Artificial Intelligence

post image

With the use of A.I. on the rise, as well as Workflow looking into implementing artificial intelligence in the future, let’s take a look at the four main types of A.I.  These are reactive machines, limited memory, theory of mind, and self-awareness.

Reactive Machines

Reactive machines are the simplest type of A.I. system.  They do not form memories or use the past experiences they have had to inform their decisions.  This type of system only uses the information presented in front of it to make its decision.  An example of a reactive machine would be Deep Blue, the chess-playing computer.  This machine is unable to analyse patterns and behaviour, but instead identifies the pieces on the board and uses those to choose the best move.

This type of machine is also unable to interact and participate in the world like a human is able to.  However, it is perceived as the most trustworthy as it is guaranteed to react the same way when presented with the same situation.

Limited Memory

An example of a limited memory system would be a self-driving car.  While this form of Artificial Intelligence is more complex than a reactive machine, it still does not have the ability to permanently store all observations.  For example, a self-driving car has a pre-programmed database of line markings, road signs, speed etc.  In this database, it can build some memories based off of previous observations/data.

Theory of Mind

Theory of Mind systems are the current phase in which Artificial Intelligence research is in.  This type of system will have the ability to create representations of the world, as well as other entities such as humans.  Systems created with an effective theory of mind will also be able to understand that people and creatures have thoughts and emotions that affect behaviour, and be able to adjust their behaviour accordingly.  Theory of Mind systems will also be able to adjust their behaviour depending on the expectations set of them.


Self-Awareness is the endgame for A.I. systems.  Self-aware systems will be able to form representations about themselves.  Before this can happen however, researchers and developers need to fully understand the concept of consciousness and then implement it into machines.

Conscious beings are aware of themselves, are able to understand their internal states (e.g., fatigue, hunger etc.) and can predict feelings of others.  This is where the Theory of Mind will help them to make inferences about things like emotions based on their own experiences.  In order to do this, researchers will need to focus on understanding memory, learning, and the influence of past experiences on present decisions.

No comments for "Artificial Intelligence" yet, why don't you let us know what you think.

Leave a Comment

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