Preceden

History of Machine Learning Timeline

(1950 - Present)

The history of machine learning traces the evolution of algorithms and systems that enable computers to learn from data. Beginning with early theoretical work by Alan Turing and Arthur Samuel's pioneering checkers program, the field progressed through symbolic AI, statistical learning, and connectionism before exploding into prominence with deep learning breakthroughs like AlexNet in 2012 and the rise of large language models. Less

Go to: Timeline | Event List | Key Facts | Directory

Foundations

Hebb publishes The Organization of Behavior

1949

In 1949, Canadian psychologist Donald Hebb published The Organization of Behavior, introducing a theoretical neural structure formed by certain interactions among nerve cells. His ideas laid early groundwork for artificial neural networks and learning theory.

Image source: Donald O. Hebb

Hebb publishes The Organization of Behavior

Arthur Samuel coins the term machine learning

1959

The term machine learning was coined in 1959 by Arthur Samuel, an IBM employee and pioneer in the field of computer gaming and artificial intelligence, giving the emerging discipline its enduring name.

Pattern recognition research continues through the 1970s

1970 - 1979

Interest related to pattern recognition continued into the 1970s, as described by Duda and Hart in their influential 1973 work, helping establish statistical approaches to classification and perception.

Image source: Pattern recognition

Pattern recognition research continues through the 1970s

Expert systems dominate AI; statistics falls out of favour

1980

By 1980, expert systems had come to dominate AI, and statistics was out of favour, shifting the field's focus toward rule-based symbolic systems rather than data-driven learning methods.

Image source: Expert system

Expert systems dominate AI; statistics falls out of favour

Algorithms and Architectures

Neural network learns to recognise 40 characters

1981

In 1981, a report was given on using teaching strategies so that an artificial neural network learns to recognise 40 characters (26 letters, 10 digits, and 4 special symbols) from a computer terminal, demonstrating practical neural learning.

Shapiro builds Model Inference System

1981

Shapiro built his first implementation (Model Inference System) in 1981: a Prolog program that inductively inferred logic programs from positive and negative examples, advancing inductive inference research.

Image source: Inductive logic programming

Shapiro builds Model Inference System

Self-learning paradigm introduced with crossbar adaptive array

1982

Self-learning, as a machine learning paradigm, was introduced in 1982 along with a neural network capable of self-learning, named crossbar adaptive array (CAA), expanding the range of learning paradigms beyond supervised methods.

AlexNet achieves breakthrough in ImageNet competition

2012

In 2012, AlexNet, developed by Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton, achieved substantially improved results in the ImageNet image recognition competition, contributing to the wider adoption of deep neural networks.

Image source: AlexNet

AlexNet achieves breakthrough in ImageNet competition

word2vec introduced for efficient word embeddings

2013

In 2013, Tomáš Mikolov and colleagues introduced word2vec, techniques for efficiently learning distributed vector representations of words from large text corpora, transforming natural language processing.

Image source: Word2vec

word2vec introduced for efficient word embeddings

Generative adversarial networks introduced

2014

In 2014, Ian Goodfellow and colleagues introduced generative adversarial networks (GANs), a framework for training generative models through an adversarial process between generator and discriminator networks.

Transformer architecture introduced

2017

In 2017, Ashish Vaswani and colleagues introduced the Transformer, a neural network architecture based primarily on attention rather than recurrence or convolution, becoming foundational to modern large language models.

Image source: Transformer (deep learning)

Transformer architecture introduced

Ethics and Society

St George's College admissions bias found by UK Commission for Racial Equality

1988

In 1988, the UK's Commission for Racial Equality found that St. George's College had used an algorithm that discriminated against applicants, an early example of algorithmic bias with real-world consequences.

Image source: Commission for Racial Equality

St George's College admissions bias found by UK Commission for Racial Equality

Vinod Khosla predicts automation of most medical doctor jobs

2012

In 2012, co-founder of Sun Microsystems, Vinod Khosla, predicted that 80% of medical doctors jobs would be lost in the next two decades to automated machine learning medical diagnostic software, sparking debate about AI's impact on professions.

Image source: Vinod Khosla

Vinod Khosla predicts automation of most medical doctor jobs

Google Photos mislabels photos of black people as gorillas

2015

In 2015, Google Photos once tagged a couple of black people as gorillas, which caused controversy and drew attention to racial bias in image recognition systems.

Microsoft tests Tay chatbot, which learns offensive language

2016

In 2016, Microsoft tested Tay, a chatbot that learned from Twitter, and it quickly picked up racist and sexist language, prompting Microsoft to shut it down within days and raising concerns about adversarial manipulation of learning systems.

Image source: Tay (chatbot)

Microsoft tests Tay chatbot, which learns offensive language

Uber self-driving car kills pedestrian in Arizona

2018

In 2018, a self-driving car from Uber failed to detect a pedestrian, who was killed after a collision, marking the first fatality involving an autonomous vehicle and intensifying scrutiny of ML safety in critical systems.

Image source: Death of Elaine Herzberg

Uber self-driving car kills pedestrian in Arizona

Study finds women hold just 16.1% of AI faculty positions

2021

According to research carried out by the Computing Research Association in 2021, "female faculty make up just 16.1%" of all faculty members who focus on AI among several universities around the world, highlighting persistent gender disparities in the field.

Google Photos still cannot recognise gorillas after label removal

2023

The gorilla label was subsequently removed from Google Photos following the 2015 controversy, and in 2023, it still cannot recognise gorillas, illustrating the lasting challenges of fixing biased image classifiers.

Image source: Google Photos

Google Photos still cannot recognise gorillas after label removal

Competitions and Applications

Netflix launches the Netflix Prize competition

2006

In 2006, the media-services provider Netflix held the first "Netflix Prize" competition to find a program to better predict user preferences and improve the accuracy of its existing Cinematch movie recommendation algorithm by at least 10%.

Image source: Netflix Prize

Netflix launches the Netflix Prize competition

Ensemble model wins Netflix Grand Prize for $1 million

2009

A joint team made up of researchers from AT&T Labs-Research in collaboration with the teams Big Chaos and Pragmatic Theory built an ensemble model to win the Grand Prize in 2009 for $1 million, showcasing the power of collaborative filtering ensembles.

Rebellion Research credited with predicting the 2008 financial crisis

2010

In 2010, an article in The Wall Street Journal noted the use of machine learning by Rebellion Research to predict the 2008 financial crisis, highlighting ML's growing role in quantitative finance.

Image source: Rebellion Research

Rebellion Research credited with predicting the 2008 financial crisis

Machine learning applied to art history research

2014

In 2014, it was reported that a machine learning algorithm had been applied in the field of art history to study fine art paintings and that it may have revealed previously unrecognised influences among artists.

AlphaGo defeats professional Go player without handicaps

2016

In 2016, AlphaGo became the first computer program to defeat a professional human Go player without handicaps on a full-sized board, using deep neural networks and reinforcement learning.

Springer Nature publishes first book created using machine learning

2019

In 2019 Springer Nature published the first research book created using machine learning, demonstrating ML's potential to assist in scientific publishing and knowledge synthesis.

Image source: Springer Nature

Springer Nature publishes first book created using machine learning

Machine learning aids COVID-19 diagnosis and cure research

2020

In 2020, machine learning technology was used to help make diagnoses and aid researchers in developing a cure for COVID-19, showing how ML accelerated responses to a global health crisis.

Image source: COVID-19 pandemic

Machine learning aids COVID-19 diagnosis and cure research

Hardware

Tensor Processing Units introduced

2016

Since their introduction in 2016, TPUs have become a key component of AI infrastructure, especially in cloud-based environments, providing custom silicon optimized for neural network training and inference.

Image source: Tensor Processing Unit

Tensor Processing Units introduced

OpenAI documents 300,000-fold increase in deep learning compute

2017

OpenAI estimated the hardware compute used in the largest deep learning projects from AlexNet (2012) to AlphaZero (2017), and found a 300,000-fold increase in the amount of compute required, with a doubling-time trendline of 3.4 months.

Image source: OpenAI

OpenAI documents 300,000-fold increase in deep learning compute

GPUs displace CPUs as dominant method for training cloud AI

2019

By 2019, graphics processing units (GPUs), often with AI-specific enhancements, had displaced CPUs as the dominant method of training large-scale commercial cloud AI, accelerating deep learning development.

Image source: Graphics processing unit

GPUs displace CPUs as dominant method for training cloud AI

Key Facts

  1. In 1950, Alan Turing proposed the 'Turing Test' in his paper 'Computing Machinery and Intelligence,' laying the philosophical foundation for artificial intelligence and machine learning.
  2. Arthur Samuel coined the term 'machine learning' in 1959 while developing a checkers-playing program that improved through self-play at IBM.
  3. The 1986 popularization of backpropagation by Rumelhart, Hinton, and Williams revived interest in neural networks after years of stagnation following Minsky and Papert's critique of perceptrons.
  4. In 2012, AlexNet's victory in the ImageNet competition demonstrated the power of deep convolutional neural networks combined with GPUs, triggering the modern deep learning revolution.
  5. The release of transformer-based models, starting with 'Attention Is All You Need' in 2017 and culminating in large language models like GPT and BERT, has driven machine learning's integration into everyday applications.

Source

This History of Machine Learning timeline was generated with the help of AI, using information found on the internet.

We work hard to keep these timelines accurate, but mistakes do get through. If you spot one, email us at [email protected] and we'll fix it for future visitors.

Want a timeline like this one?

Generate yours with AI or build it from scratch, for free.

Try It for Free