Latinx in AI Coalition

Creating harmony and opportunity in AI for the Latinx community

AI Technology has begun impacting  all sectors of society, including the Latinx community. AI developed for retail such as Amazon Go, LowesBot and Focal Systems Deep Learning for Retail solution (Y-Combinator 2016), threatens to displace workers who rely on their jobs as cashiers and inventory restockers. Out of a population of 3.5 million truck drivers in the U.S. — 3.1 million of whom are commercial driver’s license (CDL) holders 38.75% are minorities according to data compiled by the American Trucking Associations (ATA). Hispanic men make up the 2nd largest demographic of the truck driver population at 14.6%, based on 2014 data. Few autonomous trucking companies, including Google, Tesla, or Uber, are considering the effect they’ll have on the current trucking labor workforce or their families.

Additionally, Machine learning algorithms can encode a discriminative bias during training with real world data in which underrepresented groups are not properly characterized or represented. A question quickly emerges: how can we make sure Machine Learning does not discriminate against people from minority groups because of the color of their skin, gender, ethnicity, or historically unbalanced power structures in society? Even more so, as the tech industry does not represent the entire population, underrepresented populations in computing such as Hispanics, women, African-Americans, and Native Americans have limited control over the direction of machine learning breakthroughs. As an ethnicity, the Latinx population, is an interesting case study for this research as members are comprised of all skin tones with a wide regional distribution across the world.

In five years, our organization will be a professional research organization, focused on improving access to opportunities, livelihood, and education for Latinx identifying individuals around the world. We’ll foster cross collaborations in research between top universities in the US, South America, and European nations. We’ll secure access to research grants which will enable us the opportunity to fund others work while being mentored by professionals within our research committee and organizational network.

Through our own research initiatives we’ll have created an open source dataset of Latinx images and voice recordings to improve the robustness and representation of latinx populations in commercially viable AI technology. This data collection pipeline will be simplified through the development of a mobile phone application, enabling individuals without access to computers or internet cafes in developing countries to participate and ensure they are accurately represented.

Access to educational opportunities for retraining displaced workers will be cemented by our partnerships with traditional education programs as well as MOOCs (Massively Open Online Courseware). We’ll create a natural pipeline of students interested in using Artificial Intelligence to solve real world problems in their own communities.

Organizations which focus on the future of work for underserved communities and fighting the effects of bias in machine learning algorithms include other diversity focused professional organizations and nonprofits including, Women in ML organization, Black in AI organization, AI4ALL as a few examples. Each of these groups, including our own focuses on a niche population which will be uniquely affected by the future due to advancing AI technology. Our focus on the Latinx population, gains us the advantage of appealing to individuals comprised of all skin tones with a wide regional distribution across the world.

Technology is integral to our solution to ensure a simplified data collection pipeline through the development of a mobile phone application, enabling individuals without access to computers or internet cafes in developing countries to participate in our research initiatives and ensure they are accurately represented. The open source dataset of Latinx images and voice recordings will improve the robustness and representation of Latinx populations in commercially viable AI technology.

Our goals this year include spreading awareness of the oncoming effects of AI on Latinx communities, providing access to education for retraining, and conducting research to improve accuracy and representation in data sets used in the development of AI technologies.

The outcomes of our initiatives will be measured by the amount of quality data we are able to collect and the applications we are able to improve. By improving the robustness of facial recognition and language software to serve latinx populations, we’ll be improving access to technology while reducing chance of bias and reinforcement of negative stereotypes.

In the next three to five years, our organization will be a professional research organization, focused on improving access to opportunities, livelihood, and education for Latinx identifying individuals around the world. We’ll foster cross collaborations in research between top universities in the US, South America, and European nations. We’ll secure access to research grants which will enable us the opportunity to fund others work while being mentored by professionals within our research committee and organizational network.

We will motivate university professors and graduate students to conduct research in this domain regarding the impact of bias in machine learning on underrepresented groups, specifically Latinx populations. We are empowering those outside of academia by spreading awareness of the effects of this technology on their communities and their future employment opportunities. We are providing access to free educational materials and scholarships ensuring they may be retrained for the future of work. We’ll retain these individuals through a professional network across the world, hosting annual “State of Latinx in AI” symposiums alongside NIPS, and providing ongoing research opportunities.

With our education and outreach initiatives we engage 50-100 attendees monthly in person, providing information and hands on practice in the use of AI technology. Our digital outreach across most social media channels averages over 6,000 views per post. Our partnerships with other diversity focused organizations, research institutions, and education companies such as Udacity, University of California Berkeley, and Universities in South America, ensures we are reaching our target audience and are able to create pipeline for individuals to continue their education and transition into a digital workforce and economy.

Within 12 months, we expect to increase our outreach across U.S. Universities, serving thousands of students enrolled in or interested in pursuing a degree in Artificial Intelligence or a related industry. Within 3 years, our goal is to be a global organization serving hundreds of thousands across the US, South America, and Europe. The pipeline we would create from labor workers, transitioning those interested to professionals in engineering and research through scholarship opportunities will provide an immense opportunity and impact both in their individual lives, the lives of their families, and to society as a whole.

Our organization board includes professionals experienced in education, research, entrepreneurship, business development, and outreach. Our board members background and experience is comprised of traditional education and alternative education programs with history of native US (first or second generation), refugees, and those who have immigrated to this country. Combining this diversity of experience and perspective ensures we can leverage our networks and make an impact for our targeted audience around the world.

The Latinx in AI Coalition is currently fiscally sponsored under the umbrella of the Accel AI Institute, a 501c3 Non Profit education organization, which is funded by corporate sponsorships, community donations, and founders investments. Once specific milestones have been achieved, the Latinx in AI Coalition will be registered as a 501c3 Nonprofit research organization. Our strategy for long term stability and expansion includes a nonprofit membership program with student, professional, and corporate tiers. Accessible membership tier rates with appealing benefits including opportunities for access to research mentorship, professional development, a vastly growing network of professionals, platform for sharing one's own work, and voting rights for the future direction of the organization will enhance community engagement and organizational stability.

We are applying to Solve because we know the recognition and network of MIT alumni and professionals would allow us to expand our research and outreach initiatives globally. With your help we'll be able to launch to the next level in our journey of up-skilling underserved communities and reducing bias in AI ensuring a harmonious future for everyone.

Key barriers to our success are limited resources and time available to conduct the outreach needed to raise funding, hire staff, and empower volunteers. Solve can help us extend our network, strategize growth and strengthen our business development skills to drive impact ensuring our success and impact on latinx communities and the world as a whole.