Explained: Generative to conversational, understanding the many forms of AI

We are currently at Stage 1 of AI, which is Artificial Narrow Intelligence. Stage 3, where AI becomes superior to humans, is what experts are worried about

The concept of artificial intelligence (AI) has been around since 1955, but it is only recently that it has started to become mainstream with the arrival of consumer-facing tools and services such as ChatGPT.

Originally, AI was defined as a machine’s ability to perform a task that otherwise requires human intelligence. Its definition has been modified time and again over the years, but the fundamentals remain the same – powering machines with human-like intelligence. From chatbots to conversational and generative, AI is now available in various forms for people to explore and experiment with. Let’s understand AI and its various forms:

 

Artificial Intelligence

AI is classified into three categories: Artificial Narrow Intelligence (ANI), Artificial General Intelligence (AGI), and Artificial Superintelligence (ASI).

ANI: Designed to perform a single task or a narrow range of tasks, ANI is the commonly used form of AI today. Voice assistants such as Apple’s Siri and Amazon’s Alexa, and generative AI tools such OpenAI’s ChatGPT and Google’s Bard are some of the examples of ANI since these are designed or trained to perform specific tasks.

AGI: It is perceived to be the evolution of ANI with intelligence more on the level of human intellect. Since there are no examples of it, as yet, everything about it is but hypothetical. In theory, this type of artificial intelligence would be able to think and reason like humans.

ASI: This is the Stage 3 – an expected evolution of AGI and the one that those who track the space are seriously worried about even though it is not a reality yet. ASI is the type of AI that dreads people. It is touted to be superior to human intellect in all aspects, thus outperforming them in every function.

Conversational AI
 

Based on ANI, the conversational AI is a system that is trained or designed to respond to queries in a conversational manner. It uses natural language processing and, therefore, finds context in the queries and responds back in a natural way for human-like interaction experience.

Some examples of conversational AI are Amazon Alexa, Google Assistant and Apple Siri. Even chatbots are based on conversational AI models. ChatGPT and Bard AI are also considered conversational AI models, but these have created a new genre called generative AI.

Generative AI
 

Generative AI is capable of creating content such as text, images, music, videos, and even web pages. It is conversational by design but with multimodal advantage, which essentially means it recognises images and other visual forms as input besides text. ChatGPT, Bard AI and Dall-E are some of the examples of generative AI. At present, however, these generative AI tools do not support voice input.

Open-source and closed-source

When a company behind a new technology shares the source code for anyone and everyone to study it, it is called open-source. What this allows is for people to find out the possible gaps in that technology and also assess whether and how it might be using their personal data.

But when the company holds on to the source code, that’s called a closed-source technology. So far, ChatGPT, Bard and so on are closed-source, but there are unconfirmed reports that an open-source model will be launched sometime soon.