Illustration by Somnath Bhatt

TEQUIOLOGIES: Technology Creation Rooted in Community, Collaboration, and Mutuality

A guest post by Yadira Sánchez Benítez. Yadira is a PhD student at the University of Southampton, and a Research Software Engineering Fellow at Software Sustainability Institute. Twitter: @yadira_sz

This essay is part of our ongoing “AI Lexicon” project, a call for contributions to generate alternate narratives, positionalities, and understandings to the better known and widely circulated ways of talking about AI.

Companies, governments and institutions in the Global North — and even in Westernized spaces in the Global South — often communicate and think about AI through singular lenses that convey over-the-top optimisms or pessimisms about the future. AI is going to save humanity, as promised by many AI corporations.¹ Alternatively, AI is going to destroy us, with big tech corporations and personalities regularly selling us this fear and the idea that their technological innovation will save us from AI or at least prepare us from AI taking over our lives.

In the process of conveying these lenses, minoritized communities are further otherized from technoscientific processes; seen as either incapable of having any answers to everyday living problems that may require technological solutions, or seen as passive consumers of technology and capital exploitation.

In this essay, I present an alternative discourse of AI creation: one driven by communities, collaboration, and mutual support. Instead of AI as technologies, I offer the term “Tequiologies,” coined by Mixe linguist and activist Yásnaya Elena Aguilar² that roughly translates to: “technologies as collaborative work for mutual support.” The term is commonly used and lived in many towns across Mexico and is rooted in the word “tequio” (from nahuatl tequitl — work, contribution, duty — which has been co-opted by many communities as work towards the community), with similar terms such as “faena,” (labor, task, work)³ “or “minga” (from quechua mink’a — collective work done for the benefit of the community)⁴ being used further south in Latin America.

The key objective of “tequiologies” is to find a middle ground that separates the development of technologies, economic development, and consumerism. The ideas I present below are grounded in my personal history, and offer new forms of designing and creating AI technologies. Only by reconceptualizing technology itself, are we able to design and create AI systems that are driven by collaboration and mutual support. The futuristic possibilities of AI should be in service of the common good, not the market.

I was born and raised in a small rural village in Mexico where livelihood is mainly based on agriculture and farming. In school, we were taught to understand technologies such as huge infrastructures and systems with large reach and impact — like the Internet, PCs (at the time), airplanes, and mobile phones, to mention some. Yet, the technologies in my community — all with their own important use, purpose, and impact — had nothing to do with the technologies taught at school.

For example, in the crops of my village and surrounding villages, irrigation systems are used as technological tools built in collaborative effort to support each other’s crops. Farmers come together and exchange information about what tools and processes are best to build an irrigation system or the infrastructure for the most benefit.

As opposed to the way that technology is developed by profit-seeking firms, this form of technology creation is rooted in the sense of community, collaboration, and mutuality. It is not a top-down approach where the farmers wait for some external organization, institution, or “expert” to provide them with the tools and solutions to their problem. At the end of the day, they know their most pressing problems and what is needed to fix them; these decisions are made through collective intelligence.

Collective decision-making around which technologies to use, and community issues in general, are generally done through “community assembly.” In community assembly traditions, it is customary to listen to both those who are in favor of a proposal and those who are totally against, giving both positions the same attention and the same exposure time; sort of a collective intelligence process, because everyone in the assembly has knowledge that makes up the final decision. Although this is easier said than done; arguments, discontent, and rawness are a part of the process of bringing a community together to make a collective decision.

For AI infrastructures and machine learning operations⁵ to fully embrace this idea of tequiologies, collective and collaborative efforts would include not only participants involved in the design process but the entire community’s knowledge and context by following a kind of community assembly becoming more a type of collective intelligence aided by tools.

Machine learning and artificial intelligence infrastructures require significant computational resources, making it difficult for individuals and communities within or outside universities or tech corporations to develop AI. That is partly the reason why corporations and universities are the only ones leading the so called “AI race.” AI technologies are designed to increase the piercing mechanisms of capitalism and other forms of domination in each of the spheres of our lives — all efforts antithetical to the ethos of tequiologies.

For example, there are now several “artificial intelligence applied in agriculture” programmes in Mexico led by corporations and supported by the government.⁶ These programs offer hope that artificial intelligence will benefit farming where land management is becoming more difficult due to high variability of rainfall and water scarcity due to climate change. Yet such AI projects leave aside the fact that the dependence of farmers on traditional practices like rainfed practices have now become more difficult due to climate change which, as we know, has been exacerbated by industrial and exploitative practices outside the farmers’ power.

A process of AI as tequiologies would mean that its infrastructure, design, and deployment would need to fully respect the contextual needs and desires of the communities, following their communal consensus processes. This is especially true for communities without access to computational resources, or the capital required to fully implement technological projects due to power imbalances instilled through systematic decision making. It would be necessary to have a thorough process in which the communities’ own proposals (other ways of addressing technological innovation) are centered. Once the information has been heard and discussed in both directions, decision-making can become an honest exercise that starts from sincere intention to be fair, open, and honest.

A tequiology understands that communities do not need a greater presence of the state or corporations. Instead, it means strengthening autonomy and self-determination. In many parts of the Global South, there are projects and initiatives that propose other ways of facing the problems that originate from the historical oppression by the state and corporations. Before proceeding, we should listen.

Collective work for mutual support as tequio or “minga”⁷ has been at the forefront of real-world initiatives tackling racist and oppressive AI systems as well as in the development of technological and AI tools. The Mijente #notechforice campaign and Stop LAPD Spying are good examples of community movements in the US which are keeping oppressive algorithms accountable. Other examples include the linguistic rights digital movements lead by Indigenuos people across Mexico and Latin America⁸ as well as the development of community networks.⁹

The creation of AI tools as tequiologies defends and supports bodily and technological sovereignty of communities through collaborative work for mutual support. This translates into allowing communities to make their own decisions about their data, their spaces,¹⁰ and what AI frameworks may truly benefit them long term or if at all. For example, AmericasNLP encourages the development of machine translation (MT) systems for indigenous languages of the Americas and centers researchers and professionals from underrepresented communities and native speakers of endangered languages. Now, what would AI for/by/with farmers in rural communities look like? First, we would have to ask the question, is any AI tool actually needed to aid the hardships currently faced by farmers and do they fully understand the implications of such technologies in their spaces or how they can actively be a part of the process, or is it more a question of power imbalances where the farmers are seen as another point of exploitation in the race to ‘technically fix’ the effects of climate change?

It is critical to consider that creating and managing AI infrastructure is not only a technical matter, but that it involves different forms of use, appropriation, and transformation of machine learning tools in a particular territory. Seeing AI tools and any other technological tool as tequiologies means they should not be replicated nor imposed on society for the benefit of the few, but adapted and contextualised to community needs, and many times even discarded.


[1] Technology in 2050: will it save humanity — or destroy us?

[2] Yásnaya Elena Aguilar is a linguist, writer and activist of language rights. She is Mixe from Ayutla Mixe, Oaxaca, Mexico. Her mother language (Mixe) is one of the most spoken in the state of Oaxaca and has a great internal variety. Hers being ayuujk. In Spanish she is Mixe but in Mixe she is ayuujk jä’äy and speaks ayuujk.

[3] Faenas are seen as a labor tribute to the community or a cash-free form of local taxation.

[4] Minga — an indigenous word that can imply a meeting of diverse actors, knowledge and tools in search of a common goal. Mainly practiced in Colombia, Peru, Ecuador, Bolivia, Chile and Paraguay.

[5] AI infrastructure and machine learning operations, or MLOps. Both terms denote the technology stack necessary to get machine learning algorithms into production in a stable, scalable and reliable way.

[6] Crédito Real; Nava; Towards an AI Strategy in Mexico: White Paper

[7] Minga — an indigenous word that can imply a meeting of diverse actors, knowledge and tools in search of a common goal. Mainly practiced in Colombia, Peru, Ecuador, Bolivia, Chile and Paraguay.

[8] Rising Voices through Act Lenguas — an outreach initiative of Global Voices, aims to help bring new voices from new communities and speaking endangered or indigenous languages to the global conversation.

[9] Association for Progressive Communications (APC), Rhizomatica, REDES A.C., Centro de Investigación en Tecnologías y Saberes Comunitarios (CITSAC), Red de Comunicadores Boca de Polen A.C and Techio Comunitario

[10] APC, Technological autonomy as a constellation of experiences: A guide to collective creation and development of training programmes for technical community promoters


Algorithmic Justice League — Unmasking AI harms and biases.

Aguilar Gil, Y.E (2020) A modest proposal to save the world through “tequiology”.

APC (2021) Technological autonomy as a constellation of experiences: A guide to collective creation and development of training programmes for technical community promoters | Association for Progressive Communications.

BBC (2021) Qué es la minga indígena y qué papel juega en las protestas de Colombia.

Crédito Real (2021) Inteligencia artificial aplicada en la agricultura.

Draffan, E.A. (2021) Global versus Universal? | International Symbol Dictionary for AAC.

Global Voices. Activismo Digital de Lenguas Indígenas. Blog Post.

Hern, A. (2021) Technology in 2050: will it save humanity — or destroy us?. the Guardian.

Nava, D. (2019) Así es como funcionará la agricultura gracias a la inteligencia artificial. El Financiero.

Rising Voices · Helping the global population join the global conversation.

Towards an AI Strategy in Mexico: Harnessing the AI Revolution (2018) White Paper Report.

Tyagi, H. (2021) What is MLOps? Machine Learning Operations Explained.