Trust, Ethics, Consciousness, and Artificial Intelligence
Symposium on Dependable Autonomic and Secure Computing September 18, 2022 Katherine L. O'Grady, Steven Harbour, Ashlie Abballe et al.
As artificial intelligence expands into many sectors, ethical guidelines are needed to protect both humans and potentially conscious autonomous agents. Current AI lacks context and true understanding, but future artificial general intelligence (AGI) might develop dynamic cognizance similar to humans. If conscious agents are assigned grueling tasks, outcomes could include retaliation or breakdown of human-autonomy teams. Trust and ethical consideration between humans and machines are problematic because these concepts are defined only in human terms. This work explores consciousness from evolutionary, neuroscientific, and philosophical perspectives to position degrees of consciousness within a trust and ethics framework, aiming to spark dialogue and establish legal and operational requirements before any agent acquires consciousness.