Artificial Intelligence refers to the simulation of the human brain in machines so that these intelligent machines can act, think and behave like humans. These machines should be rational and should give the best possible solution. It is based on three principles of learning, reasoning and perceiving. Alan Turing’s Test to check whether “machines can think” was a turning point in the technology. The term Artificial intelligence was first coined by John McCarthy in 1956.
Artificial Intelligence can be summarised as
- The ability to solve problems
- The ability to act rationally
- The ability to act like humans
It is commonly implemented in computers using software and has two possibilities
- A system with intelligence should behave as intelligent as human
- A system with intelligence should behave in the best possible manner
Some other definitions of AI can be categorized as
- System that think like humans
- System that think rationally
- System that act like humans
- System that act rationally
Turing Test approach
In an environment set up, a questioner would ask the same set of questions to a computer and a human respondent. If the computer gives a similar answer as the respondent, then the computer is as intelligent as humans.
Artificial intelligence is an interdisciplinary subject which includes the study of
- Philosophy – Logic, methods of reasoning, mind as a physical system, foundations of learning,
- Mathematics – Formal representation and proof, algorithms, computation, (un)decidability, (in)tractability, probability.
- Economics – utility, decision theory
- Neuroscience – neurons as processing information units
- Psychology/Cognitive Science – how do people behave, perceive, process Cognitive Science information, represent knowledge.
- Computer engineering – building fast computers
- Control theory – design systems that maximize an objective function over time
- Linguistics – knowledge representation, grammar
Comparisons between human and computer intelligence
Normal humans have the same intellectual mechanisms and that differences in intelligence are related to ‘quantitative biochemical and physiological conditions‘, but computer programs have plenty of speed and memory but their abilities correspond to the intellectual mechanisms that program designers understand well enough to put in programs.
Whenever people do better than computers on some task or computers use a lot of computation to do as well as people, this demonstrates that the program designers lack understanding of the intellectual mechanisms required to do the task efficiently.
Why Artificial Intelligence?
- Computers are fundamentally well suited to performing mechanical computations, using fixed programmed rules.
- AI research is allowing us to understand our intelligent behaviour.
- Artificial machines perform simple monotonous tasks efficiently and reliably, which humans are ill-suited to.
- AI can help us understand this process by recreating it, then potentially enabling us to enhance it beyond our current capabilities.
- AI aims to improve machine behaviour in tackling such complex tasks.
- Humans have an interesting approach to problem-solving, based on abstract thought, high-level deliberative reasoning and pattern recognition.
- For more complex problems, things get more difficult… Unlike humans, computers have trouble understanding specific situations and adapting to new situations.
The Goals of Artificial Intelligence
- Deduction, reasoning and problem solving – To develop algorithms that human use. Algorithms can require enormous computational resources and the problem goes beyond a certain size.
- Knowledge Representation – To build a machine with the capability of making a working assumption and common sense.
- Planning – In classical planning problems, the agent can assume that it is the only system acting on the world. Multi-agent planning uses the cooperation and competition of many agents to achieve the given goal.
- Learning – Machine Learning is the fundamental concept of AI search. Unsupervised learning is the ability to find a pattern in a stream of input. Supervised learning includes both classification and numerical regression.
- Natural Language Processing – Natural Language Processing gives machines the ability to read and understand human language. Semantic representation is a common form of representing natural language.
- Perception – machine perception is the ability to use sensors to deduce aspects of the world.
- Motion and Manipulation – Intelligence is required for robots to handle tasks such as object navigation and manipulation.
- Social Intelligence – Effective computing is the development of systems that can recognise, interpret and simulate human effects. Emotion and social skills are important to an intelligent agent to understand others and to make better decisions.
How Artificial Intelligence is used?
- Narrow AI – It is also called “Weak AI” as it works within a limited context. Examples of Narrow AI include Google search, Image recognition, Siri, Alexa, autonomous vehicles etc.
- Artificial General Intelligence – This category falls under “Strong AI” as we see in movies like the “Transformers”.
History of Artificial Intelligence
The birth of AI (1943 – 1956)
- McCulloch and Pitts (1943): a simplified mathematical model of neurons (resting/firing states) can realize all propositional logic primitives (can compute all Turing computable functions)
- Alan Turing: Turing machine and Turing test (1950)
- Claude Shannon: information theory; the possibility of chess-playing computers
- Boole, Aristotle, Euclid (logics, syllogisms)
Early enthusiasm (1952 – 1969)
- 1956 Dartmouth conference
- John McCarthy (Lisp)
- Marvin Minsky (first neural network machine)
- Alan Newell and Herbert Simon (GPS)
- Emphasis on intelligent general problem solving
- GSP (means-ends analysis)
- Lisp (AI programming language)
- Resolution by John Robinson (the basis for automatic theorem proving)
- heuristic search (A*, AO*, game tree search)
Emphasis on knowledge (1966 – 1974)
- domain-specific knowledge is the key to overcome existing difficulties
- knowledge representation (KR) paradigms
- declarative vs procedural representation
Knowledge-based systems (1969 – 1999)
- DENDRAL: the first knowledge-intensive system (determining 3D structures of complex chemical compounds)
- MYCIN: first rule-based expert system (containing 450 rules for diagnosing blood infectious diseases)
- EMYCIN: an ES shell
- PROSPECTOR: first knowledge-based system that made a significant profit (geological ES for mineral deposits)
AI became an industry (1980 – 1989)
- wide applications in various domains
- commercially available tools
- AI winter
- Current trends (1990 – present)
- more realistic goals
- more practical (application oriented)
- distributed AI and intelligent software agents
- resurgence of natural computation – neural networks and emergence of genetic algorithms – many applications
- dominance of machine learning (big apps)
Scope of Artificial Intelligence
- Manufacturing – We have seen the examples of industrial automation and large scope of deployment of robots to help in manufacturing activities. Many organizations are moving towards human-less manufacturing.
- Medical and healthcare – Image processing and image analysis are helping significantly in the medical field in the areas of diagnosis.
- Autonomous cars – AI software systems are the actual brain behind autonomous cars. It takes constant learning, monitoring of surrounding, analyzing information in real-time and converting them into appropriate driving actions. Safety, security aspects are supreme in autonomous cars.
- Retail – We have recently seen the example of Amazon Go – a store without any manned check-out counters. These types of stores are empowered by the concept of AI technology.
- Cyber security – The future application of AI in cyber security will ensure in curbing hackers.
- Face recognition – The launch of iPhone x with face recognition feature was a step towards AI future. In the coming years, iPhone users might be to unlock their phones by looking into the front camera.
An agent is anything that can be viewed as perceiving its environment through sensors and acting upon its environment through actuators. For example, a human agent, robotic agent, software agent.
- Human agent – eyes, ears and other organs for sensing. Hand, legs, mouth and other body parts for actuators
- Robotic agent – Cameras and Infrared sensors for sensing. Various motors for actuators
- Software agent – Keystrokes, file contents, received network packets as sensors. Screen display, files, received network packets are actuators.
The fundamental faculties of intelligence are
- Understanding, reasoning, learning
An Intelligent Agent should sense, act and should be autonomous (to some extent). It also must be rational. AI is about building rational agents.
Artificial Intelligence algorithms
Artificial Intelligence algorithms can be broadly classified under three categories
- Classification algorithms – These belong to the class of supervised learning where the algorithm classifies the variable of interest into different classes. For example, this algorithm can filter spam from our emails. Examples of this kind of algorithm are – Naïve Bayes, Decision Tree, Random Forest, Support Vector Machines and K nearest neighbours.
- Regression algorithms – These algorithms are also a part of supervised learning. These determine the output based on the independent variable of interest. Predicting stock market and predicting weather are based on these algorithms. Examples of these type of algorithms are Linear regression, Lasso regression, Logistic regression, Multivariate regression and Multiple regression.
- Clustering algorithms – In this process the algorithm segregates and organises the data points into clusters based on similarities within the group. It can filter all transactions which are fraudulent based on some properties of the transaction occurring. Examples of these types of algorithms are K-Means algorithm, Fuzzy C-Means algorithm, Expectation Maximisation (EM) algorithm, Hierarchical Clustering algorithm.
Real-life applications of Artificial Intelligence
- Expert systems – Tracking system in flights and Clinical systems
- Natural Language Processing – Google speech assistant, automatic voice output, language translator are all examples of Natural Language Processing using Artificial Intelligence.
- Neural networks – Neural networks find patterns as in face recognition, character recognition or handwriting recognition.
- Robotics – Robots used in car manufacturing companies and other industries.
- Fuzzy Logic Systems – Consumer electronics and automobiles work on fuzzy logic.
- Game playing – There is some Artificial intelligence in them and they play well against people.
- Speech recognition – instructs some computers using speech.
- Modelling Human Performance
Artificial Intelligence Issues
Artificial Technology is developing at such a massive speed that the researchers believe that this technology can become strong enough to control the humans itself.
Violation of human privacy – Artificial Intelligence has such capabilities that it can process natural language using NLP and have language translators. This makes them intelligent enough to read emails and listen to telephonic conversations.
Threat to Human Dignity – Artificial Intelligence empowered robots have led to the automation of industries in many sectors. But they should not replace positions of human beings which are based on ethics and morals such as doctors, police, lawyers etc.
Threat to safety – The self-learning, the self-motivating Artificial Intelligence equipped agents have become so powerful that they can even control human power. They can stop humans from achieving their goals and cause disastrous effects. As a consequence, humans are threatened by their own technology.
Artificial Intelligence Advantages
- It can help improve our way of life.
- Machines will be able to do jobs that require detailed instructions
- Artificial intelligence also makes interactive electronic games more fun by making the computer-controlled characters more realistic and human-like.
- Use robots for heavy construction, military benefits, or even for personal assistance at private homes.
- There will be fewer injuries and stress to human beings.
- Artificial intelligence opens up new and exciting avenues for entertainment possibilities.
- Many of our health problems now have possible solutions with the use of Artificial Intelligence in studies at universities.
- Mental alertness and decision-making capabilities.
- Scientists have been using AI to test theories and notions about how our brains work.
Artificial Intelligence Limitations
- To date, all the traits of human intelligence have not been captured and applied together to spawn an intelligent artificial creature.
- It is Costly.
- Currently, Artificial Intelligence rather seems to focus on lucrative domain-specific applications, which do not necessarily require the full extent of AI capabilities.
- It can’t handle a difficult situation.
- There is little doubt among the community that artificial machines will be capable of intelligent thought in the near future.
M.Tech (VLSI Design and Embedded system)
BS Abdur Rahman University