This is the mobile-friendly web version of the original article.

WHAT DO WE TALK ABOUT WHEN WE TALK ABOUT AI?

March 2020.

ALGORITHMIC DECISION-MAKING IN LATIN AMERICA

By María Paz Canales

Video: Defender los derechos digitales

This work is available under a Creative Commons Attribution 4.0 International Licence (CC BY 4.0): https://creativecommons.org/licenses/by/4.0/deed.es

1. ALGORITHMIC DECISION-MAKING IN LATIN AMERICA

This report was made by Derechos Digitales.

Edited by Vladimir Garay.
Design and layout by Constanza Figueroa.
Style correction by Rocío Consales.
Translated by Alex Argüelles.

Access the full-length release of Latin America in a Glimpse 2020 in Spanish at: https://segu.dev/3

While the gap between rich and poor increases in the world, Latin America remains the region where wealth is distributed more unequally. According to the Economic Commission for Latin America and the Caribbean (ECLAC), there are marked imbalances between different socioeconomic levels in aspects such as life expectancy, infant mortality, illiteracy rate and access to water inside homes. The very high inequality levels that plague the continent directly influence the development possibilities of its inhabitants and the exercise of their fundamental rights.

The impetus that public actors in the region have given to the implementation of automated decision-making technologies and AI development is based on the belief that these would allow solving all of those issues more quickly or in a better way. In Argentina, Chile, Colombia, Mexico and Uruguay, national artificial intelligence strategies are being developed or are beginning to be implemented, creating road maps that seek to take charge of the possibilities and risks that this type of technology implies for the local context.

However, the first thing that is possible to verify is that there is no complete agreement regarding the definition of artificial intelligence. Some propose that it refers to the ability of a computer to perform tasks commonly associated with intelligent beings; or to the simulation of human intelligence processes by machines, especially computer systems, including learning, reasoning, and self-correction processes; or a technology that makes predictions based on the automatic detection of data patterns.¹

But regardless of the definition we choose, the truth is: there is still a long way to go in order to be able to effectively refer to forms of artificial intelligence when we talk about the technologies that are being made and implemented in our region. Most of the technological developments that are announced today are more modest, covering in most cases processes of automation in decision-making or assistance to humans² through algorithms that simplify the processing of large volumes of information. Even so, the social impli-

¹ https://medium.com/@chethankumargn/artificial-intelligence-definition-types-examples-technologies962ea75c7b9b

² Sometimes referred to as augmented human intelligence.

cations of this automation are relevant enough to pay careful attention. If our governments prefer to do it under the label of artificial intelligence, either as an aspirational slogan or by technical simplification, this is not an obstacle to examine from the perspective of the public interest the impact of these technologies, which we will generically call automated decision making systems or AI.