Why Artificial Intelligence is the New Black?12 min read
Do you know how far have we reached in the field of technological advancement? Who are intelligent agents? Do you know there’s a machine that can read human mind? Do you know that there are various machines who can successfully identify human speech? Do you know that they can perform reasoning, knowledge representation, planning, learning, natural language processing, perception and many more? Do you know that they significantly serving in the field of computer science, mathematics, psychological, logical, and many others?
They are able to perform and solve many of the human problems and can serve you till its malfunction.
But in the end, it’s major drawback that many of you might know is the risk of unemployment that it may create.
Today You’re going to have an absolute answer to all the questions and you’ll be able to know more about 21st century’s technological advancement, Artificial Intelligence (AI).
Back in the old days, the very first “Artificial Neurons” were first discovered by McCullough and Pitts in 1943.
Soon after that two undergraduates from Harward, Marvin Minsky and Dean Edmonds built the first Neuron based network computer in 1950, called SNARL. They used about 3000 vacuum tubes and a surplus automatic pilot mechanism from B24 bomber to stimulate a network of 40 neurons.
Alan Turing gave a lecture on the topic “Computing Machinery And Intelligence” at LONDON MATHEMATICAL SOCIETY explaining all about his persuasive agenda in 1950. That lead to the introduction of Turing Test, Machine Learning, Genetic Algorithm, and Reinforcement Learning.
In 1956, John McCarthy moved to Dartmouth College convincing Minsky, Cloud Shannon, Nathaniel Rochester about the same notion to bring forth the automata theory, neuron nets and the study of intelligence. They organized a workshop for the same at Dartmouth College.
In 1956, the researchers team consisting of 10 people indulged in an activity to introduce a new technology, Artificial Intelligence. The whole community worked together for about 2 months. Its aim is to build a machine that is able to incorporate different languages, form abstraction and concepts. And that’s how the birth of AI(Artificial Intelligence) took place. This was the year when the future of the human reborn as it successfully advanced the growth of human problem solving ability and helped in many other fields of work.
This was the year when AI was beginning to explode and came out as a General Problem Solver(GPS). It’s aim is to solve all the general level problem that human encounter in their day-to-day life.
McCarthy defined the high-level-language Lisp, which proved to be more dominant than AI(Artificial Intelligence) programming language in the coming years. And the prediction came to be true, now the machine learning is dominating AI without any doubt. “Machines will be capable, within 20 years, of doing a work than a man can do” – Herbert Simon (1965). Simon also stated that Machines will be able to beat humans in computer chess game in the coming 10 years and consequently significant theorem would be proved by machines. All these predictions came to be true as after 40 years a computer chess game was built and the prediction came to be true but not in 10 years rather 40 years after the prediction.
And do you know the best part?
A report came to light when an advisory committee found that “there’s no machine translation for general scientific text and none is in immediate use”. For which all US Government funding for the general translational projects were dropped.
In 1969, Minsky and Papert’s book perceptions soon derived that it could represent very little. Thier results apply nowhere to multilayer networks, more complex, so research funding for net-neutral research were soon dwindled to almost nothing.
After that, professor sir James Lighthill was elected from the UK Parliament to analyse the state of AI research that found to be huge unsuccessful in making AI to the next level. He further concluded saying that nothing being done in AI couldn’t be done in other field of sciences. He mentioned about the major drawback that it hold is “combinatorial explosion” or “intractability” that means it can only be able to solve mini or toy versions. Following which the British Government withheld its funding for AI research which came to an end on 1983. The whole scenario came to known as “AI Winter”.
But there’s a catch.
The first successful commercial expert system, RI, began into operation at the Digital Equipment Corporation (DEC). The program helped solved a lot of problems related to a computer system. By 1986, the program helped save about $40 million of the company.
Nearly every major US companies has started to employ the AI and related experts to handle the machine work.
In 1981, the Japanese announced the “Fifth Generation” project as a 10-year plan to build advanced computer system.
Adding which US formed various Microelectronics and gadgets to create national competitiveness in the growing world.
In Britain, the funding was restored that was cut by Lighthill report.
That’s not all…
In the mid 1980s, at least four different groups reinvented the back propagation learning first found in 1969.
These so called connectionists models of intelligent systems were seen by a direct competitor both to the symbolic models promoted by Simon and Newell.
In this way the neural network successfully got back to the market.
But here’s the deal:
The sudden collapse of the so-called Machine Language, LISP proved to be a major drawback in the development of AI Machines. Lisp, the major programming language that’s used in AI was replaced by an alternative programming language brought about by Sun Microsystems. The system proved to be too expensive to maintain. The system is quite difficult to update, they could not learn, they were unable to give a perfect output on some unusual input. Experts proved it a perfect machine but in some special context. By 1991, the Fifth Generation Computer System(Japan) slammed it down in 1981. Also, funding from US Government soon cut down.
This tragedy came to known as the “Second AI Winter”.
One of the most important tool of AI is the internet. Over the past 60 years, the history has revealed that the computer science department has been contributed a lot more in the development and evolution of AI. But at the same time recent study reflects more upon problem solving strategy and not on the algorithm to be used. Many day-to-day problems shows that AI could easily be able solve those problems without any doubt focusing on the problem rather than a hand-coded programming engineering.
But there’s a catch.
Till now things worked so well that AI has developed itself and serving the human all around the world. This was 2010 when significant advances in machine learning especially Deep Learning (Neuron Network).
Speech recognition and a few other computer processing were dominated by deep learning.
What is Machine Learning?
Machine Learning is a branch of Artificial Intelligence that is often used in a computer science department using statistical techniques to make computers learn with data and make healthy use of machines without being abused.
You May Also Like to Ponder About:
- Top 10 Unknown Features of Latest Android P | 2018
- Why ICO is Hotter than Jennifer Lawrence?
- The Insider’s Edge to Bitcoin (2018)
- Will Information Ever Rule the World?
What are the most widely used programming languages that are used in AI(Artificial Intelligence)?
Most highly rated and widely used programming language is the Python. Another most widely used language is Haskell.
Top 10 programming languages used in AI:
1) AIML (Artificial Intelligence Markup Language) is an XML dialect that is used in chatterbots.
2) IPL (Information Processing Language) was the first programming language to be build and came into use to solve general problems using AI. It supports problems such as lists, association, dynamic memory allocation, recursion, multi-tasking, etc.
3) LISP is one of the oldest programming language which is used in AI based on lambda calculus. Lisp is made up of linked lists and its source code is itself made of lists. Concluding which LISP programs can manipulate source code as a data structure, giving rise to macro systems to allow programmers to create their own code or domain specific programming languages using LISP.
4) SMALLTALK has been extremely used for machine learning, genetic algorithm, and neural network.
5) PROLOG is a widely used declarative language where programs are expressed as relations, and execution occur by running over these relations. It’s useful in language parsing, symbolic reasoning and database. Today, it’s one of the most widely used languages of all.
6) STRIPS is a programming language that is used for automated planning problem (also known as AI Planning). It represents the initial state, goal state and the final state(set of actions). For the actions to be performed is already stated and the output to be generated is already guessed and performed.
7) PLANNNER is a language that works both as a logical and procedural language. It gives a procedural interpretation to logical operations and shows results in a pattern based system.
8) R is most widely used programming language in a new style artificial intelligence includes numerical analysis, neural networks and Off course machine learning. It is considered as one of the major standard language. It is used in programmes like vectorial computation, functional programming and object-oriented programming. Most importantly it supports Deep Learning.
9) Python, which is quite often used in Artificial Intelligence in today’s world. It find its usage in general AI, Machine Learning, Natural Language Processing and Deep Learning.
10) Haskell is another good programming language for AI. It’s easy to understand and work with. The only drawback is that working with a graph is a bit difficult at first because of purity.
What actually Machine Learning (ML) can perform?
There are many uses of ML but let me tell you a few of the most renowned ones:
1) CLASSIFICATION: it can figure out objects clearly (differentiate between male and female user’s from web, cats and dogs in an image, spam from mails, etc)
2) Regression: predict the value of an object(the value of a house, sales in the coming month, energy consumption of the house, etc)
3) Clustering: group together the user with similarity(cluster the users with similar interest from a user’s web activity, etc)
4) Dimensions Reduction: the data can be reduced from higher dimensions to the lower dimensions (useful for visual purpose, etc)
5) Ranking: with the help of user activity it can rank the data on the internet (using clicks, relevance, quality data, etc)
6) Recommendations: using users search ability, it can filter the large collections of data helping the user with similar interest, increase sales, etc)
Deep Learning/State of Art:
Deep Learning is a subsidiary of Machine Learning that deals only with the algorithms of the functions and structure of the brain.
Everyday the technology is evolving day by day, new bots, machines were growling around the world.
There are a number of uses of AI (Artificial Intelligence) in the modern world. Stating a few of them here:
1) Expert System: A special kind of software designed to perform tasks like expertise knowledge to give advice in various domains like medical, military, chemical, geological exploration, etc. They are designed to give a perfect advice and to help assist in different research work. They are designed to help assist expert but not to replace them. It helps us clarify the problems that may occur create uncertainty in system.
Building an expert system is known as Expert Engineering. And the person doing this job are called Expert Engineer. It uses both factual and heuristic knowledge.
Factual knowledge is the knowledge that is mostly found on the books or journal.
Heuristic knowledge is the knowledge that is less exhaustive, more experimental, and stand out in performance.
Examples of Expert Systems:
3) Agricultural and many more.
2) Natural Language Processing: NLP is the branch of Computer Science and Artificial Intelligence which interacts with the human and computers using natural human language.
Most used natural language processing involves speech recognition, natural language understanding and natural language processing.
Examples: social media marketing, formulating response to questions, etc.
3) Computer Vision: The human generally see a lot of things rather than hearing, feeling, smell or taste. So we the people use our vision to see the things in our surrounding. Similarly the goal is to provide vision to the computer so that they can learn about their surrounding using camera attached to it.
In this way, the face recognition process works.
Examples: automatic inspection, identifying humans, controlling process, interaction (computer-human interaction), navigation, and many more.
4) Robotics: A ROBOT is an electro-mechanical device that can be programmed to perform various operations or reprogrammed to make it perform a variety of tasks to move materials. An intelligent robot includes a kind of feature that can be able to change its gesture with respect to a varying environment.
Examples: Honda Robot, R2-D2, Rosey, etc.
5) Automatic Programming: In computer science, the automatic programming signals the notion about the computer programming in which it generates computer programs to allow human programmers to write their code at higher level abstractions.
What is the future of AI?
The future of AI is unknown. That means no one knows that how much these gadgets/technology will help human to help solve the general problems. These can rupture at some point and at other, can be built-in to face jeopardy challenges. Automated transportation will be easier and faster. Robots will help older humans to make their life easier at such an elder age when there’s no one to assist them. Number of smart cities will be developed in future as vehicles, phones, buildings, etc.
But at the same time, researchers seem to be doubtful in their act as an issue.
With the exponential rise in the rate of development of technology it is logical to believe that human will have sophisticated life as robots will come to use in the future. The development of AI will continue to an increase in the coming years.You may even find that you are living with AI right now in your home, without even knowing it! Smart hubs are some of the first AI devices that have entered our everyday lives. The development of AI will continue to an increase in the coming years.