When Do Robots have Free Will? Exploring the Relationships between (Attributions of) Consciousness and Free Will

Abstract

This is an open access chapter distributed under the terms of the CC-BY-NC 4.0 License.

In AI We Trust: Ethics, Artificial Intelligence, and Reliability

“…AI may be programmed to have motivational states, but it does not have the capacity to consciously feel emotional dispositions, such as satisfaction or suffering, resulting from caring, which is an essential component of affective trust. An agent must be able to feel dispositions resulting from their actions; they must have mental states that are necessary for caring (Nahmias et al 2020). If an agent cannot consciously feel anything, then it would be difficult to say anything matters to that agent, even if it can carry out actions similar to ours.…”

“…Decisions made by AI do not matter to it. It does not have the ability to care about or be moved by the trust placed in it (Nahmias et al 2020). While it may fulfil what the trustor is entrusting it with, it would certainly not because of any affective reaction towards the trustor.…”

Consciousness, Phenomenal Consciousness, and Free Will

“…While the empirical evidence surveyed in the previous section suggests that the concept of phenomenal consciousness is not part of folk psychology, another prominent body of work in experimental philosophy seems to paint a different picture. A large body of research has sought to understand the ordinary concept of free will, and some of this work suggests that it bears a close relationship to judgments about consciousness, with some interpreting this in terms of phenomenal consciousness (e.g., Shepherd 2012Shepherd , 2015Nahmias et al 2020). In this section we'll survey this research, arguing that while it does indeed indicate a connection between the ordinary concept of free will and some notion of consciousness, it is at best unclear that that notion is phenomenal consciousness.…”

“…A growing body of evidence, however, suggests that non-philosophers do not tend to employ this philosophical concept (e.g., Machery 2010, Sytsma andOzdemir 2019). This work shows that non-philosophers do not tend to categorize mental states in the way that philosophers do, distinguishing between those states that are phenomenally conscious and those that are not.At the same time, another body of evidence has been taken to indicate that non-philosophers typically treat the possession of phenomenal consciousness as a necessary condition for freely willed action (e.g., Shepherd 2015, Nahmias et al 2020. But if non-philosophers don't employ the concept of phenomenal consciousness, then they couldn't treat phenomenal consciousness as a necessary condition for free will; and if they do treat it as a necessary condition, then researchers arguing they don't employ such a concept must be mistaken.1 Forthcoming in Advances in Experimental Philosophy of Action edited by Paul Henne and Samual Murray.…”

“…At the same time, another body of evidence has been taken to indicate that non-philosophers typically treat the possession of phenomenal consciousness as a necessary condition for freely willed action (e.g., Shepherd 2015, Nahmias et al 2020. But if non-philosophers don't employ the concept of phenomenal consciousness, then they couldn't treat phenomenal consciousness as a necessary condition for free will; and if they do treat it as a necessary condition, then researchers arguing they don't employ such a concept must be mistaken.…”