thanks, David, for ypur clear formulation of "AI agency" and conscious acts in assisting humans deal with today's complex challenges
John Raymaker
do you think that AI has agency
through conscious 'acts of understanding'
(as we've been attributing to Lonergan's notion of human understanding)?
I think the short answer is no, in the sense that it is not conscious and has no agency. However, it does display what we may term a virtual act of understanding, which can lead to an actual act of understanding once a connection is drawn between the meaning of the understood act, and its relation to an element of concrete reality.
In humans, this link is mediated by our perception of the external universe, and systematised by the collaborative scientific enterprise.
These virtual acts of understanding can therefore be very helpful, not for AI to perform conscious acts, but to assist humans in rising to the complex challenges posed by our existential situation, i.e. they can help us understand.
The role of Lonergan could therefore be to act as a facilitator to humans for using AI intelligently, because it is conditioned by our ability to interiorise the knowledge so gained.
Best wishes,
David