How Spark's creator is trying democratize AI with the DAWN project

Matei Zaharia has been an instrumental figure in the world of distributed computing and big data over the past decade, helping create Apache Mesos and Apache Spark, and then co-founding Spark-based startup Databricks, where he's still CTO. Now a Stanford professor, as well, Zaharia is working with colleagues Peter Bailis, Kunle Olukotun and Christopher Ré on a new project called DAWN, which aims to make it easier for everyone to take advantage of artificial intelligence. In this episode of the ARCHITECHT Show podcast, Zahara and Bailis talk about what DAWN is, the projects it includes and the specific problems they're trying to solve. The interview touches on a number of issues related to AI and machine learning, including: the difficulty of obtaining quality data (and people) to train AI models; how AI can build on the lessons and technologies of the big data era; the importance of building end-to-end systems; the differences between doing AI on the web versus in a field like health care; and why academia and open source are great places to tackle tough challenges. This week's episode brought to you by: In the news segment, co-host and Barb Darrow (Fortune) and I talk all about Amazon, and how its $14 billion purchase of Whole Foods is having competitive repercussions even in its cloud business.