Think about the stereotypical robot-what characteristics usually comes to mind? Computational, logical, maybe even predictable? For the longest time, predictability has been a given attribute when talking about a robotic system. I remember back in high school, during JV basketball practice we would run this same specific play (it was called “Bingo”), over and over. Our coach had to remind us, however, “Guys, keep in mind, the play isn’t meant to be robotic!” He was implying certain randomness on the court can happen during a game, and in that circumstance, we would have to adjust accordingly and find a way around the defense. In other words, as opposed to being robotic, we needed to be spontaneous.
Although robotics and spontaneity have traditionally been opposing forces, scientists at the University of Tokyo have been aiming to shatter this belief and make robots spontaneous. Much of their studies fall under the class of “high-dimensional chaos”, which is a dynamical system that is used to model animal brains, which includes human brains, which are traditionally spontaneous. The high-dimensional chaos system is very complex in that it takes in a lot of parameters and can be subject to change very easily by varying conditions (think “butterfly effect”).
Doctoral student Katsuma Inoue, explains, “There is an aspect of high-dimensional chaos called chaotic itinerancy (CI) which can explain brain activity during memory recall and association.” Katsuma and her colleagues claim that through high-dimensional chaos, they have generated time-series patterns that proposes a recipe for spontaneity. A recipe for spontaneity may sound paradoxical, however, it is exciting to see what further can be developed.
The implications of spontaneous robots are very intriguing in the world of AI. It may be exciting to have robots spontaneously interacting with us. In my previous blog post I talk about motivational robots, so it would be cool to see how spontaneity could be implemented in motivation. However, It is also scary to think that robots can be spontaneous, how at any given moment a piece of technology (perhaps equipped with weapons) can suddenly make its own, unpredictable decisions.
Sources:
University of Tokyo. “Robotic AI learns to be spontaneous.” ScienceDaily. ScienceDaily, 11 November 2020. <www.sciencedaily.com/releases/2020/11/201111144419.htm>.
