In 1975 Allen Newell and Herbert A. Simon initially proposed the physical symbol system hypothesis (PSSH). Their Turing Award-winning paper states that "a physical symbol system possesses the necessary and sufficient means for intelligent action."

This assertion argues that human thought is a form of symbol manipulation and, therefore, intelligent computers are possible.

Ideological breakthrough

The concept is back to 

• Hobbes (who claimed that reasoning was "nothing more than reckoning"),
• Leibniz (who attempted to create a logical calculus of all human ideas), 
• Hume (who believed PSSH could reduce perception to "atomic impressions"), and
• Kant (who thought that PSSH could reduce perception to "atomic impressions"). 
• The computational theory of mind, developed by philosophers Hilary Putnam and Jerry Fodor, is the most recent form.

Many people have strongly criticized the hypothesis, but it is still essential to AI research. One common criticism is that the idea seems to work for higher-level intelligence, like playing chess, but not so well for everyday intelligence, like being able to see. Most of the time, there is a difference between high-level symbols that directly represent things in the real world, like dog and tail and the more complicated "symbols" in a machine like a neural network.

What does the PSSH mean?

The authors think that a physical symbol system has everything needed and enough to make intelligent actions in general. What they mean is that. 

(1) a physical symbol system can be shown as any system that acts intelligently. Also, 

(2) if the physical symbols can show intelligence in the same way that human actions do.

What do people have to do with testing this theory?

Some modern models of human intelligence are like a physical symbol system explaining how people think. Researchers in "information processing," "cognitive science," "cognitive psychology," and other fields can learn a lot by using the "computer" or "information processing" metaphor. Symbol systems are used in several theories to explain how people act. This concept has significantly affected AI research since AI researchers want to use the same processes that cognitive psychologists find and define. 

How can AI be used to test this theory?

The physical symbol system hypothesis claims that developments in computer science strongly support it. The field of AI can build computer systems that are as smart as humans. There are a lot of examples of these kinds of systems in use today, and in theory, PSSH could copy any intelligent action. Several successful AI programs back up the idea. We don't know if the Physical symbol system model has any restrictions that make it impossible for the system to show specific kinds of "intelligent behaviour." Such evidence would indicate that the hypothesis is wrong.

Examples

Here are some examples of physical symbol systems:

• Formal logic: It uses words like "and," "or," "not," "for all x," and so on as symbols. The expressions are formal logic statements that can be either true or false. The rules of logical deduction are the steps in the process.
• Algebra: In algebra, the symbols are "+," "," "x," "y," "1," "2," "3," and so on. Equations are what the expressions are. Algebra lets you change a mathematical expression without changing its truth.
• In a digital computer: the symbols are the ones and zeros that make up the memory, and the processes are the things that the CPU does that change the memory.
• Chess: the symbols are the pieces, the processes are the legal moves, and the expressions are where all the parts are on the board.

According to the physical symbol system hypothesis:

• Smart people think that the symbols are in our brains. The words are like thoughts. The processes are the things that happen in your mind when you think.
• In an AI program that is running, the symbols are the data. The words are more information. The programs that change the data are called "processes."