Highlighting the social structures and power dynamics behind technological and AI building processes as a necessary step towards localised sociotechnical solutions

Invited panelist at AAAI 2021 IBM Workshop – Reframing Diversity in AI: Representation, Inclusion and Power

It is very important to note that expressions such as “the Global South” are also problematic if there is no consideration that permanent settler-colonial-states form the Global South too.

Thank you for creating this space and inviting me to be a part of this workshop. I must say that I guess the probabilities of me being here had been extremely low along with most of us or all of us if the organizers were different; considering either our geographical, racial, gender, sexual orientation, socioeconomic, religious, cultural, educational background, some of them or all of them and how these identities intersect in ways that impact how we are viewed, understood, treated and therefore very often left out of these knowledge sharing events.

But “intersectionality”, a term coined by American legal scholar and activist Kimberlé Crenshaw makes room, as she mentions, “for more advocacy and remedial practices” to create more egalitarian systems. Aiming to deeply and objectively understand the complexity of our society, an intersectional perspective is a realistic and tangible way to describe an issue but also a tool that can help us understand, describe and remediate socio-technical problems.

How can we all be a part of this process?

We are all here because we are either involved, care and/or are interested in technology, AI and their potential to more egalitarian societies. It is important though to acknowledge the intersections in how society functions.

While many AI solutions have originated and stayed mainly in the “western” world, they are increasingly becoming the answer to many challenges in the Global South. It has shown great potential for helping to achieve sustainable development objectives at the global level and national levels (Mbayo, 2020).

However, while the benefits may be many, there are a lot of risks that come with the application of AI in a social/developmental context. Particularly questions around power and how the imbalance of power impacts the use and absorption of AI.

The universal mindset, as we know, of the tech industry is leaving behind the situated contexts in which many of the issues AI is trying to solve are. Power imbalance, especially between North and South and within these is key to understanding how this unbalanced relationships can affect the way knowledge is cultivated, shared and information is accessed.

We know that most of the big tech companies are in the global north and that most of the global south looks up to the west as an example for developing technologies. However, it is important to acknowledge that technologies and AI models developed for use in the global south context and the continuation to have data accessed and stored within the west can only result in reinforcing and heightening this imbalance of power (Mbayo, 2020).

I would like to quickly position myself here, because having been born and raised in a small rural village in Mexico, being a software developer with a background in social sciences...I have seen time and time again like most of us here how the contexts of many communities like ours are completely ignored from the conversation in policy making and technology advancement ...and how racialized, oppressed communities are always seen as passive receivers of these decisions, both socially and technically.

This points to how these power dynamics also occur within the global south in which the rush to apply western techniques also denies the pluriversality of the Souths (Santos, 2016), further harming already left out communities. And ultimately, the development of AI technologies for but not by historically marginalized communities can only further amplify existing oppression and biases(Mbayo, 2020).

Can we as researchers, engineers, artists, activists involved in technology building be a part of creating a more just society?

I would like to think that being aware of how we may perpetuate oppression and bias by the position we have in these building processes is an important step.

Perhaps by asking ourselves, what social orders are being reproduced in this process I am taking part in?

I keep thinking about how inclusion alone, bringing as many Black, Latinx and Indigenous persons is definitely a step forward but also how, for as long as the same systemic structures remain...the creation of technologies with other realities in mind will keep being hindered.

I want to share here the thoughts of Yásnaya Aguilar, an Indigenous Mixe linguist and activist from Oaxaca, Mexico; on the dangers of romanticizing power [Nëmatyä’äky in Ayuuk]— focusing on power relations (Aguilar, 2020) when creating technologies might deviate us all from considering other forms of creation.

Based in Indigenous knowledge throughout Latin America, she coins the term “tequiology” which means technology as collaborative work of mutual support (Aguilar, 2020).

I would like to believe that this could offer us all here some hope in how we create technologies ...by collaborating with each other. I can think of the need to open more collaborative work between engineers, researchers from other disciplines, artists, activists and communities themselves to shape the path to a truly sustainable and just AI. I am hopeful that this can offer us some tangible results towards social justice.

Thank you!


Aguilar Gil, Yásnaya Elena (2021) Nëmatyä’äky. Romantizar el poder

Aguilar Gil, Yásnaya Elena (2020) A modest proposal to save the world

Crenshaw, Kimberlé (2020) Kimberlé Crenshaw’s Intersectional Feminism

Mbayo, Horlane (2020) Data and Power: AI and Development in the Global South

Santos, Boaventura de Sousa (2016) ‘Epistemologies of the South and the future’, From the European South, 17-29, 18.

— Yadira