Andrew Ng on Deep Learning and Technological Unemployment

 Deep Learning  

At the RE.WORK Deep Learning Summit last month, Baidu's Chief Scientist, Andrew Ng discussed how the technology is progressing, and also touched on the problem of technological unemployment artificial intelligence is leading to.

Andrew Ng, Baidu's Chief Scientist is developing deep learning applications that are helping power the Chinese search engine giant.

At Baidu, Ng is currently building high performance computing (HPC) capability for the company. "I think that approach has allowed us to scale up our deep learning algorithms pretty aggresively," he told an audience at the RE.WORK Deep Learning Summit earlier this year.

"Traditionally, supercomputers were used to do atomic bomb simulations or whatever and that's okay, but now supercomputer technology, the GPU processors, fast interconnect, latency detection, the crazy complicated scheduling work that is done...that is having a huge impact on AI, which I think will change the world," Ng says.

Ng addresses the over-hype surrounding deep learning too. "Baidu has a lot of products and services that use deep learning," he says. "I think that's a nice sweater.  If I take a picture with my cell phone, Baidu has a relatively unique feature that will use deep learning to recognize the sweater and try to where to buy a similar sweater."

Related articles Baidu Relying on Deep Learning to Make it Easier to Talk to Your Devices  Stanford Researchers and Google Create World's Largest Artificial Neural Network  Biological Transistors Created Inside Cells Out of DNA

Some of Baidu's advertising is also powered by deep learning.  Ng also tells of cases where the technology is used internally to improve Baidu's operations.

As the discussion continues, Ng addresses superintelligence.  Almost all deep learning used today uses supervised learning, he mentions.  "Deep learning is amazing at supervised learning, but it is a very narrow task." He states that he expects a lot of development in unsupervised learning for the systems in the future.

"Having said that, there is one worse case that does concern me," Ng continues. "The impact of technology on labor." Traditionally new technology empowered people and freed them from drudgery. As Ng, and many others now point out though, "There is something different about this time."

"There is one worse case that does concern me: the impact of technology on labor. There is something different about this time."

Compared to the transition from agriculture during the industrial revolution, Ng says the next transition may be much faster, and the education system will not be quick enough to adjust to the changes. "Our education system has historically found it very difficult to retrain people that are already alive today to do a different job skill, as opposed to training their descendants."

According to Ng MOOCs like Coursera offer some of the best tools for this transition.  However, he does say that it deeply concerns him that this may not be enough to train so many people.  "Evil killer robots in the world is an unfortunate distraction from that [problem]," he says.

In the question period, Ng also clarifies that deep learning and neuromorphic computing are not absolutely tied together. According to Ng, neuromorphic computing is significantly lagging the advances in deep learning.

Ng leads Baidu Research, which comprises three interrelated labs: the Silicon Valley AI Lab, the Institute of Deep Learning and the Big Data Lab. The organization brings together global research talent to work on fundamental technologies in areas such as image recognition and image-based search, speech recognition, natural language processing and semantic intelligence.

In addition to his role at Baidu, Dr. Ng a faculty member in Stanford University's Computer Science department, and Chairman of Coursera, an online education platform that he co-founded. Dr. Ng is the author or co-author of over 100 published papers in machine learning, robotics and related fields. He holds degrees from Carnegie Mellon University, MIT and the University of California, Berkeley.


SOURCE  RE.Work

By 33rd Square

Embed