What can AI teach us about ourselves and our communities? This essay by Tom Demeyer explores the cultural potential of AI.
AI in Culture & Society. Of course, where else? In this piece we’ll be looking at AI technologies mostly from the perspective of the humanities. More precisely, we’ll be looking at what the use of AI can teach us about ourselves and our communities, and how society and public policy might benefit from the advent of more and better instruments to set priorities, understand preconceptions, suggest courses of action and evaluate results.
The highly visible, and much discussed applications of AI in advertisement and surveillance have a huge impact, of course, on the structure and functioning of our societies, and not necessarily for the better. In fact they help spark the sentiment of a ‘race’ we have to participate in, or have already lost, without much (or any) consideration for the goal or direction of that race, other than achieving technical supremacy (and world domination). With the advice given to policy makers: “join the race or get left behind,” we’re missing the point altogether. A race towards more economic growth and unsustainable consumption or more effective state control is not a race we as citizens need to participate in. Instead we’ll be looking at these technologies with a designing purpose; where and how we can exploit the sometimes almost unreasonable effectiveness of some of these techniques to work on better versions of ourselves, our institutions and our societies. Value-driven, open and inclusive design will help to define ‘better’ to have an operational meaning in the relevant application context and for the relevant stakeholders.
There’s no clear picture or map of where current AI technology is deployed, in which domains and with what results or effectiveness. We cannot draw a clear line between traditional business intelligence, statistical analysis, data-driven decision making and what we currently mean by AI. It is clear, however, that huge strides have been made in various vertical domains in the last decade. In military applications, we presume, in geological surveys, financial services, in medical diagnosis, and, not to forget, in assistive applications (from translation and Alexa to sports- and weather ‘journalism’). The absolute supremacy of deep learning technology in gaming contexts would also suggest a huge potential in applications for international diplomacy and conflict management, although we don’t see any evidence as of yet…
One potential that we do not yet take advantage of is the power of AI to study our collective selves. Not in the sense that AI is comparable to human intelligence, it is not, but in the sense that we can ‘freeze’ a small section of our culture and lift it out into a model, enabling us to prod, probe and study that little bit of our existence in ways that we have never before been able to, potentially gaining insights that would otherwise be much harder to achieve. This we will explore in the rest of this paper.
Author: Tom Demeyer