A chapter in a book I am working on explores how we are not machines, yet can be duped into thinking that we are. It is part of a book that centers on how a person can flourish in difficult circumstances. As I was writing it, the Generative AI, Chat GPT, and computer future fears surfaced in the media, and we have been alerted to an acceleration of human-seeming abilities of technology. I found myself concerned with issues that are arising now, rather than long term fears, and wrote a paper for the US Office of Science and Technology Policy with recommendations for action. There are many people who can speak to so many issues raised by AI, but I thought there were some things that were being left out of discussions, and my scholarship in human relationships might be helpful in making recommendations for action and making a case for those.
We need to protect real human-to-human relationships and strengthen our relationships with one another in the wider social world. What can we do to help us see ourselves as we are, not as machines, and help us to relate realistically to human-seeming Generative AI objects with various levels of verbal, visual, auditory, and tactile characteristics? How do we encourage people to see the value of human-to-human interactions, of giving and receiving compassionate love, and accepting others and ourselves, flawed yet of value at a fundamental level? It is in our relationships with one another, embodied and truthful, that we express ourselves in a way that leads to fullness of life for ourselves and the people we encounter in our day to day lives. We need to put guardrails in place to help us to affirm these important characteristics of our lives in the midst of the increasing proliferation of human-seeming machines and technology.
Email me at lynn@lynnunderwood.com if you want the whole paper on AI and relationships that I submitted to the Office of Science and Technology Policy .
Thursday, October 25, 2012
Metaphor and the Self: A Role for the Arts in Understanding Suffering and Treating the Person in Distress, Lynn Underwood, International Neuroethics Conference, Brain Matters 3: Values at the Crossroads of Neurology, Psychiatry, and Psychology, October 24-25, 2012
Abstract:
Research shows that the metaphors we use affect our behaviors and attitudes in significant ways. (Holding a cup of hot coffee rather than a cold drink can cause us to have a more friendly attitude towards those we meet; Boroditsky’s work showing how reading about a description of crime as a virus rather than a beast can influence our decisions on the best ways to control the same criminal behaviors.) These effects usually happen implicitly – we are not aware of them. The machine metaphor recurs in our descriptions of the brain and the overall functioning of the body and has become an automatic default. It can be useful in simplifying complex systems, and medical training encourages this. Even efforts to promote humanism in medicine slide in this direction, as communication, empathy and ethical decisions are formulated in mechanistic terms.
This kind of thinking can get in the way when we treat conditions for which no “physical” cause can be found. Self-reports made by the person and their experiences of suffering are essential to the identification of the roots of the problems and opportunities for treatment. However if one buys into the machine metaphor too much, the experience of the person is given less weight in the overall assessment, while objective features such as brain scans, blood chemistry and physiologically evident symptoms are given the final say. Measurements based solely on a machine model lead to interventions that presuppose a kind of person that is incongruous with the way we live our lives, and what is most important to us.
Visual art, film and literature can give insight into the nature of the human person that offers alternative metaphors for the human person, and opens opportunities for creative approaches to treatment and evaluation of outcomes. This presentation would elaborate on those, and give specific examples of how they can help yield more effective treatments and decisionmaking.
Return to “Recent and Current Presentations”