The History of AI

Google Brain is born

Era 3 · From the expert boom to the statistical turn (1980–2011) › Hardware and data

Lineage
Machines and computation
Kind
Institutional

Who

Google

What happened

Jeff Dean and Greg Corrado of Google and Andrew Ng, a Stanford professor, start a project at Google X to train enormous neural networks on the company's computing infrastructure. It will become Google Brain.

Why it mattered

A big tech company bets on deep learning at a scale no university could afford. It marks the start of AI research inside the tech giants.

The full story

In 2009 Ng and a group of students had shown that GPUs greatly sped up the training of deep networks and allowed much larger ones. Google had what universities lacked: tens of thousands of computers and enormous amounts of data. Jeff Dean, one of the engineers who had built that infrastructure, took an interest in the idea.

In 2011 Dean, Corrado, and Ng began working together at Google X, the company's experimental projects division. The project brought machine learning research together with Google's large-scale information systems and computing resources.

The first result was published in December 2011. A network with a billion connections, trained on ten million unlabeled frames from YouTube videos, developed on its own a neuron that responded to human faces and another that responded to cats. The training took a thousand machines three days. It became known as the cat experiment.

Google Brain became one of the major AI labs. It produced TensorFlow and the Transformer, and in 2023 it merged with DeepMind to form Google DeepMind.

Sources

Original source

How to cite

“Google Brain is born.” The History of AI, revised November 13, 2026. https://thehistoryofai.com/en/milestone/google-brain-founded

This entry was last reviewed on November 13, 2026.

Open in the interactive timeline