AI Explained

Chapter 6 Generative AI

For many years, AI researchers aimed to mimic human skills and behaviours in the widest sense. They dreamed of creating machines that could think, reason, and interact like humans across all domains. However, as the challenges mounted, the costs increased, and the need for funding grew, their focus shifted towards narrower, more specific outcomes. These targeted objectives were more likely to succeed and lead to commercially viable solutions.

AI was increasingly being employed effectively across a wide range of industries, delivering both practical utility and significant financial returns. The broader goal of emulating all aspects of humanity gradually and quietly receded into the background.

However, throughout the early part of this century, several factors combined to create new opportunities for exploration. Digital data abundance, computational availability, ubiquitous connectivity, and advanced algorithms all combined to create a perfect storm of opportunity.

Suddenly, and perhaps even to the surprise of the researchers themselves, this convergence led to breakthrough advances, enabling AI to achieve feats previously only dreamt of. The early goals of human-like intelligence came back into view. Now AI systems were creating outcomes previously thought in the realm of humans only.

Overview

Let’s begin with a concise overview of the core ideas, techniques, and motivations behind generative AI. We’ll introduce how generative approaches enable systems ...