Is Artificial Intelligence exist in 1950’ - 8 real-world applications of AI in 1950's?

Why ancient Artificial Intelligence is the base of Modern Artificial Intelligence?

People think that artificial intelligence is a new emerging technology that is just born a few years ago. But actual artificial intelligence has very strong roots in reality. In ancient times also, the researcher worked with techniques like Neural networks and machine intelligence. The basic difference between today’s technologies and older methods only in terms of computation and data. Today we have a huge amount of computation power that can work with any amount of data. In older days we just have few kilobytes of RAM with small CPU power which can only do very basic operations. But today’s we have graphical processing units with a tremendous amount of RAM and processing powers to do high, exponential and complex calculations. Moreover, we do have innumerable data too. This data can help us to make better decisions with Artificial Intelligence. But today’s artificial intelligence is just an upgraded version from older days. Roots of AI started with the 1950s. Some of today’s trending applications started back in 1950. John McCarthy, who coined the term in 1956, defines Artificial Intelligence as “the science and engineering of making intelligent machines.”[1]

REVOLUTION OF ARTIFICIAL INTELLIGENCE WITH APPLICATIONS

1. The “Translation” memorandum

Warren Weaver[2] has the very first time introduced about translating the documents to natural language using computers. Warren had four proposals for the translation of the words-for-words method.

In the very first proposal, he handled the problem of word ambiguity. In English, there are lots of words that have different meanings in a different context. The same word can be treated with two or more meanings. He suggested a method to translate the word from a human language with occurrence in the vicinity and number of context words. [2] In the second proposal, he described conclusion based translation using neural networks based on logical context. The third proposal was to work with cryptographic translation with information theory. [3] the Fourth proposal said about the linguistic universals to make better language and conversations. [4]

2. Logic Theorist

The first demonstration of the Logic Theorist (LT) written by Allen Newell, J.C.Shaw and Herbert A. Simon. It was engineered to mimic the problem-solving techniques of a human being and is called “The first Artificial Intelligence Program”. It proved the 38 of the first 52 theorems in Whitehead and Russell’s Principia Mathematica.

The study of business organizations requires an insight into the nature of human problem solving and decision making.

Newell and Simon started to talk about the possibility of learning machines to think. Their first project was a program that could prove mathematical theorems. As they have implemented the program, that program succeeded to prove theorems, and that makes mathematicians shocked. Eventually, Shaw was able to complete the program and able to run the program on the machines at RAND’s Santa Monica facility.

1) Logic Theorist’s influence on AI:

• Reasoning as search: This theorem helped to explore a search tree: the root node was the initial hypothesis and each branch was a deduction based on the rules of a given logic.
• Heuristics: They realized that the search tree would grow exponentially and they need to trim some unnecessary branches, using “rules of thumb” to get the path that was likely to lead to a dead-end.
• List processing: To implement the Logic Theorist, three researchers developed a programming language, IPL, which used the same form of symbolic list processing that would later form the basis of McCarthy’s Lisp programming language.

2) Philosophical Implication:

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

3. Fuzzy Logic

Fuzzy logic is a many-valued logic form in which value should be any real number between 0 and 1 included.

People make the decision-based in non-numerical and im-precise information and using these logic Fuzzy logic is introduced. These models have the interpreting and utilizing data and information, the capability of recognizing, manipulating, representing that are vague and lack certainty.

Both degree and truth and probabilities range between 0 and 1 and hence may seem similar at first, but fuzzy logic uses degrees of truth as a mathematical model of vagueness, while probability is a mathematical model of ignorance.

1) Applications:

• The first notable application for Fuzzy Logic was implemented in Japan, and it was implemented on the subway train in Sendai, it helped to improve the economy, comfort, and precision of the ride.
• Fuzzy logic is a more important concept for medical decision making. One common application area is a computer-aided diagnosis that uses fuzzy logic.

4. Logical Games with Artificial Intelligence

Artificial Intelligence has always been great with learning by itself and played a tremendous job in logical games like Chess, Checkers, Othello and Go. In the year of 1952, Arthur Samuel [5] created the very first program to play checkers by self-learning artificial intelligence methods. After a few years of revolution in chess Deep Blue [6] was developed by International Business Machines(IBM) in 1997 which defeated the grandmaster Garry Kasparov. In the Man-Machine championship, Chinook a computer program became the world checkers champion in 1994. Logistello also computer program beat the world champion in 1997.

5. Automatic Differentiation

Automatic Differentiation is a set of techniques to evaluate the derivative of a function. AD exploits that every program has a sequence of arithmetic operations such as multiplication, division, addition, subtraction etc and elementary operations such as sin, cos, exp, log etc. By applying the chain rule, derivatives of sequence can be computed automatically, accurately to working precision.

Generally, there are two ways Forward accumulation and Reverse accumulation.

• Forward accumulation: First fixes the independent variable concerning which differentiation is performed and computes the derivatives of each sub-expression recursively.
• Reverse accumulation: The dependent variable to be differentiated is fixed and the derivative is computed concerning each sub-expression recursively.

Currently, these techniques are known as Forwarding propagation and Backpropagation

Implementation:

• Source Code Transformation: It could be implemented for all programming languages, and it makes easier for the compiler to compile and optimize time. However, the implementation of the AD tool itself is more difficult.
• Operator Overloading: Objects for real number and elementary mathematical operations must be overloaded to cater for the augmented arithmetic calculations

6. MYCIN

MYCIN is an expert system developed by Standford University to recommend antibiotic therapy in the year 1972. MYCIN diagnose the patient based on reported symptoms and medical test results [7]. It was a rule-based program that has a series of yes/no or textual questions and provides the result of bacteria ranked with probability. This has a base concept of Bayesian probability and conditional probability. It was written in LISP(list processing) programming language which was developed by John McCarthy in 1959 for the common use of artificial intelligence.

7. Polly

Polly was the first mobile robot that works advance technology with a computer vision of artificial intelligence. This robot works with behaviour-based robotics. It is developed by Ian Horswill in the MIT Artificial Intelligence Laboratory. The Polly algorithm is efficient enough which moves forward with pixel with very low resolutions.

8. Deep Blue

Deep Blue is the computer developed by IBM which is the first computer to beat grandmasters of chess Garry Kasparov in the championship. [8] This works with great strategies like a single-chip chess search engine, a massively parallel system with multiple levels of parallelism, a strong emphasis on search extensions, a complex evaluation function, and effective use of a Grandmaster game database. [9] Deep Blue implemented on Alpha-beta pruning is a search algorithm to decreases the number of nodes that are evaluated by the minimax algorithm in its search tree.

Conclusion

In a nutshell, Artificial Intelligence started a very long time ago but just becoming a mature enough technology nowadays. Today we know AI just because of more data and more power. Researchers indeed came with new and efficient algorithms to work faster. The artificial system, natural language processing, computer vision, recommender systems already exist but just made stronger roots in the last few years.

References :

[1] Science Daily, “Artificial intelligence,” Science Daily, [Online]. Available: https://www.sciencedaily.com/terms/artificial_intelligence.htm . [Accessed 20 Jan 2020].

[2] “Warren Weaver,” Wikipedia, 16-Oct-2019. [Online]. Available: https://en.wikipedia.org/wiki/Warren/_Weaver . [Accessed: 22-Jan- 2020].

[3] “Claude Shannon,” Wikipedia, 17-Jan-2020. [Online]. Available: https://en.m.wikipedia.org/wiki/Claude_Shannon . [Accessed: 22-Jan- 2020].

[4] E. Rich, “AILongForSeminar,” 04 October 2004. [Online]. Available: http://www.cs.utexas.edu/users/ear/AILongForSeminar.ppt . [Accessed 20 Janaury 2020].

[5] “Arthur Samuel,” Wikipedia, 07-Jan-2020. [Online]. Available: https://en.wikipedia.org/wiki/Arthur_Samuel . [Accessed: 22-Jan- 2020].

[6] “Deep Blue (chess computer),” Wikipedia, 18-Jan-2020. [Online]. Available: https://en.wikipedia.org/wiki/Deep_Blue_(chess_computer). [Accessed: 22-Jan-2020].

[7] B. J. Copeland, “MYCIN,” Encyclopædia Britannica, 21-Nov-2018. [Online]. Available: https://www.britannica.com/technology/MYCIN . [Accessed: 22-Jan-2020].

[8] “Deep Blue,” IBM100 — Deep Blue. [Online]. Available: https://www.ibm.com/ibm/history/ibm100/us/en/icons/deepblue/ . [Accessed: 22-Jan-2020].

[9] M. Campbell, A. J. Hoane, and F.-hsiung Hsu, “Deep Blue,” Artificial Intelligence, 09-Aug-2001. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0004370201001291 . [Accessed: 22-Jan-2020].

Thanks a lot Harmish Patel :sunglasses: