Follow Datanami:

People to Watch 2019

Pedro Domingos
Professor
The University of Washington

Pedro Domingos is Professor of Computer Science and Engineering at the University of Washington and head of machine learning research at the D. E. Shaw Group. He is the author of “The Master Algorithm,” a best-selling introduction to machine learning for a general audience. He is a winner of the SIGKDD Innovation Award, which recognizes individuals for outstanding technical contributions to the field of data science. He is a Fellow of the Association for the Advancement of Artificial Intelligence, and has received a Fulbright Scholarship, a Sloan Fellowship, the National Science Foundation’s CAREER Award, and numerous best paper awards. His research spans a wide variety of topics in machine learning, artificial intelligence, and data science, including scaling learning algorithms to big data, maximizing word of mouth in social networks, unifying logic and probability, and deep learning.

Datanami: Professor Domingos, we’re in the midst of an AI boom, largely driven by explosive growth of neural networks and deep learning over the past five years. Will these approaches continue to give us good returns, or will other approaches be needed to keep AI moving forward in five or 10 years?

Pedro Domingos: The word on the street is that current deep learning approaches are starting to run out of steam, but I think there are many important new things still to be discovered within the general field, as well as within other machine learning paradigms. In the longer term we’ll need new paradigms, though, and what those might be is a fascinating research direction. I’m working on a candidate called symmetry-based learning: discovering invariances in data and then composing them. The theoretical underpinning is symmetry group theory, and it’s potentially able to generalize much farther and more robustly than current methods.

Datanami: You’ve also done research into Markov Logic networks. What go you interested in this field, and what potential do you see for this technology?

Markov logic networks are a way to unify logic and probability, which I believe is essential for creating truly powerful AI systems. The real world is both very complex and highly uncertain; logic can handle the former and probability the latter, so by combining the two we can avoid a lot of the hacking that goes into AI solutions that use only one or the other, and may be able ultimately to do things that wouldn’t be feasible otherwise (e.g., create home robots and worldwide knowledge bases).

Datanami: Your book, “The Master Algorithm,” was seen on Chinese President Xi Jinping’s bookshelf. Are you concerned that technology created largely in academia will become the feedstock for the next arms race?

It’s already happening, unfortunately. AI is unusually fertile in both business and military applications, and there’s no clean separating line between the two. And as with a more traditional arms race, we need to both try to minimize it and be prepared to win it – a difficult combination.

Datanami: The potential for AI to help mankind is great. Do you think we’re at risk of fear of AI taking over, and if so, how can we overcome the fear?

Yes, fear of AI is rampant, and in my view out of proportion to the real dangers. One way to help overcome this is by informing and educating people about what AI really is, so they’re no longer stuck with stereotypes such as evil robots (a la Terminator), God-like AIs, etc. AI systems are just problem solvers; they don’t have free will, consciousness, emotions, etc., and for the foreseeable future there will continue to be many things that humans can do better than AI.

Datanami: Outside of the professional sphere, what can you share about yourself that your colleagues might be surprised to learn – any unique hobbies or stories?

I’m a graduate of the Clarion West Writers Workshop, which puts me in the company of some of the best speculative fiction writers of recent decades. These days AI evolves faster than science fiction, and that’s where I’ve focused my efforts, but the training I received at Clarion West came in very handy when writing “The Master Algorithm.”

 

Sri Ambati
H2O.ai		 Julia Angwin
The Markup		 Pedro Domingos
University of Washington
Ali Ghodsi
RISELab		 Cassie Kozyrkov
Google		 Hilary Mason
Cloudera
Bob Muglia
SnowFlake		 Oliver Ratzesgerger
Teradata		 Dr. Tony Slonim
Renown Health
Kostas Tzoumas
dataArtisans		 Evan Weaver
Fauna		 Steve Wilkes
Striim

 

 

 

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13

Datanami