Artificial Intelligence has emerged out to be one of the most popular terms of computer science in recent times. This article discusses one of the classifications of AI, ie. Artificial Narrow Intelligence (ANI).
The term ‘Artificial Intelligence’ first emerged in 1956 at the Dartmouth conference typically organized by John McCarthy. The ambitious six-week project aimed at creating computers that could use language, form abstractions, and numerous concepts to solve varied forms of problems now reserved for humans and to improve themselves. This kick-started numerous research centers to explore the potential of AI.
At that time, scientists convincingly believed that a “2 month, 10 man study of AI” would solve the biggest and unsolved parts of the AI equation, but more than six decades later even, the dream of successfully creating and achieving artificial intelligence still eludes us. But the progress we have made so far is also astonishing, and as a result of this, the field of AI has been divided into three major categories: Artificial General Intelligence(AGI), Artificial Narrow Intelligence(ANI) and Artificial Super Intelligence(ASI) so as to have a better understanding about each.
So in this article, we will be learning about ANI.
Why Narrow AI?
Artificial Narrow Intelligence (ANI) also commonly known as weak AI or narrow AI. It is particularly that type of AI that we have been able to successfully realize and implement in the present date. It is a goal-oriented, narrow rangeability holding perspective that executes specific focused tasks, without the ability to the self-expand mechanism(functionality). Machines that are focused on one narrow task operate under a narrow set of constraints and limitations, that’s why they are commonly referred to as WEAK AI. Narrow AI doesn’t replicate proper human intelligence, it basically simulates/mimics human behavior based on a narrow range of parameters.
Narrow AI has faced a lot of ups and downs in the past few decades incorporated by the progress made in machine learning and deep learning. Narrow AI is basically a combination of two words: AI meaning the technology to simulate human behavior in machines through a set of algorithms and the second word being Narrow, so it basically means implementing the concept of AI but with a narrow range of abilities.
Narrow AI’s machine intelligence is basically achieved through the concept of Natural Language Processing (NLP). NLP concept is the common functionality in chatbots and similar AI domains in which the machines are basically programmed in a way to interact with humans using speech and text recognition mechanism.
Different Types of Narrow AI
Narrow AI has two possibilities, either it can be reactive or can have a limited amount of memory.
- Reactive AI: It is the basic version, having no memory or data storage capabilities. It emulates the human mind’s behavior and responds to different interpretations without any prior experience.
- Limited Memory AI: It is more advanced, having great memory and data storage capabilities enabling machines to interpret precisely using statistical data. Most of the AI is the Limited Memory AI, enabling machines to use a large amount of data especially in the domains of Deep Learning to give results with utmost accuracy.
Examples of Narrow AI
- Virtual Assistants like Siri by Apple, Alexa by Amazon, Cortana by Microsoft, etc.
- Used in medications and prediction tools to diagnose cancer and other health-related issues with extreme accuracy through human behavior cognition, replication, and reasoning.
- IBM’s Watson: It is capable of answering questions asked in natural language interpretations.
- Rankbrain: It is an algorithm used by Google to sort the search results.
- Self-Driven Cars.
- Facial/Image recognition and interpretation software.
- Social Media Marketing tools for keeping a check on violating contents and Email spam filters.
- Manufacturing and Drone-based robots.
- Tools for bringing out the best recommendations based on search results, purchases, history, etc. in the entertainment and marketing domains.
Limitations of Narrow AI
Weak AI has limitations because of its limited capabilities which add to the prospect of causing harm if a system fails. If we take into account an example of a driverless car that calculates the location of oncoming vehicles but somehow it miscalculates the location which causes a deadly collision or a system failure if it is used by some third party or destructive organizations, then it can lead to enormous destructions to the human mankind. Another drawback is to determine the faulty identity for a malfunction or a design flaw.
A further concern is regarding the huge loss of jobs caused by the automatic functionality and performance of an increasing number of tasks. Slowly the machines are taking over human manpower for intelligent automation through AI. So it must be thought-provoking that whether there will be an increase in the job opportunities or the machines will gradually take over the entire physical manpower with the emerging levels of AI.