History of AI: Part IV —  The Boom (80s)

History of AI: Part Four— The Boom (80s)

The 1980s stood out as a key era in Artificial Intelligence (AI). In these years, AI evolved from a field largely confined to academic study to a crucial element within the business world, alongside major technological progress in its core foundations. The 80s saw some notable innovations that had a far reaching impact. Let’s take a closer look.

Rise of ‘Expert Systems’

At the forefront of this boom were AI programs called ‘Expert Systems’. These were designed to emulate human expertise in specific domains and quickly became a staple in corporations worldwide. Early systems like Dendral and MYCIN, pioneered by Edward Feigenbaum and his students, demonstrated AI’s potential in practical applications. They provided solutions in confined knowledge domains, sidestepping the ongoing challenge of commonsense reasoning, and showcased AI’s usefulness in real-world scenarios.

XCON was originally developed for the Digital Equipment Corporation and was a success, saving the company millions annually. This sparked a surge in corporate investment in AI throughout the mid-80s, and consequently the rise of the burgeoning AI industry.

Knowledge Revolution

The 1980s were also a time when knowledge became a central theme in AI research. The focus shifted towards building systems that could leverage vast amounts of diverse information. This approach was a departure from earlier efforts, recognizing that intelligent behavior is deeply rooted in handling detailed, domain-specific knowledge. Projects like Cyc attempted to tackle the common knowledge problem by creating extensive databases of everyday facts.

AI also made headlines in the world of chess. Programs like HiTech and Deep Thought, both precursors to the famous Deep Blue that began defeating skilled human players, signaling AI’s growing capability in specific, complex tasks.

Global Investments and the Revival of Neural Network

During the 1980s, investment in Artificial Intelligence significantly increased worldwide. A prominent instance was Japan’s Fifth Generation Computer Project, which aimed to develop computers capable of human-like thinking and comprehension. This endeavor influenced comparable projects in other nations, including the UK and the US, leading to increased financial support for AI research.

A key event of this time was the renewed interest in neural networks. John Hopfield’s 1982 research demonstrated the effective learning and processing capabilities of neural networks. This discovery, along with Geoffrey Hinton and David Rumelhart’s work on backpropagation, rejuvenated neural network research, setting the stage for later achievements in areas like speech and optical character recognition.

In terms of technology, advancements in metal–oxide–semiconductor (MOS) technology, especially CMOS, significantly impacted AI. This technological progress facilitated the practical application of neural systems, as evidenced by the seminal 1989 work “Analog VLSI Implementation of Neural Systems.”

However, as with every rise there comes a fall and the late 1980s saw a decline in the AI sector. Overhyped expectations led to an economic bubble in AI which eventually burst, resulting in a phase often referred to as the second ‘AI winter’. The specialized AI hardware market plummeted, and the upkeep of early expert systems became unsustainable, both financially and technically. This downturn caused drastic reductions in AI funding and raised doubts about the field’s commercial prospects.

New Directions

The field continued to evolve albeit at a steadier pace and newer approaches to AI, focused on robotics and sensorimotor skills, gained traction. These ideas, emphasizing the importance of physical interaction with the world, were a significant departure from previous AI paradigms. Experts such as Rodney Brooks supported a fundamental approach to understanding intelligence, challenging the exclusive focus on symbolic processing.

The 1980s were marked by notable progress, commercial ventures, and a period of reevaluation. These developments played a key role in guiding the direction of AI, paving the way for its growth into the 1990s and beyond. Looking back at this eventful decade, it’s a tale of persistence, flexibility, and an ongoing effort to comprehend and replicate intelligence.

Stay tuned for the next chapter in The History of AI where we dive into the 1990s and advances in data mining, robotics and speech recognition.

History of AI: Part IV —  The Boom (80s) was originally published in Fetch.ai on Medium, where people are continuing the conversation by highlighting and responding to this story.