13.4. Types of AI

AI systems are different from humans and machines, and the degree of variation helps classify the different types of AI.  AI can be classified based on ability and functionality. There are three categories of AI based on ability: narrow AI, general AI and super AI.

Types of AI: Based on Ability

Categories of Artificial Intelligence Based on Ability are narrow AI, general AI and super AI
Types of Artificial Intelligence Based on Ability
Artificial Narrow Intelligence (A.N.I) A form of weak AI refers to a computer’s ability to perform a single task well.  This is the type of AI we see today. Natural language processing used in smart assistants like Siri and Google are examples of narrow AI.
Artificial General Intelligence (A.G.I) A form of strong AI that mimics human intelligence. It means that machines have the ability to apply what they have learned across different tasks, take in new information, and apply reason.
Artificial Super Intelligence (A.S.I) A form of supernatural AI that exceeds human capabilities.

Despite having developed complex machines and applications such as Siri, Alexa, or other virtual assistants, all AI systems that currently exist are classified as A.N.I. According to a consensus among researchers, A.G.I and A.S.I are still decades away (Dilmegani, 2021, Aug 11).

Example: Chatbots

Chatbots and virtual assistants (i.e. Siri and Alexa) are computer programs that use AI and natural language processing (NLP) to understand customer questions and automate responses to them, simulating human conversation (Ibm.com, n.d.). These bots are cloud-based and able to gather information, such as user preferences or browsing history, to provide solutions that are specific to each individual. Chatbots improve the individual customer experience as the Cloud is able to provide personalized and relevant information.

In the hospitality industry, Best Western, in collaboration with IBM Watson Advertising, managed to use AI and NLP to create a more personalized customer experience for those looking to book Best Western hotels for their holidays. Specifically, “Conversations”, an AI-powered advertising model from IBM, was used to provide one-on-one connections and recommend options based on the user’s intent (Ibm.com, 2021). For example, Conversations was able to provide users with ads containing information on their destination and other tips and tricks!

Types of AI: Based on Functionality

AI can also be categorized based on functionality: reactive machines, limited memory, theory of mind, and self-awareness.

Categories of Artificial Intelligence Based on functionality
Categories of Artificial Intelligence Based on functionality include reactive machines, limited memory, theory of mind and self-awareness.
Reactive Machines AI that conducts tasks for specific objectives. It is the first step of any AI system where no learning takes place. An example of this type of AI is Deep Blue the IBM program that beat chess champion Garry Kasparov.
Limited Memory It is the ability of AI to use past experience and information to make predictions.
Theory of Mind AI with the awareness that others have their own beliefs, objectives, and intentions. This type of AI is not yet developed and will be able to distinguish and understand emotion.
Self-Aware  AI with the sense of self and consciousness. This type of AI is not yet developed.

Module 1: Introduction to Artificial Intelligence” from  Thriving in the Age of Artificial Intelligence by the Faculty of Engineering and Applied Science at Ontario Tech University under a Creative Commons Attribution license, unless otherwise noted.

License

Icon for the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License

Information Systems for Business and Beyond Copyright © 2022 by Shauna Roch; James Fowler; Barbara Smith; and David Bourgeois is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License, except where otherwise noted.

Share This Book