Why AI could not take the place of humans?

Jul 08,2021

Artificial Intelligence (AI) has no sense of concept, no values which could only be a human tool, and can not be beyond human finally.


In recent years, the AI technology develops extremely rapidly, each kind of intelligent equipment & the intelligent software has entered each family, has changed our life style and the working way. Therefore, many people believe that AI will completely replace human intelligence, even beyond human intelligence in the near future. However, this view is too pessimistic. The most important and unique part of human thought and behavior can not be achieved and replaced by AI.


AI, no matter how advanced it is, comes down to solving problems within a given human framework. For example, if someone commutes to work every day, the distance between the office and home is 15 kilometers. His options include taxis, buses, subways, self-driving cars, shared bikes, and a combination of these. If he asks the navigation software, it can give him an optimal travel plan based on his requirements and real-time traffic conditions. This is often useful in the real world. However, while many people will choose to buy or rent a house near their workplace to solve their commute problems, the navigation software will never suggest a move, because the navigation software runs the program, or run the framework has no such means to choose from, but people are not bound by the existing framework.



AI is also unable to proactively identify problems that need to be solved. To take a very simple example, if I ask the smart answer software: "Where is Kiki, did you see Kiki?" It either answers "NO" or gives a wrong answer. And if I take this question to ask people, the first reaction of people who are asked, regardless of their level of knowledge, is probably: "Who is Kiki?" I will tell him, handsome is my puppy, what is it like, how big, and what are the characteristics, and so on. Obviously, the Human Beings can firstly take the initiative to determine what is the problem to be solved, that is to say, determine the goal. The example I've given is simple enough that AI might be able to handle it, but it,s not that the software has learned how to identify problems, it's that the designers have expanded or changed the programs or frameworks that make the software work.

In short, if human beings identify the problem, the means or information available, AI can give the answer, even near perfect answer. But AI is not going to design a framework for this purpose or break through it.

Human beings are intelligent in pursuit of meaning and therefore they have their own values. Only when human beings endow some things with meaning or value can they form the logical relation of the means of purpose. In other words, people could know what they want and how to achieve it. The notion that AI has no meaning requires humans to encode meaningful logical relationships in order for AI to work on those relationships, but it can't build those relationships on its own.

People understand the world by its meaning and communicate and cooperate with others by it.  Human beings have similar mental structures and can understand each other even if they do not speak the same language. What if it was a robot? Unless it's an actor playing a robot in a science movie, it searches and outputs code without really trying to understand what's going on.


Obviously, the biggest difference between man and AI is that man connects with the outside world through meaning and value. This is the basis of human being not only as subject but also as object. It is also the basis of human cooperation and innovation. AI has no sense of concept, no values, in the end can only be a human tool, and can not be beyond human.


It is noteworthy that we can see that the rational economic man hypothesis of the mainstream neoclassical economics treats man as a machine seeking optimal solution under the framework of a given purpose. This actually omits the most essential and innovative part of human behavior. Rational economic man constitutes an economy that is static, mechanical, and passive. At best, it is an AI world, not a human society.


Economics is a value-neutral science, and therefore it does not study and does not have sufficient capacity to study what specific goals and means of someone, but economics should bring the behavior pattern of the individual in the means framework of the goal into the research category, otherwise it has the defect which is difficult to overcome.



Contact us