Five years ago, Stanford computer science Professor Fei-Fei Li became worried about the direction AI was headed. The technology had the potential to impact all of humanity, yet a very narrow group of people were developing this technology. As philanthropist Melinda Gates has said, AI’s researchers and developers mostly consisted of “guys in hoodies.”
In 2015, Li and several colleagues founded a program designed to inspire young women to study artificial intelligence. Originally called SAILORS, it's now called AI4ALL. In the four years since it was founded, more than 100 women have gone through its intensive, two-week sessions at the Stanford AI Lab, focusing on introducing AI technology in the context of its human-centered applications and ethical issues.
Stanford AI4ALL graduates have gone on to run middle school AI workshops, start AI and robotics clubs in their own schools, conduct AI research in university labs and connect AI with disciplines ranging from medicine to the environment, and from dance to art.
At the launch of the Stanford Institute for Human-Centered Artificial Intelligence (HAI), two Stanford AI4ALL graduates described how the program had changed their lives — and then conducted a fireside chat with philanthropist and Microsoft co-founder Bill Gates.
Amy Jin was in ninth grade when she entered the inaugural session of A14All. She worked on a project to use computer vision to combat hospital-acquired infections. When the program ended, her Stanford mentors guided her in further study and Jin embarked on a project to use artificial intelligence to monitor surgical performance.
“Seven million people suffer surgical complications every year, and about half of those are preventable because they were the result of poor performance by surgical teams,” she said. But evaluating surgical performance is time-consuming and quite labor intensive. “Athletes and musicians benefit from coaching, why not surgeons,” Jin said.
Jin, now a freshman at Harvard, and her team developed a deep learning approach to monitor and evaluate surgical skill. They tracked the hand movements of surgeons as they performed gall bladder operations. A paper describing her work was named the best paper at the Machine Learning for Health workshop at the 2017 Neural Information Processing Systems Conference. NeurIPS, as the conference is known, is one of the largest gatherings of AI researchers in the world.
Stephanie Tena-Meza, a junior in high school, is a first-generation Mexican American who lives in an agricultural community with high rates of poverty. “I experienced how growing up in an underserved community can hinder one’s intellectual curiosity,” she said.
In middle school, she came across one of the few free coding clubs in the area. The program inspired “a passion for computer science,” said Tena-Meza, who had seen the damage to the water supply caused by agricultural and pesticide runoff. She wanted to know more about that issue and explored it via a ninth-grade science project.
Joining AI4ALL, she worked with her mentors to develop an AI-focused project to measure the quality of water in the Colorado River. “None of my projects would have been possible without programs like AI4ALL that exposed me to the computing field at an early age,” she said. Hoping to inspire other young people, Tena-Meza founded a computer science and AI club at her former middle school and a Girls Who Code club at her high school.
By 2016, Stanford AI4ALL was overwhelmed with applicants from around the world, Li said. She met Melinda Gates that fall, and they discussed their desire to make AI and tech more inclusive. Gates coordinated an early funding round, and what began as a Stanford program became a national non-profit in 2017. AI4ALL now serves young men and women who are underrepresented in AI and will be on 11 campuses this summer, including Princeton and UCSF, Li said.
Bill Gates: AI’s greatest ROI
During his discussion on stage with Jin and Tena-Meza, Bill Gates said computers in classrooms have not yet made significant improvements to education. He was much more bullish on the benefits of AI, saying it is already contributing to better health and lower rates of premature births in the developing world.
Fighting disease in the developing world is a major focus of the Bill and Melinda Gates Foundation.
The Gates Foundation has been exploring the uses of AI to help identify the causes of premature birth and malnutrition. “With AI we are able to take all that data and find some really low-cost interventions,” Gates said.
“At the classroom level, the value of computers is virtually nil. So that’s good news: We didn’t accentuate the digital divide; the schools with computers are just as bad as the ones without them,” he said.
When asked by the AI4ALL graduates about the challenges around ethics in AI, Gates said, “The world hasn’t had that many technologies that are both promising and dangerous,” he noted, pointing to nuclear energy and nuclear weapons. “With AI, the power of it is so incredible, it will change society in some very deep ways, so it’s great that Stanford’s stepping up.”
“Given the general-purpose nature of these technologies to find patterns and insights,” said Gates, “It’s a chance to … look not just at correlation but to try interventions and see causation, as well. It’s a chance to super charge the social sciences.”
Although it’s early days, AI has had some notable successes fighting disease in developing countries. “If you give kids in some countries an antibiotic once a year that costs two cents – Azithromycin — you save a hundred thousand lives,” Gates said. Without machine learning, it would not have been possible to understand the solution.
Africa, Gates said, is where investments in artificial intelligence are likely to have the highest return on investment. “If everybody in the United States lived to 100, it would not match what we can do in the developing world in terms of net change to human benefit,” Gates said.
Before breakthroughs in machine learning, says Gates, “many deep societal problems were not tractable.” Now, however, with the help of AI, “It’s a chance, whether governance, education or health, to accelerate the advances in all the sciences.”