Manulife’s Lab of Forward Thinking (LOFT) is undergoing efforts to accelerate its adoption of new technologies. In April, the Canadian insurer's U.S. unit
More recently, Manulife began exploring artificial intelligence and deep learning in its innovation labs through
INN: What exactly is deep learning?
Jesse Bean: It is unsupervised learning by the machine. Previously, AI (artificial intelligence) involved having to train the machine on everything new. With peak learning, as it’s also called, the machine learns on its own. Startups are concentrating on developing these new deep learning techniques.
INN: Why should insurers pay attention to it?
JB: Deep learning gives you the ability to process data faster than original machine learning. It is big a reason why we partnered with indico data. It has models that can start to learn with thousands of examples, instead of hundreds of thousands. If I feed an AI platform thousands of analysts’ articles rather than of hundreds of thousands, it takes less time to get results.
INN: How is Manulife exploring it?
JB: We are using it in the research space to be able to provide our research analysts with insights; using analysts’ reports, news feeds and social media data. Our researchers look for insights like they would on a normal basis, but with more input data to be better, active managers.
INN: How does deep learning fit into Manulife’s innovation strategy?
JB: LOFT likes to look at tech that can help and disrupt our business from a customer value adding perspective. This is the first prototype. The end goal of anything in LOFT is to develop a platform that others can use throughout the firm in any business unit.
INN: Any differences between the capabilities Nervana Systems and indico data offer Manulife LOFT?
JB: Definitely. indico is a deep learning tool while Nervana provides a platform to run models on, which is 10 times faster than its nearest competitor. It [indico] is a model data play while Nervana is a platform processing play.
INN: How does this project relate to the company’s recent efforts with blockchain?
JB: They're related only in that LOFT is working with these technologies and at some point they will connect, but not now. We do everything from a customer value perspective to solve problems in different ways with different tools.