Originally published in Portuguese on March 24, 2023.
In the early 1960s, the Brazilian philosopher Álvaro Vieira Pinto developed a definition of technology that highlighted its complex and multifaceted nature, emphasizing the need for critical reflection and ethical considerations in its use and development. For him, technology is a form of knowledge that transforms nature, being much more than a set of tools or machines. Through relationships between humans and their environment, it creates a specific way of understanding and acting in the world.
In his conception, technology is part of the transformations produced by each society within its characteristics, to address its contradictions. In this process, technology is fundamentally linked to social, economic, and political systems, and, as a fundamental part of human culture and history, it is never neutral. It reflects the values, interests, and power relations of those who create and use it, thus having issues of national sovereignty and development possibilities intrinsically linked to the transformations it promotes.
In the context of using design in the production of ICTs (Information and Communication Technologies), and considering the periphery of capitalism, it is essential to reflect on these power relations and how they occur, especially in Brazil and other countries with a significant history of colonialism and imperialism. These relations lead to the exploitation and oppression of subalternized people and communities, resulting in significant economic inequalities and social injustices.
However, following Álvaro Vieira Pinto's own principles, it is a mistake to "substantialize" technology, which can lead to empty moral judgments, such as viewing technology as "salvationism" or as a "threat." It is important to understand technologies within this spectrum of human relations and to seek alternatives to decolonize their production and use.
In this sense, we can think of an approach to decolonize artificial intelligence. How can we initiate a discussion for the development and deployment of AI technologies in a more socially just manner? This text proposes 5 minimum requirements to achieve this goal:
Diversify AI Development Teams: Ensure that AI development teams are diverse and inclusive, representing a variety of perspectives and experiences.
Decolonize Data: Critically examine how data is constructed and how it is used to ensure that it is culturally appropriate and respectful.
Center Marginalized Communities: Shift the perspectives and experiences of marginalized communities to the center of AI development and deployment. This may involve collaborating with communities to co-create projects where AI meets their needs and respects their values.
Promote Transparency and Accountability: To ensure that AI systems are designed and deployed ethically and responsibly, it is necessary to understand the potential impacts of AI on marginalized communities and take steps to mitigate negative effects, but perhaps even more importantly, to amplify the creations and uses by these communities.
Rethink Ownership and Control of AI: Promote collective ownership and collaboration. This may involve creating open-source AI systems designed to be owned and controlled by communities, or fostering projects where the use of AI primarily generates public benefit.
These points imply rethinking the ethical issues related to AI, understanding ethics as the way of being in the world relative to the people developing these technologies, thus critically evaluating their implications and ensuring that they are aligned with the values of the communities that will be impacted by them. I believe that the only way to achieve this goal is by giving voice to the production and use by these communities. Anything less than that is just another form of colonization.Is a socially just Artificial Intelligence possible?
Comentários