Artificial Intelligence Explained

AI is the ability of a machine to execute cognitive functions we usually associate with human minds, such as reasoning, learning, perceiving, interacting, problem-solving, and creativity. The most common and rudimentary applications of AI most people will have experienced are voice assistants like Siri and Alexa as well as some customer service chatbots. Generative AI like ChatGPT and DALL-E have grown in popularity and become poster children for AI capabilities over the last few years, but AI technologies are diverse with a range of applications.

Modern iterations of AI that we see and use can be classified into two types:
 • Capability based AI
 • Functionality based AI

Capability Based AI

 Capability based AI includes the following:
  • Artificial Narrow AI
  • General AI
  • Super AI

Artificial Narrow AI

Artificial Narrow Intelligence, also referred to as Narrow AI or weak AI is the most common and only type of AI that exists today. It can be trained to perform a single or “narrow” task faster and better than a human mind. Its capabilities however are limited to its defined task and the AI cannot perform outside of it. Digital assistants like Siri and Alexa are common examples of Narrow AI. The current iteration of ChatGPT is also classified as a Narrow AI since it is unable to understand and perform tasks outside of its programmed capabilities.

General AI

Artificial General Intelligence (AGI) commonly referred to as General AI is a purely theoretical AI with human-like capabilities. Researchers predict that someday AGI might rival human-like cognitive abilities like perception, reasoning, problem-solving, learning and language comprehension. It is hypothesized that when AI abilities are indistinguishable from humans it will pass the Turing test I mentioned earlier. To date, no AI tool including ChatGPT has passed the Turing test.

Super AI

Super AI as of now is the stuff of science fiction. Artificial Superintelligence, much like General AI is a theoretical concept that, if realized, would lead AI to think, learn and make judgements with cognitive abilities surpassing that of humans. This is what dreams and nightmares are made of, it could hypothetically transform humanity altogether or cause a Terminator scenario where AI takes over the world. Whether it becomes reality or not is up for debate but it makes for very entertaining sci-fi.

Functionality Based AI

1. Reactive Machine AI

Reactive Machine AI systems have no memory and are used for very specific tasks working with presently available data. They use statistical principles to analyze vast amounts of data and produce intelligent responses and solutions. One of IBM’s supercomputer AI beat a chess grandmaster in the 1990s by analyzing chess pieces on the board and predicting the probable outcomes of every available move.

2. Limited Memory AI

Unlike Reactive Machine AI, Limited Memory AI can recall and process past and present moment data to process and present outcomes. It can use past data for a specified amount of time but cannot retain data as past experiences to use over a longer period. If trained on more data over time, limited memory AI can improve, consider ChatGPT 3.5 vs ChatGPT 4. By training on a larger and more diverse dataset, GPT4 has considerably improved its capabilities. Other examples of Limited Memory AI are Siri, Alexa, Microsoft Copilot and Tesla’s self-driving cars.

3. Self-Aware AI

Back to sci-fi. Self-aware AI as the name suggests, would be a fully functional super AI that would have its own emotions, needs and beliefs. It would be fully aware of its traits and be able to understand comprehend and respond to human emotions and thoughts. It would essentially have every human capability but instead of flesh and bones, it would be metal and processors.

AI Applications

Rapid development in technologies like Machine Learning and Neural Networks has allowed AI to make great strides from being a conceptual framework of ideas to practical applicability. It is no longer a buzzword and is slowly becoming an integral part of everyday lives, from generating ideas to optimizing business processes, AI is influencing sectors like healthcare, security, manufacturing, customer service, content creation, education and finance among many others.

To give you an idea, a recent Forbes survey of CEOs and business leaders found that nearly half of them are using AI for cybersecurity and fraud management, while two-thirds believe that AI will improve customer relations. Almost all of them (97%) believe Generative AI like ChatGPT will help their business and one in three are already using ChatGPT to write web content for their businesses. A brief look at how AI is being applied in different sectors is summarized below.