What is Artificial Intelligence?

Sometimes referred to as Machine Intelligence, Artificial Intelligence (AI) in computer science is machine-demonstrated intelligence – different from the natural intelligence of a human. 

Leading AI textbooks define the field as the study of “intelligent agents”: any device perceiving its environment and taking action that maximises its chance to achieve its goals successfully.

In common terms, people use the word, AI to refer to machines or computers that imitate reasoning functions, such as learning and problem-solving, which are normally associated with the human mind.

A more detailed description characterises AI as “the ability of a program to correctly interpret external data, learn from it, and use it to accomplish particular goals and tasks.”

artificial intelligence

History

The first time AI appeared, it was used as elements of storytelling and often used in fiction. Examples were R.U.R. by Karel Čapek (Rossum’s Universal Robots) and Mary Shelley’s Frankenstein. 

These fictional characters and their fates threw up many of the same issues that dominate the debates on the ethics of artificial intelligence.

John McCarthy coined the word “artificial intelligence” in 1956 at a laboratory at Dartmouth College. He used the term to differentiate that field of science from cybernetics and to avoid the influence of cyberneticist Norbert Wiener.

The founders and members of AI research were Arthur Samuel (IBM), John McCarthy (MIT), Marvin Minsky (MIT), Allen Newell (CMU), and Herbert Simon (CMU). They and their students created systems that the press characterised as “surprising”. 

Machines mastered playing skills for checkers (c. 1954) (and by 1959, some reportedly beat human “opponents”). They interpreted language, solved algebra problems and demonstrated logical theorems – amongst others.

By the mid-1960s, U.S. research was heavily funded by their Defense Department, and labs had been established around the world. The designers of AI were positive about the future: Herbert Simon expected that “machines will be able to do whatever job a man can do in twenty years’ time.”

Marvin Minsky agreed, saying, “the question of developing ‘artificial intelligence’ would be effectively solved within a decade.”

In 1955, the world saw the first known use of artificial intelligence

How Does AI Work ?

Artificial Intelligence (AI) refers to computers’ simulation of human intelligence, where computers are programmed to think like humans and imitate their behaviours. The word can also refer to any computer that displays human mind-related characteristics such as thinking and problem-solving.

The perfect function of artificial intelligence is its capacity to rationalise and take decisions that have the greatest potential to accomplish a particular goal. The first thing that most people generally think about when they hear the words ‘artificial intelligence’ is robotics. That is because large-budget movies and novels weave tales about human-like robots that (mostly) wreak havoc on earth.

There may be nothing further from the truth, though. Artificial intelligence is based on the premise that human intelligence should be described in such a way that a computer can readily mimic it and execute functions, from the easiest to the more complicated.

Artificial intelligence targets include comprehension, thought, and understanding. AI is evolving continuously to benefit a lot of different industries. Machines are wired using a mathematics, computer science, linguistics, psychology and more cross-disciplinary based approach.

Applications of Artificial Intelligence

The possibilities of artificial intelligence are endless. The technology can be applied to a great many different industries and sectors. In the pharmaceutical sector, AI is being studied and used in clinics for dosing medications and multiple therapies, and in the operating room for surgical operations. 

Certain examples of AI machines include computers which play chess and self-driving cars. Each of these devices must consider the consequences of each action they take, as each action determines the ultimate outcome. 

For chess, game-winning is the final goal. The computer programme will compensate for all external data for self-driving vehicles, and measure it to function in a manner that avoids collision. 

Artificial intelligence also has uses in the financial industry, where it is used to identify and report banking and finance events such as irregular use of debit cards and big account deposits – all of which support the bank’s fraud department.

AI applications are also used to help streamline and promote trading. This is achieved by making it easy to estimate the supply, demand and price of securities.

Potential Harm

Potential Harm

Widespread use of artificial intelligence could have hazardous or undesirable and unintended consequences. Scientists from the Future of Life Institute, among others, outlined some short-term research goals to see how AI affects the economy, the laws and ethics involved with AI, and how AI security risks can be minimised.

The scientists have proposed that optimising function in the long term should continue, while minimising potential safety risks that come with new technologies. Microsoft founder Bill Gates, Physicist Stephen Hawking, and SpaceX founder Elon Musk have wondered if AI could evolve to the point where humans could lose control of it. Hawkins theorised that if that ever happened, it could “spell the end of the human race”.

Once machines are developed with artificial intelligence, they could potentially take off on their own and redesign themselves at ever-increasing speeds. People, who are constrained by slow technological evolution, couldn’t compete and would be superseded.

Challenges of AI

Artificial intelligence’s ultimate work goal is to develop technologies that allow computers and machines to operate as ‘smart machines’. The general problem of intelligence simulation (or creation) has been broken down into sub-problems. They consist of unique characteristics or attributes that researchers expect to see in an intelligent system.

Conclusion

In brief, exceptional developments have been made in recent years in AI systems’ ability to integrate intentionality, intellect, and adaptability into their algorithms. 

Instead of being mechanistic or deterministic on how the computers work, AI software is learning as it goes along and integrating real-world knowledge in its decision making.

In this way, individual efficiency is improved and the capacities of humans are expanded. Such developments, of course, often make people anxious about the doomsday situations sensationalised by film-makers. Situations in which AI-powered robots take over from humans or weaken basic values scare people and lead them to wonder if AI is making a useful contribution or if developers are running the risk of endangering humanity’s essence.

Countries can move forward with the appropriate safeguards and gain the benefits of artificial intelligence and emerging technologies without sacrificing the significant qualities that define humanity.

FAQ'S

AI’s rise has been received with mixed reviews. The late Stephen Hawking announced that AI may be the “worst thing in the history of our species” and Elon Musk said that “AI is a profound danger to the life of human society.” 

Nevertheless, AI is an important part of our daily lives and shouldn’t be seen as so intimidating. When AI continues to expand into our factories and workplaces, it streamlines and enriches operations to the degree that businesses that employ AI are better equipped for long-term success than those that don’t.

Technology has made our lives very easy, particularly in the field of artificial intelligence. Even without us realising when this happened, Alexa’s technology and other apps have seeped into our houses.

Here is a quick overview of the future where AI will play an important role in every phase of our lives.

  • Continuous machinery maintenance is an enormous expense to manufacturers and the shift from reactive to predictive maintenance has become a must for all manufacturers. AI has helped to save companies precious time and money by leveraging sophisticated AI models and artificial neural networks to devise forecasts about equipment loss and training technicians ahead of time.
  • Driverless cars: autonomous vehicles now exist in many countries. The U.S. Department of Transportation has gone ahead and published some concepts and guidelines relating to the various types of automation that can be implemented.
  • Automation of robotic processes: automation of robotic processes refers to the use of machine learning to automate tasks which rely on rules. It should help people concentrate on certain important facets of their job and leave the repetitive tasks to machines.

Additionally, automated servicing has helped prolong system life, resulting in substantial decreases in labour costs.

With regard to knocking out jobs, AI will be very effective, one narrow vertical at a time. These would be more than the boring manual tasks which most people believe a computer should perform.

In fact, unlike previous technological revolutions, AI would harm the highly skilled almost as much as, or perhaps even more than the unspecialised workers. Aside from truck drivers and customer service staff, also highly-specialised areas such as oncology, dermatology, music composition, architecture and immigration law can all be updated by Ai.

Artificial Intelligence (AI) is all about simulating machine intelligence. A sub-set of AI is Machine Learning (ML). AI can be accomplished in two major ways.

The first way to imitate intelligence is to create a series of rules/conditions or handmade versions. Such rules/models are based on human experience. Because of the flow of information, the devices behave as a human person would (Stockfish chess engine or Computer Vision-based tracking etc).

The second approach or ML approach is data-guided. We take a flexible model and use the data to automatically fine-tune its parameters. The model learns by itself from data and acts smartly (Alpha Zero chess engine)

Like every other technology, AI can be used for both good and evil, but it is neutral in and of itself. Thinking that artificial intelligence would be good or evil would mean it would assume human traits – it will have to be mindful of the choice it takes to become very good or bad.

No. Nothing can become infinitely smart. In fact, infinitely nothing else. 

In the real world, as well, an AI cannot affect anything unless it is connected to peripheral machines. It cannot connect to anything, because it doesn’t have any hands. It is simply a (computer) program that runs inside a box.

Besides, unless it is expressly designed to enslave mankind (with clear instructions about what this entails, and how to do so), it cannot. Since machine systems, however clever they can be, have no motives, they don’t have any feelings and so they have no urges.

Robots are hardware, and the software is artificial intelligence (AI).

AI should not be regulated as a basic technology. It also seems impractical for the government to halt the installation of a neural network on your desktop. 

However, there are applications of AI that need regulation, for example, autonomous driving. AI also has new effects on antitrust (monopoly regulation), which policymakers have not yet thought about but should have.

Many of the AI control debates originate from unfounded concerns of “sentient AI” or “evil killer robots”, rather than a better understanding of what it is able and not able to do.

Since AI is still embryonic today and is rapidly evolving, any country’s heavy-handed regulation would stall the AI development of that country.

However, certain cases concerning AI usage require legislation both to secure individuals and to speed up their adoption. To ensure protection, the automobile industry is already heavily regulated. Thinking of how these rules will evolve in the light of new AI technologies such as autonomous driving, should benefit the industry as a whole. 

The same happens to other fields, including pharmaceuticals, weapons control, financial services, etc. But the legislation needs to be industry-relevant and based on a thorough analysis of the use-cases and the consequences we do/don’t want to see in particular industries, rather than on the basic technologies.

Leave a comment

Send a Comment

Your email address will not be published. Required fields are marked *