The intertwining of humans and machines is ever increasing, with integrated human/machine ‘teams’ under development and even being deployed. Challenging is how to replicate interactions between humans and machines, given how we now humans interact with one another. Humans constantly value intermediate decisions with respect to context through internal models of their confidence, expected reward, risk etc, before they generate a behavior. Such information about human decision-making is expressed not just through behavior, such as speech or action, but more subtlety through physiological changes, small changes in facial expression, posture etc. These are cues that human beings utilize to infer the current disposition of one another. Socially and emotionally intelligent people are excellent at picking up on this information and using it to guide their decisions and interactions with team members. The ability to pick up on these cues is important for developing trust bonds between team members, enabling just-in-time prediction of team member states and needs. In this talk I will present some of the work we are doing, using non-invasive physiological sensing, to infer cognitive/physiologically states of humans while they interact with machines/agents. We show examples of how this state information can inform autonomy so as to improve human-machine interaction.

https://mit.zoom.us/s/97261895976

PW: 312454​​​