AIToday
Large Language ModelsZenn AI/MLPublished: Nov 4, 2026, 06:00 JST

AI since 1956: three booms, then ChatGPT in 2022

Send on LINE
AI since 1956: three booms, then ChatGPT in 2022

3 Key Points

  1. What happened

    AI began as a research field in 1956 at a Dartmouth conference, went through three booms, and entered the generative AI era after ChatGPT launched in November 2022.

  2. Why it matters

    Each boom was defined by what the leading technique could and could not do, and today's systems still show uneven ability: one test can be aced while a simple task fails.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

AI's path goes from a 1956 Dartmouth conference through three distinct booms, each defined by the leading technique of its time: search and reasoning in the 1950s and 1960s, expert systems in the 1980s, and machine learning and deep learning from the 2000s. The third boom's milestones include a deep-learning team's decisive win in a 2012 image-recognition contest, AlphaGo's 2016 victory over a top Go player, the 2017 Transformer architecture that later underpinned large language models, and ChatGPT's November 2022 launch.

There is no single established definition of AI. The OECD and EU focus on what a system does, describing a machine-based system that infers from input how to generate outputs. UNESCO declines to fix a definition, saying it changes with technology. Japan's AI law instead frames AI in relation to human abilities, defining it as technology that substitutes for human cognitive, reasoning, and judgment capacities.

Capability is contrasted with consistency. Stanford HAI's AI Index Report notes that AI now matches or exceeds human levels on many conventional tests and is improving quickly in math, programming, and computer operation, yet it still cannot reliably read an analog clock. The report also notes that scaling laws face criticism, but the compute used for training continues to rise, with performance gains still observed.

FAQ
How is AI defined internationally?
The OECD and EU define an AI system as a machine-based system that infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions.
How does Japan's AI law define AI?
Japan's AI law defines AI-related technology as technology that artificially substitutes for human cognitive, reasoning, and judgment abilities, and systems that process information and output results using it.
How much did global private AI investment grow in 2025?
Global private investment in AI reached $344.7 billion in 2025, up 127.5% from the previous year, with about $285.9 billion in the United States alone.
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleDrax and Brookfield stake £20 billion on 1GW UK AI data center