beings; because
- They can collect more data in less time.
- They can process the collected data in a multi-dimensional virtual math space.
- They have a much higher processing power than human beings.
For now, the most ambitious field of artificial intelligence is pattern recognition. Google’s Alpha Go that defeats world-go champion using a pattern recognition technique with deep learning.
The human brain is an organic network of neurons and learns new things with this network intelligence. So if we manage to imitate, copy, derive and simulate the structure and functions of the brain, we can also give machines cognitive abilities.
This is what we mean by robots that think like human beings and are as intelligent as human beings. In AI, it calls as neural networks. Artificial neural networks are the most complex and deepest learning system of machine learning. That’s why we call this technique deep learning.
In this context, there are multiple deep learning techniques. These are different types of deep learning that imitate the ability of the human brain to learn new things in different ways. There are even deep learning techniques that tell them how and according to what do they make decisions. In this way that they can carry both legal and financial responsibility.
MEMORIES AND EXPERIENCES
Human beings can remember the past, for example, what they ate last night. In the field of artificial intelligence, we can establish an artificial neural network system that can remember the past in a limited way. We call this a recurrent neural network (RNN).
In short, artificial intelligence works in two ways: symbolically (visual recognition) and data-driven (machine learning, pattern recognition, etc.). Robot science and image recognition are associated with symbolic learning. Statistical learning (IBM Watson) and deep learning, as well as machine learning, are associated with data-processing.
WHAT ABOUT PATTERN RECOGNITION?
Pattern recognition uses both symbolic computing (image recognition) and data-driven computing (for example, recognizing letters from lines, recognizing objects in the picture, recognizing faces, recognizing the relationships between letters, reading words and sentences, even associative-conceptual thinking)
The human brain essentially recognizes patterns; thinking, associative, contextual, conceptual and symbolic. Thus, it builds systemic structures such as institutional religions, philosophical views, and states by taking advantage of creativity, inspiration and intuitive thinking.
HOWEVER, IT IS DIFFICULT TO DO FOR AI
In fact, when it comes to machine learning, we have to feed artificial intelligence with big data, very big data. For example, IBM Watson needs to read millions of pages of patient records and research articles to make indirect recommendations to doctors in cancer research.
ADVERTISING AND MARKETING
At its simplest, if you have millions of data points that show the relationship between sales and ads, you can pattern them (patterns) by patterning them. If a computer learns this pattern with the help of artificial intelligence, it can make predictions that sales will increase based on what they have learned.
We have seen what artificial intelligence is, the difference between AI and ordinary software, and how AI imitates the most basic features of the human brain, such as pattern recognition and learning.
Artificial intelligence seems to be extremely useful and life-facilitating, but names like Elon Musk and Stephen Hawking are very pessimistic about the future of artificial intelligence. We are all curious about how artificial intelligence will develop.