A short history of AI
AI research began in the 1950s. Big steps came when computers got faster, data got bigger and learning methods improved, leading to today's chatbots.
📌 Key points
- In 1950 Alan Turing asked "Can machines think?" He suggested a test, now called the Turing test: if a person chatting by text can't tell a machine from a human, the machine passes.
- The name "artificial intelligence" was coined by John McCarthy for the Dartmouth workshop in the USA in 1956. It is often called the birth of AI as a field of study.
- Early AI used rules written by hand. Progress was slower than people hoped, and there were times when interest and money for AI research fell. These are called AI winters.
- In 1997 the chess computer Deep Blue beat the world chess champion Garry Kasparov in a six-game match. It searched through millions of possible moves every second.
- In 2012 a deep neural network won the ImageNet picture-recognition contest by a big margin. This started the deep learning boom, helped by more data and fast chips called GPUs.
- In 2016 AlphaGo beat Lee Sedol, one of the best players of the board game Go, by 4 games to 1. It learned much of its skill from playing games.
- In 2017 researchers introduced the transformer, a new design of neural network. Today's large language models are built on it.
- In November 2022 the chatbot ChatGPT was released to the public, and millions of people started using AI chatbots.
📖 Key words
- Artificial intelligence (AI)
- Computer systems that do tasks that usually need human intelligence, such as recognising speech or making decisions.
- Turing test
- A test where a person chats with a hidden machine and a hidden human and tries to tell which is which.
- AI winter
- A period when interest and funding for AI research dropped.
- Deep learning
- Machine learning that uses neural networks with many layers.
- GPU
- Graphics processing unit: a chip that does many calculations at the same time, which makes training much faster.
- Transformer
- A neural network design from 2017 that today's large language models are built on.
✏️ Worked example
AI ideas existed in the 1950s, but AI improved very quickly after about 2010. Why?
- More data: the internet, phones and digital photos created huge amounts of examples to learn from.
- More computing power: GPUs could do millions of calculations at once, so big networks could be trained in days instead of years.
- Better methods: deep learning (2012) and transformers (2017) learned much better from all that data.
- Answer: data, computing power and better learning methods came together at the same time.
Remember the dates as a story: 1950 a question (Turing), 1956 a name (Dartmouth), 1997 chess, 2012 deep learning, 2016 Go, 2017 transformers, 2022 chatbots for everyone.
Test yourself8 questions from a bank of 19. A new mix every time.
Start the quiz →