Understanding Critical AI Research in 'Majority World'

Learnings at #SIMC Part III**

Recently, I attended an interesting talk on Digital Humanities and Critical AI Research. The session started with an interesting argument: Why should developing nations like India, Bangladesh, Sri Lanka, etc. be called “Third World”? Furthermore, why should they even be called “Global South”, a term first used by American writer-academic-political activist Carl Oglesby in 1969? The term has connotations of uneven power dynamics in the world, and may not be the corrected reference to the developing nations. So, what should it be called?

The answer to this – and many other critical questions related to AI – were found in this extremely informative session titled “Humanistic Inquiry and/in Critical AI”, by Dr Dibyadyuti Roy, Assistant Professor (Cultural Studies, Media Studies and Digital Humanities), University of Leeds, UK, held at Symbiosis Institute Of Media And Communication, Pune.

Dr Roy, while referring to Bangladeshi photographer, writer, and social activist Mr Shahidul Alam, said the correct representative term for such countries is “Majority World.” Mr Alam had coined the term “Majority World” in the 1990s, to highlight that the majority of the world's population lives in the regions traditionally known as the developing world or Global South.

The correct representative term for such countries is “Majority World”: Bangladeshi photographer, writer, and social activist Mr Shahidul Alam.

The point is well taken.

Thereafter, there were other important deliberations on how a “text” at the fulcrum of three major disciplines – Cultural Studies, Media Studies and Digital Humanities. Then, what is text? Almost everything - from old relics to all social media posts. He quoted a scholar (whose name I can’t figure out in my notes): “There have long been two texts: one, an object; and another, a paradigm.” Mentioning David Beer’s 2017 book “The Tensions of Algorithmic Thinking”, he connected it to “paradigms of Critical AI.

“There have long been two texts: one, an object; and another, a paradigm.”

Coming to AI, Dr Roy said that AI or Artificial Intelligence is not a new phenomenon; rather it has already gone through two previous “hype cycles” in the 1970s and 1990s (before the current craze). Similarly, he argued that LLMs and NLPs have been there, and quoted an important paper titled “Attention Is All You Need” (Vaswani, A., Shazeer, N., et al, 2017) to support this.   

During the session, he also raised some pertinent points such as:

• closed or analytical readings of texts;
• reading of literature also being a part of research methodology;
• cultural logic (Enfield, 2007);
• deductive logic v/s inductive logic;
• from surveillance capitalism to cognitive capitalism;
• network’s environment culture;
• big data v/s thick data,

and many more.

Plus, he also talked about over a dozen important texts by scholars on the topic that one should read to understand Critical AI Research. They are all very helpful.

Since I have done a minuscule research on AI for my joint work on AI and Journalism, this was a highly informative session for me and added to my continuous learning on the topic.

Thank you Symbiosis Centre for Research in Media and Creative Industries for organising such a wonderful session. Another feather on your cap. Thanks are also due to SIMC Director Prof. Ruchi Kher Jaggi , Prof. Swapna Gopinath and SCRMCI's very own Dhara Shah

Lastly, thank you for enriching us with your insights Dr Dibyadyuti Roy .

PS: If there are any errors or misconceptions in my notings, it should be considered my shortcomings and not that of the session.