Preparing for the Future of Artificial Intelligence (in Finance)

(Note: this is really just a quick update/share, but LinkedIn's odd 255 character limitation prevents posting it as such, so I guess it's on to this long form..)

Recently the White House issued an RFI entitled Preparing for the Future of Artificial Intelligence to recommend directions for research and determine challenges and opportunities.  

IBM's response is predictably self-promotional, but highlights a few interesting points.

In the context of the finance industry, we have for many years led in the application of Machine Learning towards fraud prevention.  Recent advances in ML will only improve those capabilities.  And as security in financial systems remain critical, we have applied ML to internal systems such as intrusion detection and prevention.

IBM enumerates a few areas for future research that are not difficult to imagine as applied to finance including decision techniques (analyzing tradeoffs, analyzing data while preserving privacy, making decisions under uncertainty, etc), data assurance and trust (both as applied to AI research but also I think as applied to financial data), and more efficient computing infrastructure (see the recent buzz about Google's own application of AI dramatically reduced their power consumption). 

Most interesting and specific from IBM's response is applied to finance:

"By lowering the cost of asking basic financial questions, cognitive systems promise to create more financial inclusion by lowering the wealth threshold that people must cross to access financial advice. An advisor can conceivably increase their client workload by one or two orders of magnitude, by offloading many of more mundane and trivial information retrieval tasks they perform, and in doing so open them up to serving many people that otherwise could not afford that advice."

We already know from recent digital banking surveys that a majority of customers want their bank to proactively recommend products for their financial needs (in contrast, I suspect, to simple 'advertising').  They are telling us that even though they're increasingly self-serve, mobile, and low-touch, they are also open to financial advice, and that just might be something that AI can do fairly well.