First Ever LatinX in AI Research Workshop at NeurIPS Recap

Sebastian Anaya
Published in
9 min readJan 7, 2019


NeurIPS 2018 proved to be a historical conference as it has been considered to be the most inclusive and accessible so far, yet there is still a lot of work to be done. Of note are ticket demands and visa issues which continue to provide barriers to attendance for less privileged individuals and marginalized populations.

One of the major accomplishments this year was the conference name change. After much protest, the conference changed its name from “NIPS” to “NeurIPS” after it was recognized that some attendees were purposefully misusing the acronym for sexual puns. One such example, Elon Musk reportedly made a joke relating to breasts. Anima Anandkumar pioneered the name change in the public domain via her Twitter account. NeurIPS was brought forth by the community organically, not decided upon by the conference organizers.

The biggest change thus far came from the representation of organizations such as Women in ML, Inclusion in ML, Black in AI, Queer in AI, Jews in ML and LatinX in AI were present to promote fairness and inclusion.

A couple of other firsts this year was bringing on two Diversity and Inclusion chairs as well as parental care being offered to help ease the burden of women and researchers with families in general.

History was made on December 8, 2018 when the first ever LatinX in AI research workshop @ NeurIPS was hosted. The workshop brought together faculty, graduate students, research scientists, and engineers for an opportunity to connect and exchange ideas. While all presenters identified primarily as latinx, all were invited to attend.

This marks the first of many as we have developed a strong relationship with the NeurIPS board and affinity groups for joint collaborations!

Opening Ceremony

Laura Montoya helped kick things off with the opening ceremony by sharing why the first ever LatinX in AI Research workshop @ NeurIPS was hosted as well as the mission of LatinX in AI coalition.

Main Purpose of this workshop

  • Use of AI across industries, specializations, and countries
  • Opportunity to connect and exchange ideas
  • Highlight Latinx faculty, graduate students, research scientists and engineers

This workshop is so important because we want to ensure we had Latinx representation at this conference, giving them an opportunity to share their work in the global market by lowering the barriers to entry.

Creating Harmony and Opportunity for Latinx in AI

Opening Keynote

Can AI be unfair?

One day Omar woke up and decided to answer the question, What is the average Latino face on the internet.

We are not correctly represented on the internet, or in Hollywood based on the faces that have been compiled over time.

Omar’s background is an outlier and one could say he had to work twice as hard. He was born in Peru and needed to navigate a much more difficult journey in order to be in the field of AI. He had to learn English to complement his graduate education. Immigrating to the US was extremely difficult given the barriers to being granted even a tourist visa. Lastly, Omar had to obtain his Ph.D. via scholarship due to financial circumstances. Omar had to go through many hurdles in order to become a Senior Researcher in AI at a major financial institution like CapitalOne.

Problems with AI

1. AI makes decisions on our behalf often based on maximizing an objective function. And sometimes we do not understand its consequences.

An example is with Amazon Prime, in Atlanta, the service was mostly available in the North rather than South. This could be a result of race as most black neighborhoods in Atlanta are in the South whereas predominantly white neighborhoods are in the North. This same scenario happened in Boston, Chicago, and Dallas.

2. AI enables automation and also the displacement of manual work.

Changying Precision Technology Co has been able to create automated production lines for mobile phones in China

After implementing these changes, Changying saw the following:

  • Decrease in 90% human workers
  • 250% increase in productivity
  • 20% decrease in product defects

“The automation has already decimated jobs in traditional manufacturing, and the rise of artificial intelligence is likely to extend this job destruction into the middle classes, with only the most caring, creative or supervisory roles remaining” — Stephen Hawking

3. AI reinforces stereotypes because it reflects the unfairness and hierarchies that are stored in our datasets

During the keynote, Omar highlighted screenshots from Google Image results of doctors and cab drivers to support his point above.

Why does AI work like this?

  • Most of our mathematical techniques are designed to “predict by generalization”
  • These favor data that frequently occur and penalizes missing values and outliers
  • The outcome often misclassifies patterns from minority groups (LatinX, Afro-Americans, Women, Native Americans) with noise
  • Under-sampled minority groups

Potential Solutions

  1. Use AI — our problem is the answer

Autoencoder for faces

2. Create AI content in Spanish so the language does not become a barrier for Latinx talent

Created the AI English to Spanish dictionary

3. Create opportunities for raising Latinx talent

Can AI be unfair?

  • Aside from being an optimization problem, this is also a problem of education, ethics, and values
  • We need more Latinas and Latinos in AI to represent our culture correctly
  • It is the mission of our generation to create fair AI (F-AI)

Build ML one can truly rely on

Not only safe/secure/fair but also “better”

Oral Presentations

We brought 60 researchers who identify as Latinx from South, Central America, the Carribean, US, Spain, and Europe to present their work. The researchers came from various distinguished universities and companies such as University College London, University of San Francisco, National Tsing Hua University (Taiwan), MIT Media Lab, Google AI, IBM Research,

Topics presented included: Predicting Criminal Behavior, Singing Voice Detection, Sign Language Recognition, Dopamine Research Framework as well as many others.

If you would like to watch any of these talks, the recording of the live stream will be made available soon!

Poster Sessions

For the workshop, over 30 different researchers were able to showcase their work and connect with others in various fields of Artificial Intelligence. Below are the names of the researchers who presented a poster for the workshop:

  • Hugo Jair Escalante Balderas, Associate professor, INAOE
  • Fernando Alonso Delgado, PhD Student, Cornell University
  • Alejandra Márquez Herrera, MSc Student, San Pablo University(Arequipa-Perú)
  • Leonardo Cayo Leon Vera, Student, Universidad Nacional de Ingenieria
  • Paula Kintschev Santana de Moraes, MSc, University of São Paulo
  • Leissi Margarita Castaneda Leon, PhD student, University of Sao Paulo
  • Albert Manuel Orozco Camacho, BSc Student, UNAM
  • Diana Mabel Diaz Herrera, PhD Candidate, Wayne State University
  • Jorge Ortiz, Ph.D., Assistant Professor, Rutgers University
  • Pablo Fonseca, Visiting Researcher, MILA
  • Francisco Cruz, Universidad Central de Chile
  • Nils Murrugarra-Llerena, PhD Student, University of Pittsburgh
  • Mariano Phielipp. PhD., Senior Deep Learning Data Scientist at Intel AI Labs
  • Javier Andrés Orduz Ducuara, Professor at UNAM
  • Wilbert Santos Pumacay Huallpa, Msc. Student
  • Diana Silvia Patiño Polar, Student, Universidad Católica San Pablo — Arequipa
  • Ximena Gutierrez-Vasques, PhD, UNAM
  • Andres Campero, PhD Student, MIT
  • Karol Baca-Lopez, PhD Student at Autonomous University of the State of Mexico
  • Marcell Llerena Paricahua, Student, Universidad Católica San Pablo — Arequipa
  • Dehua Chen, Msc, Federal University of Minas Gerais
  • José Luis Olivares Castillo, Student, UNAM
  • Rel Guzman Apaza, Research Assistant, Universidad Nacional de San Agustin
  • Romer Rosales, Director of Artificial Intelligence, LinkedIn
  • David Camilo Alvarez Charris, MSc, University of Edinburgh
  • Lisette Elizabeth Espín Noboa, PhD. Student at GESIS and University of Koblenz-Landau
  • Pablo Francisco Hernandez Leal, Researcher, Borealis AI
  • Verónica Vilaplana, Associate Professor, Technical University of Catalonia
  • Héctor Ricardo Murrieta Bello, Computational Engineering, Universidad Nacional Autónoma de México
  • Alfredo Alejandro De la Fuente Briceño, MSc, Skolkovo Institute of Science and Technology
  • Kevin Christian Rodríguez-Siu, Research and Teaching Assistant at Universidad Católica San Pablo in Arequipa-Perú
  • Sergio Martin del Campo Barraza, Ph.D., Post-doc, Luleå University of Technology, Sweden
  • Francisco Xavier Sumba Toral, Student, Concordia University

Global AI Strategy

To conclude, Omar also covered how AI will be impacting the global economy. Many countries have developed an AI strategy, with the US and China competing to be the best. However, Mexico is the only Latin American country with an AI strategy.

Latin America is lagging behind simply because there is not enough adoption of AI and lack of resources and infrastructure for adequate training.

Some of the adoption challenges include:

  • Lack of investment and awareness of the effects of AI
  • Shortage of talent specialized in AI
  • Lack of infrastructure and training data

To address these challenges, LATAM governments should try to seek ways of retaining talent, as part of the shortage comes from many engineers coming to work in the US. They should also be willing to take more risk to incentivize companies to adopt AI as well as incorporate AI into education to generate more talent.

If you are interested in learning more about the State of AI in Latin America, Endeavor, a non-profit specializing in helping LATAM entrepreneurs has published a report on AI Adoption (Spanish). There is also a great podcast from Game Changer that discusses this report.

Special Thanks

This workshop would not have been possible without the support of our Sponsors:





Over 90 abstract submissions were received. Thanks to the following members of the Latinx in AI community and supportive allies for helping review the submissions.

  • Meire Fortunato, Deep Mind
  • Andrés Campero, MIT
  • Alejandro Galindo, Iris Automation Inc
  • Andres Duque,
  • Jose Lugo-Martinez, Carnegie Mellon
  • Jorge Pérez, Universidad de Chile
  • Anthony Ortiz, Montreal Institute for Learning Algorithms (MILA)
  • David Uminsky, University of San Francisco
  • Gilmer Valdes, University of California, San Francisco
  • Gabriela de Queiroz, IBM / R-Ladies
  • Luis C. Lamb, Federal University of Rio Grande do Sul, Brazil
  • Romer Rosales, LinkedIn
  • Sergio Martin del Campo Barraza, Luleå University of Technology
  • Sergio Guadarrama, Google AI
  • Jose Oramas M., KU Leuven ( ESAT-PSI )
  • Henry Anaya-Sánchez,
  • Ángel Castellanos González,
  • Pablo Samuel Castro, Google Brain
  • Edgar A Duenez-Guzman, DeepMind
  • Mariano Phielipp, Intel
  • Pablo Francisco Hernandez Leal, Borealis AI
  • Andrea Treviño Gavito, Northwestern University
  • José Oramas M., KU Leuven ( ESAT-PSI )
  • Diana Diaz Herrera, Wayne State University
  • Juan Camilo Gamboa Higuera, McGill University
  • Joseph E. Gonzalez, UC Berkeley
  • Jorge Luis Guevara Diaz, IBM Research
  • Mario Banuelos, California State University
  • Alisa Zhila, CIC-IPN
  • Elvis Saravia, National Tsing Hua University
  • Pablo Fonseca, Montreal Institute for Learning Algorithms
  • Alvaro Riascos, University of los Andes and Quantil
  • David Camilo Alvarez Charris, University of Edinburgh
  • Hugo Jair, INAOE
  • Tiago Pimentel, Universidade Federal de Minas Gerais
  • Gonzalo Barrientos, University College London
  • Juan Pablo Muñoz, Intel Labs
  • David Alvarez Melis, MIT
  • Iván Venzor, FEMSA
  • Pablo Rivas, Marist

Lastly, a thank you to our volunteers to helping ensure the first ever Latinx in AI workshop @NeurIPS was a success.

We are creating a public directory of LatinX individuals active in artificial intelligence, machine learning, and data science. This directory will be maintained by the LatinX in AI (LXAI) organization.

If you are organizing an event related to artificial intelligence, this list serves as a resource for potential speakers.

Add to our directory —

Stay Up to Date with LXAI

Subscribe to our monthly newsletter to stay up to date with our community events, research, volunteer opportunities, and job listings shared by our community and allies!

Subscribe to our Newsletter!

Join our community on:

Facebook —

Twitter —

Linkedin —

Private Membership Forum —

If you enjoyed reading this, you can contribute good vibes (and help more people discover this post and our community) by hitting the 👏 below — it means a lot!



Sebastian Anaya

Salesforce Analytics Champion | Consultant @ Accenture | Co-Founder at Fuerza Ventures