THE FUTURE OF AI

Criticism is not as common in science and engineering as it is in art and literature. Scientists review each other's writings, write commentaries on them, or try to replicate a reported experiment, but they rarely criticize each other's work in the sense intended here – that is, in the sense of a reflexive evaluation of background assumptions, biases, and principles, and an examination of the way they guide the creation of theories, models, and technical systems. Traditionally, scientists leave such criticism to “outsiders” – philosophers, sociologists, or even to religious or political ideologues. This is not to say that science totally lacks critical practices. Scientists also have dedicated venues – societies, workshops, conferences, committees, panels, journals, as well as very rigorously established procedures of peer review – whose function is to keep things in constant check, presumably mitigating the need for self-criticism. Nevertheless, it is not common to find reflexive accounts of science from within the scientific community. Science changes, but the driving engine of change is seldom an explicitly critical encounter on the part of scientists themselves.

Contrary to this received tradition in science, the present work conducts a critical assessment of a scientific and technical field, Artificial Intelligence (AI), from a constructive standpoint that aims to contribute, rather than refute.