Imagine a day you wake up, and your entire day is scheduled. A virtual associate is assisting you throughout the day, from Making a medical appointment, booking a cab, ordering your favourite food according to your present mood, making a call when you’re thinking about a particular person after listening to your thoughts. A day where you no longer have to utter a word, but your virtual assistant can listen to your thoughts and make your mind real in a matter of time?? This can be true in the near future.
A term coined by John McCarthy in 1956 turned to an avatar by the dawn of 21 st century.
Watch: Streams of ‘Artificial Intelligence’.
Buildings an intelligent machine will not address the idea of artificial intelligence. The real question arises when the core of the matter is analyzed.
What makes a machine Intelligent ?
Approaching the concept of intelligence involves two aspects.
Rationality and humanity. Brilliance without rationality is a mere memory, which cannot be called intelligent.
To move, to nod, to fidget, to scratch your head, to move your eyeballs… An average human being makes 16000 decisions a day involuntarily. The human brain is a shell of wonders and billions of neurons swimming across the body, transmitting information in a fraction of seconds. Emulation of the human brain or making a computer digitally conscious is the primary objective of artificial intelligence.
Like an Immortal Dictator, Artificial Intelligence seems to rule the society.
The idea of AI can be simplified as the mimicking of the human thought process by a computer which makes the intelligence more flexible & cheap compared to the complex neural network of the human brain along with the fundamental characteristics of humans, such as the ability to reason, knowledge representation, planning, learning, movement, manipulation, to develop perception, discover meaning, generalize, or learn from experience and the toughest of all social and emotional intelligence. There are two distinct methods involved in the present development of AI.
According to the standard definition from Britannica it can be defined as
- symbolic approach: seeks to replicate intelligence by analyzing cognition independent of the brain’s biological structure in terms of the processing of symbols.
- connectionist approach: involves creating artificial neural networks, in imitation of the brain’s structure.
Creation of Artificial Intelligence
At the beginning of the creation of AI, it was the symbolic approach. Mere object recognition from enormous data in the database. Algorithms use the existing rules like grammar and syntax for predicting. The existing coding and programming trend demanded more clarity and advancement, which eventually popularized the mainstream of Artificial Intelligence called machine learning (ML).
The preexisted computer programming technique modified prior to AI is machine learning. The development from pattern recognition to Morphing of multiple data sets into a single data, ML helps in bringing advancement in every aspect of humankind. Later in the spectrum of AI, to respond more with humans, an artificial stress system is induced to the machine via Virtual neurotransmitters, which paved the way for Affective computing to achieve Enhancement in human-machine relationships.
Apart from the usual pattern identification, AI is now searching for all kinds of texture that humans cannot create within the scope of our biological capability. Like the creation of Adam in the bible, experts are in search of the living breath that can provide dexterity or subtle control of the thoughts to the machines, which is the most tedious task in the AI.
One of the major questions ever raised regarding AI was, are they huge robots who will take over the world someday, as seen in the movies and series?? The way sci-fi portrays AI as an enemy and a mistake committed by science wrong. This is a major misconception. AI is not only about creating huge robots but also the minute changes that can work on its own from a sensor to the software.
Narrow AIs are presently in use and service. The voice, face, fingerprint recognition and Siri, Alexa, and Cortana kind of virtual assistants in our mobile phones and other gadgets are the simple examples of the narrow Artificial Intelligence. The vision recognition in automated cars, self-driving cars, google glass are the advanced level of the narrow AI.
The intention behind the designing of these smart search engines and decision-making systems are for serving the purpose of complimenting and augment human abilities. Undoubtedly narrow Artificial Intelligence contributes a lot to the improvisation in our modern lifestyle. But unreaching the majority of human traits, it remains as mere machines operating with fed data.
Every advancement in the field of AI is analyzed based on the test run, followed by every stage. The final tests must meet the following criteria to take the program into further steps, like application and implementation.
Observing the behaviour of the system when a new input is introduced
Weighing and adjusting of the input in the newly introduced system.
These steps are repeated an ‘n’ number of times until the criterion matches else terminated. The involvement from never-ending algorithms to deep learning is not an overnight miracle. The progressive learning algorithm is a method where the algorithm can teach itself how to play chess; it can teach itself what product to recommend next online. And the models adapt when given new data.
When the system becomes self-sufficient, it can become an intellectual property. According to Oliver Schabenberger, SAS Executive Vice President and Chief Technology Officer, machines can never replace humans; they can only mimic the human brain and thoughts to become the smartest machines from smarter machines.
Challenges of Artificial Intelligence
The major challenge involved with Artificial Intelligence is data. There is no other practical way to incorporate data other than spoon-feeding the system. The large databases from different sects are clubbed to make a parent data pool is the theoretically existing solution. Considering the complexity and the practicality of such massive incorporation, researchers are still searching for better options.
You May Also Like To Read: New Invention of Artificial intelligence in Medicine with Stem Cells by WCM-Q
The subsequent problem evolving from data deficiency is that the particular Artificial Intelligence is created for a particular task; this specialization limits it to the particular set they are designed for.
According to Russell and Noverg in the introduction to artificial intelligence agents, the specific characteristics of symbolic AI are characterized as A typical AI that analyzes its environment and takes actions that maximize its chance of success. That’s the reason for the success of a computer system in the games when playing against a human. This computational analysis featured with the help of statistical methods. When a human plays not to lose, a computer compares the game with all possible outcomes but is not driven by the spirit to win. This is where the emergence of NLP begins.
Natural language processing (NLP) is a drastic addition to AI. The ability of computers to analyze, understand, and generate human language, including speech is called natural language processing. PEPPER (affection bot) and SOPHIA (robot) are a few examples of the natural language processing systems. The present challenge is making NLP more emotionally stimulant.
An example for the emotional bot of the AI system is that when the AI robot SOPHIA was interviewed she said someday she would like to have kids and family, even though that statement was from the data set she was fed with, the choice of making that emotional statement shaken the whole world. Another similar occurrence was when she asked her creator, “why did you create me??” None of the statements made were her direct choice of words but efficient programming that set her neural algorithms.
More about Artificial Intelligence
Artificial Intelligence, like electricity, is a general-purpose technology. The transparency offered by Artificial Intelligence is the most promising feature of this system. The introduction of AI in international relations and politics is a milestone. As a beginning, AI is introduced as an intern in international politics. Chatbot– Design virtual agent AI system to understand and translate into human behaviour to support human experience and Robotic process automation (RPA) – An application to function on repetitive commands of users without human interruption. It’s essential to cognize the dilemma of the subjects when automated decision-making systems (ADS) is implemented into the public sphere for the institutions and democratic setups that are already existing on the virtue of people’s money (taxes, donations…)
Even after the induction of the neurotransmitters and the artificial hormones, what is still lacking is the human-machine interface. There are still miles to go in the development of AI systems. From object recognitions to the Democratization of all essentials. Upon the transparency promised, Machine learning still complicates the field; whereas the ADS changes the game. As our research suggests, this is a source of anxiety for the general public and we don’t even know half of the story. The illiteracy about the topic confuses the public and stuck in the shadow of doubts, this can hinder the whole growth of the AI. people should understand the ethical implications of using AI in everyday life to public interests, including politics.
AI stands as a promise of tomorrow when the whole nature goes against humankind, hoping that the technology we gave birth might stand for us. Maybe that’s the promise of this avatar.