These two lectures were given as part a conference at Ralston College, a privately run liberal arts college in Savannah, Georgia. Together these lectures by McGilchrist and Pageau make interesting bookends, providing two very contrasting approaches to AI, and helpfully open out an imaginatively clearheaded appraisal of the pro and cons of it.
Iain McGilchrist -Information is not Understanding
McGilchrist perceives the way AI is being sold to us as deeply flawed. Consciousness is not a bank full of a wealth of information. The distinguishing factor in human consciousness and cogitation is an ability to imagine outside of the box of knowledge we already possess, stimulated through what we bring our attention to. Directed by the motivations of our soul (our overall sense for what has meaning, of ourself and purpose in life) not a computer programme. McGilchrist applies his left and right brain hypothesis to AI, which is essentially a very left brain technology. AI deals primarily in the acquisition of already known knowledge. Wisdom or understanding, however, is not really achieved simply by the amount of information we have available. Arising more often unexpectedly as though from nowhere. He uses the analogy of our being more like hopeful gardeners, who plant, feed and water ideas without any certainty of what will grow.
Jonathan Pageau - Whom Does AI Serve?
Pageau's talk is the more direct and punchier of the two talks. He asks from the outset 'Who does AI serve'. The conversation about whether AI is conscious is a red herring in this regard. AI is a grab for power by people whose moral framework appears at best to be rather vague or non-existent, which is in itself worrying. AI also appears to have no fixed objective that we might see as being good for humanity. Where is the holy grail here? AI simply wants our attention because it extracts information from us like bloodsucking vampires. The presence of AI asks of us, to know what we ourselves are serving through our lives, and our interactions with AI programmes. Asks us to question how this aligns with our own sense for whether this is for our own or humanities good. It's not that we can set a limit on what AI is applied to. It's already out there, growing exponentially, with no apparent end in sight to its voracious hunger for factual details.
No comments:
Post a Comment