Overcoming the AI hype — and what enterprises should really focus on

Deep learning for all its hype is brittle, non-generalizeable, and its learnings are not readily transferable from one application to another. Since we are unlikely to see anything close to Artificial General Intelligence in the next few decades., we should instead focus on how enterprises can capitalize on the state of the art in machine learning and re-implement successful algorithms and follow the data science lifecycles that generate highest ROI. This talk will cover the current state of the art in AI, its limits vs hype, and discuss concrete steps that enterprises can take to achieve desired ROI by reimplementing production-grade-ready machine learning algorithms, that have been hardened and demonstrated to work very well in specific, constrained domains. By the end of this talk, attendees should have a better grasp on how to avoid costly and unnecessary investments into yet unproven technologies, be better equipped to navigate the complex space of AI, and understand where to best focus their resources to maximize ROI.

Robert Hryniewicz has over ten years working on various projects related to Artificial Intelligence, Enterprise Software, Robotics, Blockchain and more. Currently he's a Data Scientist / Architect and an Evangelist at Hortonworks Inc. Previously he held multiple roles as a CTO @ Authentise (a 3D printing Singularity Labs startup), Sr. Software Architect @ Cisco, NASA et al. where he developed multiple products & prototypes in the Machine Learning, IoT and Robotics space. Robert is also a frequent speaker on Enterprise Big Data and AI at all major conferences in the U.S.