One of the pioneers of modern MOOCs, Andrew Ng, is stepping down from his day-to-day role at Coursera, though will remain Chairman of the Board at the company he co-founded with Daphne Koller in 2012. You can see Andrew’s internal memo to the company, which was posted on Coursera’s blog page, where he says:

I joined Coursera to launch MOOCs out to the world. Coursera has grown to a level where I am confident that it will continue to grow and thrive. In my continuing role as Chairman of the Board, I will work with the team to make sure that we do so. In this role, I will continue to act as a spokesperson and evangelist for Coursera, support the company in building out its Teaching & Learning team, further Coursera’s work in China, and remain deeply involved in company strategy. 

Undoubtedly, many will be surprised at this news, as the MOOC space is growing, and Coursera is the leading platform, with millions of customers, and ample VC funding. Although criticized for lacking a business model (as parodied in this April Fools Post), one popular MOOC, Intro to Marketing MOOC offered by Wharton, alone generated $500K in revenues from Signature track enrollments. Coursera also hired the recently retired president of Yale University, Rick Levin, as CEO–someone with clear education management and university relations credentials. So is this a case where investors push out a startup founder because the company needs more seasoned executive managers? Likely not, because of the huge opportunity that lured Andrew away.

I am joining Baidu as Chief Scientist to work on AI; will remain engaged with Coursera as Chairman. http://t.co/SiEgCtMGJT

— Andrew Ng (@AndrewYNg) May 16, 2014

Andrew has been hired on to be Chief Scientist at Baidu, the $54B company that controls over half of the online search market in China. Andrew will oversee three research labs, two in Beijing, and a new one to be built in Silicon Valley, with an investment of $300M over five years. The new lab will hire 200 people and focus on artificial intelligence, specifically, deep machine learning, which utilizes massive computing power to develop better unsupervised (or ‘self-learning’) algorithms–applications could likely range from better voice recognition to advanced robotics to more targeted online ads. This is an area where Google, Facebook, Microsoft, and others are competing fiercely for talent. Andrew is a recognized expert in deep learning, and this would provide him with a large amount of resources (though one must wonder what Google would have been willing to offer).

Thus, it appears that Andrew has traded one hot emerging area (online education), for an even hotter one (AI). Coursera is on a strong growth trajectory and is in the hands of a seasoned education executive, and so it is perhaps natural that Andrew would feel the pull of his original field of expertise. In a way, things have come full circle, from teaching machine learning at Stanford to 400 people, to putting the course online and reaching 100,000 people, and now moving from online learning back to machine learning.

The biggest impact of MOOCs is on continuing education 

Andrew spoke last Wednesday at an IEEE Silicon Valley event focusing on online learning, and we were able to attend. His talk, on May 14th, must have been after he had decided to join Baidu, but before it was publicly announced. He seemed a little pensive and nervous before the talk, but never having seen him talk, I assumed he was probably just not a natural public speaker. I wrote down a few quotes from him, and looking back, two of them stand out:

• The biggest impact of MOOCs is on continuing education
• In the future, colleges won’t just reach tens of thousands, but millions, and we want to be that platform

Learning of his decision to leave, these quotes seem to be imbued with more meaning: perhaps the first quote is a realization that MOOCs alone are not going to meet the utopian ideal of providing everyone in the world with a free education, and a recognition that most of the beneficiaries are educated individuals. The second quote describes a more modest, though still large goal for Coursera, but also suggests that it could take some time. This might be reading too much into this, but if it is on the right track, and it is Coursera’s fate not to change the world quickly, but to help change the world, somewhat slowly, then it may be the case that Andrew Ng can make a bigger impact, at this point in time, in machine learning than in online learning. But this should not cause us to diminish the clear, important impact he has made, and hopefully will continue to make, through Coursera.