Do insurers collect enough data to warrant the use of AI, or are they better off concentrating on analytics?  If they don't have enough data now, should they start new initiatives to collect data for the future?  

Some of insurance's strongest assets are it's aged data.  Few industries have been around as long as insurance and this depth of data is great for predictive analytics but is it broad enough to leverage true benefit from AI? 

Maybe the answer is to focus on what's in front of us already and benefits both AI and analytics - data quality!