History of AI: Part Three — AI Winter (70s)

Published in
3 min readMar 15, 2024


Artist illustration of AI in a 70s art style.

As we step into the 1970s, the story of Artificial Intelligence enters a phase characterized by profound self-examination, assessment and adjustment. This era was frequently regarded as a complex segment in the history of AI and was defined by a mix of critical skepticism and inventive thinking.

The early 1970s rode on the wave of high expectations set by the 60s, with researchers hopeful about the ability of machines to match human intelligence. However, the mid-70s brought a swift reality check. The AI community soon recognized the journey to intelligent machines was more complex than anticipated.

A major obstacle was the limited computing power available. Computers lacked the processing speeds and memory needed for AI’s ambitious goals. For example, early natural language processing efforts were hindered by limited memory, which restricted program vocabularies. This period highlighted the need for greater computational power to achieve true machine intelligence.

Researchers also faced issues like intractability and the problem of combinatorial explosion. Many AI problems had an overwhelming number of potential solutions, making it tough to find optimal solutions quickly. This was especially problematic for AI programs dealing with smaller, controlled problems but struggling in practical, larger-scale applications.

The 70s also brought forth a deeper understanding of commonsense knowledge and reasoning. It became apparent that AI applications, such as vision or natural language processing, required vast amounts of information about the world — much more than what was initially anticipated or than what could be feasibly programmed into a machine.

One of the paradoxes that emerged during this period was Moravec’s paradox, highlighting a counterintuitive aspect of AI. It became clear that tasks requiring high-level reasoning, such as playing chess or solving mathematical problems, were easier for computers than the sensory motor skills that humans and animals perform effortlessly.

Funding Crisis

As the decade progressed, AI research funding began to wane. Disillusionment with AI’s slow progress and the realization of its complexities led to reduced financial support. This phase (dubbed AI Winter) saw a significant drop in funding in AI research.

The 1970s were still pivotal for AI. This era saw advancements in logic programming and enhanced commonsense reasoning. Researchers developed new logics and algorithms, pushing past traditional methodological barriers. The decade marked a period of deep contemplation and reassessment in AI, balancing early excitement with much-needed pragmatism. This era played a pivotal role in directing the future course of AI, emphasizing the need for consistent progress, cross-disciplinary efforts, and more sophisticated computing resources.

Reflecting on this decade, it’s clear that it laid the essential groundwork for AI’s significant revival and evolution in the years that followed. The experiences and insights gained during the 1970s were key in steering AI towards more grounded, focused, and attainable objectives, setting the stage for the impressive advancements we witness in the present.

AI’s Resilience and Future Direction

The early chapters of AI’s story go beyond technological advancement, encapsulating a tale of human ambition, relentless pursuit of knowledge, and a drive to break traditional boundaries. The mid-20th century was a crucial time that fostered the growth of today’s AI, blending achievements with challenges.

The story of AI mirrors the creativity and intellectual capacity of humans. Moving forward, we will explore how AI is progressively becoming a vital component in different areas of life, including healthcare, education, entertainment, and personal productivity. The post 1970s show the revival of neural networks, the emergence of personal computing, and the advent of the internet era, each significantly contributing to shaping the contemporary landscape of AI.




Build, deploy and monetize AI apps and services.