If you work in the actuarial profession and over the past few years have been
- attending a professional conference or
- sitting in a presentation by an actuarial consulting firm or
- just googled “Actuary” and “AI”
you will have come across the following questions or variations thereof.
Most, if not all of the times I heard or saw a question like this the answer was some type of “yes, but no”, arguing that advances in AI will be a boon to actuaries yet not potent enough to replace them.
At precisely this point we would like to disagree. We do believe that it is possible to construct an Actuarial Machine capable of simulating a human’s decision-making and learning processes. While the machine would still need initial training (just like any new employee would) and human review of and sign-off on results should never be omitted (at the very least for regulatory purposes), we are certain that it will be able to take over a huge chunk of what actuaries do today. This leads to two major implications
- Large actuarial departments can either down-size to reduce costs and/or assign actuaries new tasks.
- Small and mid-size insurers who today struggle to attract sufficient actuarial resources will find themselves empowered to run and oversee their entire product spectrum with only a hand-full of actuaries, thereby significantly leveling the playing field.
To explore this potential and ultimately develop an AI capable of doing so, we founded Perasoft.