The Origin & Future of NDM Research

A few days ago, I had to go to the grocery store to buy daily necessities. There were police cars, fire trunks, and gas company's car near the shop. I immediately noticed there was a gas leak, but I didn't smell it.

When I look at fire fighters, I always remember how my specialty, Naturalistic Decision Making (NDM) was initiated. As I already introduced in my past blog several times, Gary Klein and his peers were very interested in how the expert made a judgment and decision under the life-threatening situation. They chose fire commanders as research targets.

After intensive observation and interviews, Klein concluded that the fire commanders depended on their intuition to find out the best solution. Then, he developed the cognitive model of intuition that is known as Recognition Primed Decision (RPD) model. The essence of RPD model lies in pattern recognition and mental modeling (or imagination).

Although artificial intelligence (AI) can simulate human pattern recognition and mental modeling somehow, but it cannot copy them exactly. In this sense, these two features human expert acquired through training and experience are unique.

If we NDM researchers continues our studies, there is a possibility that our findings will be applied to AI research and development. However, the purpose of AI R&D is not to create AI that surpass human ability. Rather, we need AI that support our judgment and decision making and augment our ability. This type of AI is called 'Human-centered AI' or 'Human+AI'.

In short, our NDM research has contributed to not only the understanding of the expertise (like fire fighter's judgment- and decision-making skills) but linking human and AI.