Box Breaking : Data, Dismantling and Developing

​The workshop gathers startups and stakeholders interested in artificial intelligence (Al) development in Zambia to explore the importance of representative datasets. Through discussion and collaboration, this event seeks to unpack the definitions, obscurity and limitations of data on which Al relies and promote technology's ethical, responsible, and inclusive use in addressing local challenges for sustainable development.

​'Boxbreaking' spins off from 'blackboxing', an abstract term referring to the obscure internal workings of a system that make it difficult for users to understand how it works. Instead of treating the system as a blackbox where its mechanisms are predominantly invisible, this workshop aims to dissect datasets, Al's inherent bias, and considerations to humanise technology for better ethical practices.

​This workshop forms a part of Artificial Intelligence in Maternal Health in Zambia (AIMZ), a Mozilla Foundation Africa Mradi funded project uncovering the extent of bias in Al designed for maternal health in Zambia. The visual output of this workshop will feature its key takeaways on dissecting bias, the importance of representative datasets, promoting ethical Al and humanising technology.