The Internet of Things is about data, not things. Some forecasts that by 2018 the number of connect things will exceed the combined number of personal computers, smartphones, and tablets. Each ’thing’ can produce a tremendous stream of data from sensors and other sources. This presentation will discuss progress, examples, challenges, and opportunities with machine learning for the IoT. A short presentation will be done on some recent applications of ML (using H2O) to the domains of machine prognostics / health management (PHM) and agriculture. Presented by Hank Roark, Data Scientist / Hacker at H2O.ai.