The Ethical AI Supply Chain: Protecting the Soul of AI

A few months back, I had the opportunity to visit our offices in Nairobi and Uganda, as the final step in my on-boarding at Sama.My colleagues had warned me of the vibrancy and grit to be expected, but in the 8 days I spent in East Africa, the “human” in human-powered AI began to come alive.

PHOTO: Liliosa Mbirimi (left), Heather Gadonniex and Renuka Kaimal (right), pictured with Julius. Julius is a former Sama employee that now employs over 25 people in the business he started with Sama GiveWork Challenge funds.If you’re not familiar, Sama provides training data and algorithm validation to 25 percent of the Fortune 50. While our headquarters is in San Francisco, we’re one of the largest AI providers, and one of the largest digital employers, in East Africa.Before joining Sama, I knew AI had a human element - that the algorithms we train couldn’t exist without humans.After all, every marketing slick touts how the combination of human and machine intelligence will make us smarter, more efficient -- superhuman.What we’re not told is that behind every machine learning algorithm, there’s a band of people tasked with labeling the data used to train AI technologies.For example, a self-driving car gets its initial smarts from a human reviewing thousands of images and/or video and labeling the cars, trucks, sidewalks, pedestrians, etc., that are present. This is how the car learns to be autonomous.These humans creating training data for algorithms are the soul of AI.Earlier in my career, I had a deep focus in measuring and managing environmental and social impacts of supply chains and implementing corporate impact programs. Carbon footprint and circular economy were part of my everyday lexicon.Now, staring into the eyes of our human workforce -- the real-life AI supply chain -- I started to think of the implications for supply chain ethics and sustainability. And, I realized how serendipitous it was to be working at an organization that embraced paying a living wage and equally distributing opportunity from the start, thanks to the vision of our founder Leila Janah.

PHOTO: The Sama Gulu team reviewing the results of our Gulu Living Wage Audit. Team members learned that Sama pays above a living wage in Gulu, Uganda.Like most media coverage, this episode of the McKinsey Podcast discusses bias, fairness, and ethical use of AI, however, the humanitarian aspects of today’s digital supply chain are often overlooked.How do we manage and measure the human factors present in the digital supply chain? And, how do we ensure that we’re using each and every opportunity to build a more equitable, up-skilled workforce as the Fourth Industrial Revolution explodes?As an industry, we need to connect tech laborers to dignified digital work, by aiding in the life and professional skills development required to excel in today’s digital economy.It provides people with training and support and equips them with work experience to launch long-term careers. We also need to ensure we're adopting business practices that help us advance progress toward overarching Sustainable Development Goals.Protecting the soul of AI means making sure the humans powering AI technologies know they’re a part of something much bigger than labeling data, but rather supplying Fortune 50 companies with the training data they need to fuel their AI algorithms.It also means ensuring workers in the AI Supply Chain are treated with dignity and respect, paid a living wage, and are given the opportunity to build life and career skills that set them up for long term success.AI is a force that will drive the next, clean industrial revolution. From smart cars to automated anti-poaching systems, improved healthcare outcomes to targeted product selection -- AI has the power to fuel humanity.As we continue to solve the many training data challenges associated with adopting AI, let us, as an industry, ensure we’re also advocating for ethical supply chain standards in the AI Supply Chain.Learn more about our impact sourcing mission.