AI Luminary Series: Pedro Domingos Explains Machine Learning

Artificial Intelligence Luminary: Pedro Domingos, Professor of Computer Science & Engineering at the University of Washington.

In this video, Pedro Domingos answers the question of what is machine learning, describes its relationship to AI and big data, and explains how a learning algorithm has the power to revolutionize our world.

Read the transcript:
People often ask me - what's the relationship between AI and machine learning and big data? Machine learning is the subfield of AI that deals with getting computers to learn. So you can think of AI as the planet that we're going to, and machine learning as the rocket that will get us there, and big data as the fuel for that rocket.

There are many examples of AI and machine learning at work in the world today, that touch people's everyday lives, but they aren't even aware of it. For example, every time you do a web search, when Netflix recommends a movie, when Facebook selects posts, when Amazon recommends a book, it's machine learning that's doing that. Then there are people who apply machine learning and AI in things like robotics, and vision, and natural language processing, or medicine, or oceanography, or social science, you name it.

We've gotten very far in AI in the first 50 years. There's a million miles more to go. So we're going to need a lot of compute power that is specialized for things like machine learning. I think Intel has something very important to contribute to all of this which is at the end of the day, it all starts with the hardware.

Intel is in the leading position to bring us the hardware and the architectures to try to foster this open community that we really do need to make progress.

We're actually now for the first time in history at the point where you could say you can have a supercomputer that is about as powerful as the human brain. So the thing that is really holding us back is that we don't understand well enough how, for example, learning works. If we were able to devise a learning algorithm that is truly as good as the one in the human brain, this would be one of the greatest revolutions in history. And it could happen any day at this point. At this point this is a problem that if we could solve it we solve all other problems. If I come up with a better machine learning algorithm, that algorithm will be applied in business, in finance, in biology, in medicine across the board.