Founder of AI for Good, Kriti Sharma, wants to highlight the good work people including herself are doing to solve societal problems with the help of AI technologies.
One is improving women’s access to pre-natal healthcare in the Democratic Republic of Congo. “A pregnant woman has to walk 17 hours to her nearest, rural, prenatal clinic to get a check-up. What if she could get a diagnosis on her phone instead?” asked Sharma.
Another example is that of a woman in South Africa who is suffering domestic abuse. Sharma asked: “What could AI do for those one in three women? If it isn’t safe to talk out loud, they could get an AI service to raise the alarm or get financial and legal advice.”
She said that these great opportunities to use AI for the good of the world are often overshadowed by Hollywood-inspired media panic. She did admit, however, that there is serious cause for concern about the designers of AI continuing the status quo because this is resulting in humans having flawed decisions being made about us by machines.
Sharma (like me) has had to sit through thousands of pre-roll adverts for pregnancy tests on YouTube videos but sees this as an annoying result of an algorithm having made a decision about her based on her age, gender and other factors as an annoyance that won’t really affect her life chances.
Serious problems do arise for individuals and communities when algorithms make assumptions such as; ‘a Black or Latino person is less likely than a white person to pay off their loan on time’ and ‘a person called John makes a better programmer than a person called Mary’. The individual suffers an immediate negative outcome when such an assumption is applied to them and the community they belong to suffers as those historical decisions mean they will not be given a fair chance to prove themselves but will instead be tainted by association.
Sharma also highlighted the worrying tendency to have personal voice assistants, such as Siri, Cortana and Alexa, coded as female. Meanwhile, Ross the robot lawyer and highly knowledgeable IBM Watson are coded as male. “Poor robots. Even they suffer from sexism in the workplace,” Sharma lamented. In her view, this has come about because the “brightest minds” who are creating this technology have chosen to create it in the style of 1950s Mad Men secretary.
But all is not lost. “The good news about AI is that it is entirely within our control,” she said, with three specific actions in mind. “One, we can be aware of our own biases and the bias in machines around us. Two, we can make sure that there are diverse teams building this technology. And three, we have to give them diverse experiences to learn from.”
Expanding on her point about diverse teams, Sharma said that to make AI, people from all kinds of backgrounds have to be involved in its creation. They include storytellers who can help create personalities for the AIs, as well as problem solvers and solutions finders who have faced different challenges.
“Because when people from diverse backgrounds come together, when we build things in the right way, the possibilities are limitless,” she said.
Kriti Sharma was speaking at TEDxWarwick.