By Ravi Srivatsav, NTT Innovation Institute, Inc.
Recently I attended the AI World Conference and Expo 2016 in San Francisco and participated on a panel discussion on Machine Learning in the Enterprise. Along with thought leaders and pioneers in the AI space from Yahoo Research, Facebook, Intel Machine Learning Solutions, deepsense.io, NVIDIA and CodiLime – we had a charged debate on how and why the enterprise needs to develop a new way of thinking about and engaging with machine learning today to build a competitive advantage for the future.
To begin the conversation, we agreed on the framework of how we see the world of machine learning and its diverse eco-system of hardware, platforms, software and algorithms.
“Autonomous processes for driving insights out of massive quantities of data that, in turn, can help the enterprise deliver more personalized experiences.”
Many enterprise companies are still struggling with understanding and implementing the most basic forms of “digital transformation,” so it is not surprising that there are difficulties in understanding and using machine learning technology. We are still in the early stages for many reasons:
- Data scarcity, annotation, and quality
- Data sensitivity
- Operational challenge of managing investments in machine learning
- Deployment at scale versus experiments within enterprise labs