Sachi Shenoy

Great technological breakthroughs, like AI today or the Internet a few decades ago, hold great promise of augmenting human experience. But left unchecked, they can upend industries and risk deepening the divide between the haves and have-nots. Unlike yesterday’s technological upheavals, today, we are equipped with more data and the hard-earned lessons of the past. These lessons tell us that we must be intentional about distributing the benefits of technology across our communities.

In the hubbub surrounding artificial intelligence and its role in social impact initiatives, the discourse usually focuses on bias, and a fear that algorithms and automated workflows will perpetuate, and potentially exacerbate, human bias. This is certainly a risk if the models are trained on biased historical data.

The temperature was over 110 degrees Fahrenheit the day I met Roopa, a widowed mother of three, on her tiny plot of land in rural Jharkhand, India. She recalled that in years past, she could grow vegetables for a little income. Now, she told me, all she wanted was to grow enough to eat. That was in 2009—14 years ago. Roopa was a woman boxed in by multiple, simultaneous crises: food insecurity, lack of education, poor housing, chronic illness—and global warming.