Chinese search giant Baidu has hired Andrew Ng, the noted artificial intelligence researcher and co-founder of online education startup Coursera, to be its chief scientist, as it kicks off a new Baidu Research initiative with labs in Beijing and Sunnyvale, Calif., expecting to hire 150 to 200 people by the end of 2015.
Baidu had previously created a group in Ng’s specialty area of deep learning, which is the incredibly trendy approach to teaching machines how to process large quantities of data by setting them up to model human neural networks. It’s trendy because it’s useful for applications like speech recognition, object identification, recommendation engines and ad matching.
Ng worked on deep learning as a professor at Stanford and created the Google Brain team, which famously got 16,000 computers to train themselves to recognize images of cats in videos, simply by making them watch a bunch of YouTube.
He had been on leave from Stanford and left Google while starting Coursera, which is the largest massive-open-online-course company, with 600 free courses, seven million registered students and $85 million in funding.
Ng joins Baidu as Coursera had just hired former Yale president Rick Levin as CEO, replacing himself and his co-founder Daphne Koller. Ng moved to chairman and Koller to president in March.
Ng said today on the Coursera blog that he would “remain deeply involved in company strategy.”
Ng’s long-time collaborator and grad student Adam Coates will also join Baidu. He recently published a paper on the world’s largest trainable neural network, with more than 11 billion parameters made by just 16 machines.
Deep learning — which won’t be the only focus of Baidu Research, the company said — is an older technique, but over the past five years it has proven to be extremely effective for Internet companies that are trying to make their products more intelligent. “This trend is because of the rise of big data,” Ng said in an earlier interview with Re/code. “All of us have more access to data than we did. Deep learning is the single most powerful tool to turn this data into actionable insights.”
Microsoft said last year it has more than 65 PhD-level researchers working on deep learning, while Google bought London-based research startup DeepMind for $500 million (we had earlier reported $400 million) to add bodies to its own efforts to bring artificial intelligence to search. Facebook hired deep learning professor Yann LeCun late last year for its own AI Lab.
Join the conversation: