(upbeat music) >> Hi, this is Peter Burris with another Wikibon Action Item Quick Take. Jim Kobielus, where are we with the next step in AI and deep learning? >> Yeah, there's a big to-do going on, a big buzz around reinforcement learning. I think that will become ever higher for working data scientists building applications for robotics and the analytics. And so what I'm encouraged by is the fact that there are open frameworks coming into being, development frameworks, for reinforcement and learning that are integrated to some degree with the investments companies are making on deep learning. In particular, a fair number of frameworks now support TensorFlow or integration with TensorFlow. So I advise our listeners, especially the developers, to look at and evaluate frameworks like TensorFlow Agents. Rate our LLB Global School, Machine Learning Agents in coach. These are not well-known yet, but I think these will become, at least one of these will become a standard component of the data scientist workbench for building the next generation of robotics and other applications that are used for adaptive control. >> Excellent, Jim. This is Peter Burris with another Wikibon Action Item Quick Take. (upbeat music)