This website uses cookies and similar technologies to understand visitors' experiences. By continuing to use this website, you accept our use of cookies and similar technologies,Terms of Use, and Privacy Policy.

Dec 22 2018 - 07:06 AM
"Explainable" AI

One of the challenges of implementing AI is avoiding the creation of black boxes where processes and operations become opaque, even to AI owners, let alone clients, customers, and regulators.  This is particularly important if we are to avoid bias or the perception of bias in operations and delivery of services. In this ZDNet article, Nitzan Mekel-Bobrov, the chief of AI at Capital One, argues for what he calls "explainable" AI that can be achieved by teasing apart the inner-workings and making them interpretable. This is useful for us to keep in mind as we approach AI projects. Be sure to read to the end for his thoughts on how advances in mobile will be necessary for certain types of processes to improve customer interaction.

|By: Gary Natriello|1160 Reads