2019: developing ai governance
AI is already an everyday part of our lives. From voice recognition on smart assistants to advertising, very little of what we do is not being funneled into some enormous data processing machine. Although you can choose to disable Facebook and avoid using Google (disclaimer: I work here currently), these individual actions don’t really allow you to escape from reality: the domain of AI is not only in the digital sphere, but also in the physical. The advent of autonomous vehicles, government-sanctioned mass surveillance and facial recognition and AI being integrated into criminal sentencing all are signs of the drastic societal shifts which are occurring.
Data sovereignty and AI governance are critical policies that we need to implement. AI is not some magical science – at its core is data that has been defined, curated and determined by humans. Our action (or inaction) can either further perpetuate issues of injustice or stop them. Furthermore, it also is not enough to think of making sure that AI treats everyone the same, from the AI Now 2018 Report:
Definitions of fairness face a hard limit if they remain purely contained within the technical domain: in short, “parity is not justice.”
This is complex work, and systemic change is needed. There are no easy answers, but this is the time to push for change.