AI’s Looming Ascendance Requires a Global Response

It was wonderful to see the article on artificial intelligence in the March/April edition (“Preparing for Artificial Intelligence Before It’s Too Late”). Although not a “traditional” actuarial topic, AI falls into the broader “societal risk” category—an area that the profession has historically been less inclined to take on, to put it mildly.

As Mr. Benjamin’s article points out, there are many variables, making it a complex job to get a good grasp of how AI will unfold and evolve, and the consequential societal impact thereafter.

The business imperative for growth and profits is an important fuel for AI development. We have seen this with technology sophistication and its disruptive impact to established business models in such industries as entertainment, news media, the “sharing economy,” etc. AI is already taking this a step further. Uber, for example, is testing driverless cars; IBM is linking up with other organizations looking to apply Watson’s capabilities in medical, tax preparation, and other businesses. This trend suggests a difficult future for labor—both blue and white collar work.

A second fuel for AI growth is the military ambitions of nations. An aggressor nation with superior AI capability could gain significant control over other nations and their resources. This too is a powerful force—not only as aggressor but also in deterring and containing. We have examples even today of relatively smaller nations that exhibit outsized military ambitions.

Perhaps a good starting point to understand the AI future is through scenario-building. By developing and exhaustively exploring scenarios in multidisciplinary format, involving AI/tech subject matter experts, historians, sociologists, economists, and others, some of the opportunities and dangers of increasing sophistication of AI can be fleshed out and brought before global leadership in order to make a persuasive case for worldwide partnering and agreement on policy options (many of which may be hard to sell)—something along the lines of what was done for nuclear nonproliferation, chemical weapons, etc., maybe?

I feel that actuaries, with their skills in risk evaluation, scenarios, and their long-term thinking, have a role to play—in forums with the other disciplines. However, if left to itself, there will be considerable inertia for any group to dive into this important topic, which is both important and urgent—with potentially significant societal implications. Leadership to take on this topic is badly needed.

Shiraz Jetha, MAAA, FSA, CERA
Olympia, Wash.

Medical Costs and the Price of Eggs

A marvelous article by Mark Troutman and Debbie Stubbs (“Catastrophic Medical Claims: Trends Impacting Managing Risk Associated With Value-Based Reimbursement,” January/February 2017), but the perspective regarding “today’s cost” for eggs, milk, and oranges—and also medical care—is “apples and oranges” and unfortunate.

Today’s eggs are the same as those in 1945—maybe those were even better, as the hens had access to real ground—but they are still just eggs, and oranges are still just oranges, albeit with a different load of pesticides. The Troutman/Stubbs article very effectively lays out that the practice of medicine today is very different—and enhanced—from that in 1945, when genetic testing and a thymus transplant were not even a gleam in anyone’s eye. So comparing the change in the cost of medical care to that of a basket of consumer goods is inappropriate and misleading.

For the sake of credibility, we actuaries need to be careful when trying to make a case; perhaps in this circumstance a comparison of the cost of an aircraft or automobile, both of which have seen significant enhancements, would be more illustrative.

Joan Ogden, MAAA, FCA
Salt Lake City, Utah

Letters

Letters

Letters