7 Ways You Can Help Correct Bias In Artificial Intelligence (AI)

Generative Artificial Intelligence, also known as AI, is no longer a thing of the future. It is here now and is reshaping industries and life in ways previously unimaginable. 

From education to healthcare, entertainment, finance, retail and manufacturing, just to name a few, generative AI is leaving its mark on the overall human experience and how businesses and industries operate, innovate, and compete. 

At The Bean Path, we are recognizing May as AI Month. Our goal is to demystify AI by showcasing its transformative power across various sectors and bringing together individuals from all backgrounds to learn, connect, and explore the future of technology. 

You don’t have to be a techie to participate or appreciate AI Month. In fact many common- non-tech people are using AI on a daily basis and may not know it. Artificial Intelligence is an integral part of our everyday lives, reshaping how we work, communicate, and interact with the world around us. From virtual assistants to personalized recommendations, AI is revolutionizing the way everyday people experience technology and navigate the complexities of modern life. 

Here are a few stats to consider.  97% of mobile users are using AI-powered voice assistants. More than four billion devices already work on AI-powered voice assistants. Think Amazon’s Alexa and Apple’s Siri.  AI spending in the retail industry is expected to reach $20.05 billion by 2026. Many companies are embracing AI to support their customer service. 80% of marketers use chatbots as part of their customer experience strategy.

Although groundbreaking, AI technology is not without its challenges. As the Principal Applied Scientist For Amazon Artificial Intelligence, my job focuses on creating fairness and identifying biases in these technologies.

“Traditional AI” can be thought of as algorithms that scan previous data from the past in order to create predictions about things in the future. “Generative AI” more specifically are algorithms that when given a prompt will create new content (in the form of language/text, images, or video) that is similar to patterns in content from the past. The concern comes when there isn’t a wide balance of information that was used to create the algorithms and it’s partially discriminatory towards certain groups. Other concerns involve results that contain toxic/discriminatory info, inaccurate information, or infringing upon intellectual property/copyrighted material. 

Biases in AI algorithms stem from various sources, including biased training data, flawed algorithms, and human biases encoded in the design process. Training data, collected from real-world sources, may reflect societal biases and perpetuate existing inequalities if not carefully curated and balanced. Moreover, algorithmic biases can emerge from the design choices and assumptions made by developers, leading to skewed outcomes that disproportionately impact certain groups.

How do we fix this as a community? 

• We have to continue to educate ourselves about the potential advantages and limitations of the technology. Understanding the basics of algorithmic bias and its implications helps you recognize biased AI systems and advocate for fairer alternatives. 

• Follow reputable sources and experts in the field of AI ethics to stay updated on the latest research, discussions, and solutions related to bias in AI. 

• Advocate for transparency and accountability. Provide feedback on AI products and services you use, highlighting any biases or unfair outcomes you encounter. User feedback can drive improvements and prompt companies to address biases. 

• Support diversity and inclusion. Advocate for greater diversity in tech companies and AI research teams. Diverse teams bring varied perspectives that can help identify and mitigate biases in AI algorithms.

• Use ethical AI products. Choose to use products and services from companies that prioritize ethical AI practices and demonstrate a commitment to fairness and inclusivity.

• Promote ethical data practices. Be mindful of the data you share online and understand how it may be used to train AI algorithms. Opt for platforms that respect user privacy and ethical data practices.

• Volunteer with organizations and initiatives focused on AI ethics, digital rights, and fairness. Your time and skills can contribute to efforts aimed at addressing biases in AI. Collaborate with academic and community groups working on AI fairness projects, providing support and spreading awareness within your community.

As AI continues to evolve and permeate society, it is essential to recognize its potential to empower individuals, enhance quality of life, and drive positive social change. Embracing this technology with caution and foresight, businesses, creatives and everyday people can harness its potential to drive innovation, enhance productivity, and create value for society.

*AI was used to help write parts of this editorial.

About Dr. Nashlie Sephus

Dr. Nashlie H. Sephus is the Principal Applied Scientist For Amazon Artificial Intelligence (AI) focusing on fairness and identifying biases in these technologies. She formerly led the Amazon Visual Search team in Atlanta, which launched visual search for replacement parts on the Amazon Shopping app in June 2018. This technology was a result of former startup Partpic (Atlanta) being acquired by Amazon, for which she was the Chief Technology Officer (CTO). Additionally, Dr. Sephus is the developer of the JXN Tech District and founder of The Bean Path, a  non-profit organization based in Jackson, MS dedicated to creating equity in STEAM opportunities by increasing access to tools, knowledge & networks to underserved communities, particularly in Mississippi.