UN Young Leader Kriti Sharma reveals how AI can be used to help tackle real issues such as domestic violence.
When Kriti Sharma was 15 and living in north-west India, she built her first robot. She had also built her first computer from spare parts, long before maker kits were a thing. Since then she has been building AI technologies to solve global issues, from productivity to education to domestic violence.
Today, she is focused on being a rational voice around AI and its use for effecting positive change to society. A TED talk she gave on the subject in March went viral and was viewed more than 1m times in its first week.
‘The next wave for me is when we start solving deeper social challenges around us using this technology’
– KRITI SHARMA
Sharma, who spoke at the recent Dell Technologies Innovation Series in Dublin, was named in the Forbes 30 Under 30 list for advancements in AI and in the Recode 100 list of key influencers in technology in 2017. She was invited to be a Civic Leader by the Obama Foundation Summit for her work in ethical technology. She is a Fellow of the Royal Society of Arts, a Google Grace Hopper Scholar and has given expert testimony on AI policy to the UK parliament in the House of Lords.
She was appointed a United Nations Young Leader at the General Assembly in 2018 and is an adviser to the UN Technology Innovation Labs.
AI for social good
Speaking with Siliconrepublic.com, Sharma said that people are very much caught up around the fear of AI and the potential for it to take jobs, but are missing the point on how it could be used for social good.
“There are plenty of misconceptions and fears. But it is true that job losses are a very real issue. One thing I have noticed when I teach teens how to build AI, they have their own misconceptions but once they get started and build things hands-on, the perception changes. But it is more about what technology can help us to achieve.
“A lot has to do with moving beyond the clickbait headlines to actual experimentation and experience in this technology and starting to realise that yes, whilst there are ethical issues and we have to make sure to teach machines the right values, the opportunities that are ahead of us are normal,” she said, adding that people’s behaviours and attitudes change and shift.
Sharma continued: “Secondly, about social impact – personally for me because I’ve been in this field for quite some time and leading AI and machine-learning teams in big businesses – I have realised that, as an industry, we could be doing better than what we are doing with it so far.
“If you look at big applications of AI in production today, a lot of it is in advertising and marketing. The Facebook and Google models are really about helping people click more ads by surfacing the right information, feeding a digital addiction and selling through recommendations.
“The next wave for me is when we start solving deeper social challenges around us using this technology and moving beyond the early intervention you are seeing now into more serious issues of healthcare, education, public service, government sector. There are so [many] more opportunities to drive efficiencies there, and part of what I am doing with the AI for Good movement is to bring some of these to life.”
Last year Sharma launched a project in South Africa called Rainbow, which provides predictive insights for victims of domestic violence.
“It is a huge issue. One in three women face domestic violence. It’s the largest cause for women’s homicide and yet it is not talked about. There is too much shame, stigma, embarrassment and judgement associated with the topic.
“However, we found that using machine-learning and AI capabilities, we could start to do better risk assessment of who might be in that difficult situation and also giving them the right information and advice when contacting a human might not be safe or too much for them, and machines can give information in a non-biased, non-judgemental way.”
Sharma said she is doing similar work in healthcare. “Women’s rights can be obstructed sometimes and not everyone will have equal access. Using AI diagnostics, we can give people the right information at the right time. What I am doing is a tiny drop in the ocean but this movement of using technology for social impact is continuing to grow, and I’ve watched amazing people in the last three to four years who are getting more aware … it is part of [the] ‘techlash’ we are starting to see.
“The side effects of early intervention using AI, data technologies and people in the industry themselves are more committed to looking at a broader horizon than just short-term decisions that we have made in the past.”