Sunday, March 13, 2016

Need to manage technology is more urgent than ever

Note: These are my personal views and not those of my employers, Tata Consultancy Services
I just read this article by HfS regarding TCS Ignio. I had written this article long ago regarding Ignio. HfS mention that over next few months "We’ll also shine a light on how the lessons of incident management can be ported to business process delivery to make it more intelligent". As I keenly await their view on this, I am sharing my view on this, since I think we are at the cusp of something radical. And I will take the example of insurance.
As mentioned in this article and this article, the competitive advantage for commercial insurers lies in their ability to capture and manage the knowledge of the individual and aggregate causal matrices of losses and translating this knowledge into tools for risk survey, risk selection, risk pricing, exposure management, capital allocation, claims adjudication and reserving decisions. Risk survey and loss adjustment data are static secondary sources of potential and actual loss data respectively, while IoT data can be the primary source of potential and actual loss data. Just imagine, if it were to be possible to apply a technology first approach to gather the knowledge needed to support the above types of tool, where will that take insurance to? Just like automated underwriting and claims processing has change the scenario in general insurance, will the same happen in commercial insurance too. Training new underwriters and claims adjusters used to be simpler before, but is no longer so.
If AI "tools" do more and more of the "use-cases", what will the role of humans in business processes be? Clearly it will be to handle the most complex, never-seen-before "use-cases". But how do we train humans to do this. And how do we "test" and "control" the AI over time, whether it remains correct and/or accurate. As humans lose knowledge and AI contains more and more of it, will we lose the ability within humans to make good quality decisions without AI? As we become more and more dependent on AI, will we lose the ability to survive without AI? I speculated on a lot of these fearful scenarios in 2013 in my blogs on blogger. Never felt that I would start thinking of those scenarios coming real so soon.
It is important to think long and hard about how we manage technology going forward to make sure that we survive and remain in charge over our destinies.
Regards
Pratap

No comments: