The AI Mirror

Shannon Vallor is Professor of the Ethics of Data and Artificial Intelligence at the University of Edinburgh, where she directs the Centre for Technomoral (a new word for me) Futures in the Edinburgh Futures Institute. She is also former AI Ethicist at Google, and the author of an earlier book entitled Technology and the Virtues: A Philosophical Guide to a Future Worth Wanting. In this timely new book, she asks how we can reclaim our humanity in an age of machine thinking. Her important and central insight is that today's powerful AI technologies only reproduce the past. In other words, our new digital mirrors dominated by a wealthy Silicon Valley mindset can only point us backwards. Instead, we need to recover our sense of the possible within a virtue ethics framework. In this respect, I strongly agree with the author that moral and intellectual virtues like wisdom, courage, justice and honesty hold the key to meeting the tremendous challenges we now face as a species. (p. 67) She quotes the Spanish existentialist philosopher Jose Ortega y Gasset and his notion of ‘autofabrication’, our lifelong task of making ourselves. We cannot afford to lose ourselves in the AI mirror, which reflects our own (largely left-hemisphere) intelligence back to us. The myth of Narcissus serves as a cautionary tale.

We are under an obligation to reclaim our self-knowledge and self-determination ‘from the predictive power of an opaque algorithm’ and to rescue ourselves from the dominant commercial form that aspires only to cognitive perfection and is generically incapable of wisdom. The author advocates that we use AI technologies ‘to strengthen our humane virtues, sustaining and extending our collective capabilities to live wisely and well… nourishing the relational bonds that sustain us in solidarity with others.’ (p. 8) We need to ask ourselves what we might now become, caught as we currently are ‘in the grip of a gradual and accelerating mechanisation of the human personality: the systematic replacement of reflective discernment with mindless prediction; the efficient sacrifices shared flourishing to expected utility; the exchange of humane creativity and open-ended progress for local optimisation of content delivery. In short, the surrender of humane wisdom to machine thinking.’ (p. 9, my emphasis)

A series of lucid chapters elaborates this thesis, reminding us of the perils of ‘handing over our power and responsibility to secure the flourishing of future generations to decision-optimising algorithms that are mathematically guaranteed to reproduce the unsustainable patterns of the past. This is a calamity – a betrayal of life and its possibilities.’ (p. 36) There is an influential school of thought advocating handing over the future to AI on the grounds that humans have created so much havoc and destruction. However, this constitutes a fundamental abdication of our moral responsibility. Making a deeper commitment to humane flourishing in the future is an act of ‘moral creation, expression, self-reinvention, and love, which no mirror can perform. This remains hard for us, yet wholly within our power. But it requires our will to unmake the pattern that has come before.’ (p. 160) The agenda is fundamentally about reclaiming AI and technological culture for a sustainable moral vision in an act of collective autofabrication inspired by creative practical wisdom. (p. 193) Vallor shares the vision of Lewis Mumford that science and technology must one day return to their roots in the service of life, a service that has in fact continued in the humane arts that tend to be side-lined by technological consciousness in the name of ‘relevance.’ The significant book carries a vital cultural message relating to the creative potential of the human being.