The “Why in AI”

Original article was published on Artificial Intelligence on Medium

Image Source: Unsplash.com

The Philosophy of Artificial Intelligence is concerned with whether machines can be made to learn, think, and reason as well human beings. It also asks whether machines should be made to do such things, and how to build these machines to augment human problem-solving. While some basic tasks are more difficult for machines than for humans, e.g. walking and talking, a computers’ ability to utilize oodles of information is unparalleled.

Teaching computers to think and learn requires creativity to direct the processing and translation of data-based information to give people what they
need. The core skill for using machine intelligence to solve real-world problems is posing the right question, statistically and algorithmically, to reveal patterns in the abundance of data and aid in understanding important, person-driven phenomenon.

While this is largely a technical problem for computer scientists and a moral problem for ethicists, it is an equity problem for African peoples. It’s an equity problem because much of data science has been used to make easier the lives of those who already have a comparatively easier and more stable quality of life. It’s an equity problem because the data for African peoples can be dissimilar in volume and misrepresentative of real-world issues. It’s an equity problem for African scientists who may not have access to expert training on the topic. But mostly, it’s an equity problem because precious AI-focused human and machine capital is spent repeating the same trends of inequity that have emerged from historical and imperial asymmetries.

AI4Afrika addresses one issue not addressed by machine scientists, programmers, marketing professionals and society at large. This latter group considers the can, the how, and the should of Artificial Intelligence and machine learning design. AI4Afrika considers the why in AI. The “Why in AI” for African peoples means ensuring a people-centered as opposed
to a technology-centered approach. It refers to choosing technological advances that help individuals become more a part of their community. It refers not just to increasing sales but increasing access to technology itself and access to the resources the technology can be used to free up. Finally, it refers
to knowing when AI as a tool is not appropriate.

As long as African peoples are underrepresented in technology and in the marketplace the data trends that describe all peoples will be skewed. Our job is to work to correct this misrepresentation and to make sure that when AI is used to address issues related to African peoples, that the Why of the AI is at the fore.

by Jacquelyn H. Berry, Ph.D, AI4A Contributor