SLICE: Creating futures of our own imagination

 Imagining a future when South Africans are part of creating global technologies that take on board local contexts. 

In 2017 I took a course called Utopian Studies offered by the department of political studies at Wits University. Utopian Studies allows us to construct a coherent imagined future, and to consider all philosophical, ethical and theoretical possibilities, to determine an ideal towards which we can strive because when we do not have a collectively imagined ideal, it becomes harder to know what we are working towards.

This made me think about what an ideal South African state should be. Should it be one where everyone is happy, or one where everyone has money?  

At the time that I did this course, the university was coming into a self-awareness of the way that institutions have a culture that is historically white, and was seeking ways to transform itself into a space that was accessible to all the people in it.  

So, in this context, the coordinator of my Utopian Studies course, Julian Brown, began to deconstruct the ways in which media genres that offered projections of humanity in the future (mostly sci-fi films and books) were often predicting “a vision of a [white] future where assimilation, not diversity, is the goal”.

It speaks to the extent to which a diversity of voices and ideas exist within the spaces where the media content is produced. 

This provides a lens to understand the need for a diversity of voices where artificial intelligence (AI, the programming of machines to mimic human intelligence) development is concerned, to place a diversity of developers in the spaces where AI is trained. Because we run the risk of recreating much of the socio-political dynamics we have today, in our more technologically advanced future. Unlike with search engines and social media platforms, AI requires us to develop the technologies that make a South African AI possible.  

In November 2022, OpenAI, a US technology research lab, launched ChatGPT, an AI computer programme that can interact in a chat-based conversation with humans. The programme is trained on data from across the internet and is able to mimic human cognitive processes in its conversational responses to a prompt.  

This means that unlike regular search engines such as Google, ChatGPT uses deep learning techniques to build context and give more in-depth answers in a way that a human would. This is an incredible developmental milestone for AI technology considering that until now, most AI programmes could do little more than just following an instruction.  

Now, because AI technology is dependent on being pre-trained by human beings, it makes sense that it possesses, to a certain degree, subjective, biased and sometimes even prejudiced data.  

For this reason, the arrival of AI technology as advanced as ChatGPT creates a serious impetus for South Africa to invest more intentionally in the development of our own AI technology. Not necessarily to compete with Open AI, but because we know that the knowledge and information generated by foreign AI may not be sensitive to our cultural contexts and may continue to perpetuate a false sense of cultural and moral universality that makes us the ‘other’.  

The AI Institute of South Africa (AIISA) launched an AI hub at the Tshwane University of Technology in Pretoria on March 24, 2023, in collaboration with the University of Johannesburg. Reporting on their website, the two institutions promised that through their hubs, they would “generate knowledge and applications that will position South Africa as a competitive player in the global AI space”. 

The hubs provide us with an opportunity to create futures of our own imagination. This has the potential to create global technologies that take into consideration local and contextual issues.  

FEATURED IMAGE: Morongoa Masebe, Wits Vuvuzela student journalist. Photo: File


Wits Vuvuzela, April 2021.

Wits Vuvuzela, May 2021.

Wits Vuvuzela, Oct 2019.