Specialized Topics

Generative AI

Generative AI (GenAI) is the branch of AI that focuses on generating new human-like content (text, images, speech, and videos) by learning and analyzing existing data patterns. The concept of GenAI traces back to the 1950s–60s, when the first chatbot, ELIZA, was released. There were further advancements during the 1970s–80s, such as PARRY, a chatbot designed to simulate patients with paranoid schizophrenia. In the 1990s–2000s, ML and AI gained popularity again with some practical advancements, such as the concept of personal digital assistants (PDAs) and interactive voice response (IVR). In 2010–2011, Apple launched Siri as a virtual assistant for the iOS platform. Siri could understand the natural language—Natural Language Understanding (NLU)—and respond to it in the natural language—Natural Language Generation (NLG). In 2012, Google also entered the game with Google virtual assistant. Some other examples are Cortana by Microsoft, Amazon Echo, and Alexa.

In this era, the concepts of GenAI are also shifting from PDAs or virtual assistants to other fields. Some examples include Grammarly, QuillBot, and Wordtune for enhancing the writing experience, DALL-E, Jasper, and MidJourney for generating art, and DeepBrain, Synthesia, and AI avatars for video and speech generation. All these innovative applications become a reality with the advancements of deep learning—Generative pretrained transformers (GPT), Bard, generative adversarial networks (GANs), variational autoencoders (VAEs), and stable diffusers.

GenAI finds utility primarily in the ...