Foundations of AI: Module 1 – Introducing Artificial Intelligence

Is Artificial Intelligence (AI) a recent development?

Undoubtedly, Artificial Intelligence (AI) has taken centre stage as the most widely discussed topic over the past years. Social media, especially LinkedIn, has been flooded with posts about how AI has been evolving, especially since the launch of ChatGPT. We have seen a drastic increase in the number of conferences discussing the rise of AI and the positive and negative impact that this will have on our society in the coming years. As a result, it should come as no surprise that the introduction of ChatGPT has led many individuals to learn about AI for the first time. Some even believe that ChatGPT has discovered AI.

However, this is not the case….

Artificial Intelligence dates to ancient times, when the concept of creating human-like intelligence was explored in myths and folklore. However, the formal development of AI as a scientific field was officially established in 1956 at the Dartmouth Conference, where John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon among others, gathered to discuss the possibility of creating intelligent machines. 

I do not wish to go into the history of AI because this is not the scope of this post (a very good timeline can be found here), but it is worth noting that AI was not just recently discovered with the introduction of ChatGPT. There is a history behind AI, one that can be considered a roller coaster of successes and setbacks with what is termed “AI Winter” (i.e., periods of reduced investment, interest, and progress) experienced between the 1970s-1980s and the late 1980s-1990s.

What are the different types of AI?

The science of Artificial Intelligence has evolved over the years, and as this field has progressed, different types of AI started being created, helping researchers, practitioners, and learners to focus and specialise in specific areas of AI based on their goals and interests.

The main categories of AI are typically defined as Narrow AI (Weak AI), General AI (Strong AI or AGI), and Artificial Superintelligence (ASI). These are referred to as “Category 1” in the table below and are based on the scope of tasks an AI system can perform and its ability to generalise knowledge.

Other concepts like Reactive AI, Limited Memory AI, Theory of Mind AI, and Self-Aware AI, referred to as “Category 2” in the table below, are used to describe the various aspects of AI capabilities and functions. 

The following table gives a high-level overview of the types mentioned above.  

Category 1	Weak or Artificial Narrow Intelligence (ANI)	AI systems that do not possess any thinking ability and can only perform what they are programmed to do.  These correspond to the Reactive and Limited Memory AI discussed underneath. Examples of ANI include Alexa and Siri, as they are both programmed to do a limited set of functions.
	Strong or Artificial General Intelligence (AGI)	AI systems that can function like human beings – i.e., they can understand, learn, think, and make independent decisions like human intelligence.
	Artificial Superintelligence (ASI)	This is still a theoretical concept representing AI systems that, apart from having the capabilities of AGI, have much faster data processing and greater memory, making them the most capable forms of intelligence on earth. These systems will have abilities far beyond what humans can achieve.
Category 2	Reactive AI	The oldest form of AI systems that had minimal capabilities. These type of AI systems do not have memory-based functionality and, therefore, cannot learn from experience. Instead, they have pre-defined responses based on specific inputs. For example – Board playing programs like chess that make moves strictly based on predefined rules rather than past experience.
	Limited Memory AI	Builds upon Reactive AI with the capability of learning from experience. These systems are fed with large amounts of data to be used as a reference model for solving future problems. For example – Self-Driving cars have ‘learning’ capabilities, including the ability to identify traffic signals, civilians, and car speeds to avoid incidents.
	Theory of Mind AI	Systems that can understand human mental states such as emotions, needs, beliefs, thoughts, and intentions. This means that these systems can predict someone else’s intentions or goals.
	Self-Aware AI	As the name suggests, this refers to AI systems with self-awareness, consciousness,  and needs. Elon Musk and Stephen Hawkings have consistently warned scientists about the potential dangers of Self-Aware AI on the human race.

Whether the future looks promising with the advancement in AI largely depends on how effective and willing we are to find a balance between the advantages and disadvantages of AI. Collaborative efforts among researchers, industry leaders, policymakers, and the public are essential in ensuring that technological advancements are used for the benefit of society. People need to feel secure and protected from ‘unwanted species’ that cannot be controlled once they emerge, thus making transparency, accountability, and continuous dialogue essential for both the advancement of AI and the benefit of society as a whole.  

As the objective of this post is to offer newcomers in the field of Technology a foundational understanding, I have intentionally simplified certain aspects of the content. If you seek more comprehensive and precise information, please don’t hesitate to reach out.