AI can be biased. Lack of representation in datasets have been one of the most dangerous things about building algorithms for the future. One of the ways to fight this is to have a diverse set of eyes looking at the AI process of data collection and building models.
The best way to fight bias is to democratize AI.
“85 percent of AI professionals believe the industry has become more diverse over the past few years; of those, 91 percent think that shift is having a positive impact. 74 percent of AI professionals believing diversity hasn't improved say the industry must become more diverse to reach its potential.”
Let Joy Buolamwini explain:
This is just one of the more recent news headlines we see in AI, but we have actually been seeing a diversity problem in AI for several years prior:
Lack of representation is a major challenge the AI industry faces and we must work harder to increase diversity in AI to avoid algorithmic bias.
We have compiled a list of organizations focusing on inclusivity and diversity in the AI and tech industry. If you know of any more organizations or resources, let us know and we will add them to the list!
“Black in AI (BAI) is a multi-institutional, transcontinental initiative creating a space for sharing ideas, fostering collaborations, and discussing initiatives to increase the presence of Black individuals in the field of AI.”
The organization hosts annual technical workshop series, runs mentoring programs, and maintains various fora for fostering partnerships and collaborations. Simply fill out this Google form to join. They also accept applications of allies.
Diversity.AI’s bio reads: “Preventing racial, age, gender, disability and other discrimination by humans and A.I. using the latest advances in Artificial Intelligence”
The organization leads research in using AI to detect, analyze, prevent and combat human biases and discrimination.
Their goals include:
- Establishing a discussion forum for thought leaders in AI, racial, gender and ageism issues, biologists, policy makers and ethicists;
- Scheduling meet ups to discuss the possible discrimination issues in AI and strategies to minimize exclusion and bias;
- Developing a range of guidelines and validation mechanisms to test the deep-learned systems and other cognitive computing solutions for racial, gender, age and ethnic bias;
- Developing open access datasets to allow developers to train the algorithms on minority data sets.
As a frequent publisher of research papers, the organization shares datasets they’ve used in their AI models.
Blacks in Technology is a leading community-based media organization and aims to establish “a blueprint of world-class technical excellence and innovation by providing resources, guidance and issuing a challenge to our members to surpass the high mark and establish new standards of global innovation.”
The organization has 13 chapters in the U.S. and is constantly growing to increase representation of Black men and women in the tech industry.
POCIT is an extensive content resource for people of color with articles, interviews, a newsletter and a podcast. They also have a large Twitter community to discuss the most important topics for people of color in tech today—including diversity in AI. On top of that, they boast a frequently updated job board.
The program is aimed to pave the way for underrepresented students into the tech industry.
According to a blog post, “Tech Exchange, a student exchange program between Google and 10 Historically Black Colleges and Universities (HBCUs) and Hispanic-Serving Institutions (HSIs), hosts students at Google’s Mountain View campus and engages them in a variety of applied computer science courses. The curriculum includes machine learning, product management, Theory of Computation and database systems, all co-taught by HBCU/HSI faculty and Google engineers.”
Women in Machine Learning + Data Science
Established in 2013, WiMLDS have over 100 chapters globally. Its mission is to, “create opportunities for members to engage in technical and professional conversations in a positive, supportive environment by hosting talks by women and gender minority individuals working in data science or machine learning, as well as hosting technical workshops, networking events and hackathons.”
Khipu is an event promoting diversity in AI in Montevideo, Uruguay. It’s a meeting of AI leaders to promote diversity in the AI community in Latin America. According to its website, the goal of the event include:
- To offer training in advanced machine learning topics, such as deep learning and reinforcement learning.
- To strengthen the machine learning community by fostering collaborations between Latin American researchers, and creating opportunities for connections and knowledge exchange with the broader international community.
- To grow awareness around how AI may be used for the benefit of Latin America
Read its 2019 event recap here.