Logic Theorist facts for kids

The Logic Theorist was a special computer program created in 1956 by three smart people: Allen Newell, Herbert A. Simon, and Cliff Shaw. It was the very first program designed to think and solve problems like a human. Because of this, many people call it the "first artificial intelligence program."

The Logic Theorist was able to prove 38 out of 52 math problems from a famous book called Principia Mathematica by Alfred North Whitehead and Bertrand Russell. For some of these problems, the program even found new and shorter ways to prove them!

How It Started

In 1955, when Newell and Simon began working on the Logic Theorist, the idea of "artificial intelligence" (AI) didn't even have a name yet. The term "artificial intelligence" was actually created the next year.

Herbert A. Simon was a political scientist. This means he studied how governments and groups of people make decisions. He had already done important work on how large organizations, like government offices, work. He also developed a theory called "bounded rationality", which explains that people make decisions based on the limited information they have. He later won a Nobel Prize for this idea. Studying how businesses work, like AI, needs you to understand how humans solve problems and make choices.

Simon remembered seeing a printer at RAND Corporation in the early 1950s. It was printing a map using regular letters and symbols. He realized that if a machine could move symbols around, it could also pretend to make decisions, and maybe even think like a human.

The program that printed the map was written by Allen Newell. He was a scientist at RAND who studied how things are organized. For Newell, a big moment happened in 1954. Another scientist, Oliver Selfridge, came to RAND to talk about how computers could find patterns. As Newell watched, he suddenly understood how simple computer steps could lead to very complex actions, even intelligent ones like humans do. He later said, "It all happened in one afternoon." It was a rare moment where everything just clicked for him.

Newell and Simon started talking about teaching machines to think. Their first idea was a program that could prove math theorems. These were like the ones in Russell and Whitehead's Principia Mathematica. They asked Cliff Shaw, another computer programmer from RAND, to help them build the program. Newell said that Cliff was the "real computer scientist" of the three.

The very first version of the Logic Theorist wasn't run on a computer. They acted it out by hand! Simon remembered:

In January 1956, we assembled my wife and three children together with some graduate students. To each member of the group, we gave one of the cards, so that each one became, in effect, a component of the computer program ... Here was nature imitating art imitating nature.

They successfully showed that the program could prove theorems just as well as a skilled mathematician. Later, Cliff Shaw was able to run the program on a real computer at RAND's Santa Monica office.

In the summer of 1956, several important scientists, including John McCarthy and Marvin Minsky, held a conference. They called it the "artificial intelligence" conference. Newell and Simon proudly showed off the Logic Theorist there. But it wasn't met with much excitement. One writer, Pamela McCorduck, said that "nobody except Newell and Simon themselves understood how important what they were doing was." Simon felt that "we were probably fairly arrogant about it all" and added:

They didn't want to hear from us, and we sure didn't want to hear from them: we had something to show them! ... In a way it was ironic because we already had done the first example of what they were after; and second, they didn't pay much attention to it.

Even with the quiet start, the Logic Theorist soon proved 38 of the first 52 theorems in chapter 2 of the Principia Mathematica. The proof for one theorem (2.85) was even better than the one Russell and Whitehead had worked so hard to find by hand. Simon showed this new proof to Russell himself, who was "delighted." They tried to publish this new proof in a math journal, but it was turned down. The journal said that a new proof of a simple math problem wasn't important enough. They seemed to miss the fact that one of the "authors" was a computer program!

Newell and Simon continued to work together for many years. They started one of the first AI labs at the Carnegie Institute of Technology. They created many other important AI programs and ideas, like the General Problem Solver and Soar.

How Logic Theorist Worked

The Logic Theorist was a program that performed logical steps on logical ideas, which they called "expressions."

Expressions

• An expression is made of elements.
• The program used two types of memory: working and storage.
• Working memory held small parts of an idea. The Logic Theorist usually used 1 to 3 working memories.
• Storage memory held full ideas or lists of elements. This included all the basic rules (axioms) and math problems that had already been proven.
• An expression was like a tree diagram, where each part of the idea was a "node" with different details.

For example, a math idea like was shown as a tree. The main part was the symbol. This main part pointed to the smaller ideas like and .

Processes

The program used four main types of processes, from simple to complex:

• Instructions: These were like very basic computer commands. They could do simple things, like moving a part of an idea from one memory spot to another.
• Elementary Processes: These were like small mini-programs made of several instructions.
• Methods: These were sequences of elementary processes. There were four main methods:
  • Substitution: This method tried to change an idea to match a known rule or proven problem by swapping out parts.
  • Detachment: If the program wanted to prove idea B, it looked for a known rule like "If A, then B'". If it found one, it then tried to prove A.
  • Chaining Forward: If the program wanted to prove "If A, then C", it looked for a known rule like "If A, then B". If it found one, it then tried to prove "If B, then C".
  • Chaining Backward: If the program wanted to prove "If A, then C", it looked for a known rule like "If B, then C". If it found one, it then tried to prove "If A, then B".
• Executive Control Method: This was the highest level. It decided which of the four methods to use and in what order to try and prove each new problem.

Logic Theorist's Impact on AI

The Logic Theorist introduced several big ideas that became very important in artificial intelligence research:

• Thinking as Searching: The Logic Theorist explored a "search tree". Imagine a tree where the starting point is the problem you want to solve. Each branch is a step you can take using logic rules. Somewhere in the tree is the answer you're looking for. The path you take to get to the answer is the "proof"—a series of logical steps from the problem to the solution.
• Heuristics (Rules of Thumb): Newell and Simon realized that the "search tree" could grow incredibly large, like a tree with endless branches. They needed ways to "trim" some branches that probably wouldn't lead to a solution. They used "rules of thumb" to decide which paths to follow. They called these special rules "heuristics". This word came from a book by George Pólya about solving math problems. Heuristics became a very important part of AI, helping programs deal with problems that have too many possible solutions to check them all.
• List Processing: To make the Logic Theorist work on a computer, the three researchers created a special programming language called IPL. This language used a way of organizing information called "symbolic list processing." This same idea later became the basis for another important AI programming language called Lisp, which is still used by AI researchers today.

Big Ideas from Logic Theorist

Pamela McCorduck wrote that the Logic Theorist was "clear proof that a machine could do tasks that were thought to be intelligent, creative, and only done by humans." Because of this, it was a huge step in the development of artificial intelligence and in understanding what intelligence really is.

Simon once told his students in January 1956, "Over Christmas, Al Newell and I invented a thinking machine." He also wrote:

[We] invented a computer program capable of thinking non-numerically, and thereby solved the venerable mind-body problem, explaining how a system composed of matter can have the properties of mind.

This idea, that machines could have minds just like people, was later called "Strong AI" by a philosopher named John Searle. It's still a big topic of discussion today.

Pamela McCorduck also saw the Logic Theorist as the start of a new way to understand the mind. This was called the "information processing model" (sometimes called computationalism or cognitivism). She wrote that "this view would come to be central to their later work, and in their opinion, as central to understanding mind in the 20th century as Darwin's principle of natural selection had been to understanding biology in the nineteenth century." Newell and Simon later put this idea into a more formal theory called the physical symbol systems hypothesis.