Supporting the AI researchers of tomorrow
Peter Henderson Peter Henderson
September 18 6 min

Supporting the AI researchers of tomorrow

As the modern AI revolution has made headlines around the world, machine learning has exploded as a research field. New subjects, new students, and new approaches abound, and it’s hard to keep up.

Enter Machine Learning Summer School, an event held around the world that brings students, researchers and experts from machine learning, AI, and other related fields together for a chance to swap ideas and study the latest techniques.

Element AI was a proud sponsor of the July MLSS event in London, U.K., helping to bring scientists from Ghana, Sudan and South Africa to participate in the two-week program. We caught up with three of them after the event to talk about what it’s like to be immersed in machine learning, how collaboration can help push the science forward, and where they’re going next.

Sicelukwanda-Zwane

Sicelukwanda Zwane recently finished a master’s in computer science at the University of the Witwatersrand in Johannesburg, South Africa. For him, the London MLSS was a chance to learn more outside of his own research, which primarily focuses on autonomous robot intelligence. He noted the diversity of the event — nearly 200 people from more than 50 countries attended, according to the organizers.

“The effect of MLSS is that you end up fostering a lot of collaboration between researchers that wouldn’t otherwise end of working together,” he said. “You get stronger researchers coming out because we’re exposed to the world’s leading experts presenting on topics they’re passionate about.”

For Zwane, he also got to see the human side. Of the many informal events and gatherings that surround the two-week MLSS agenda, one focused on what it means to do a Ph.D. — for your personal life, for your job prospects, and for your mental health. He hopes to start his doctorate in the next two years, and said that gathering reinforced his decision.

“Sometimes things just line up with the way you’re thinking, and that allows you to be more confident in your decisions,” he said.

Zwane plans to apply for residency programs and long-term internships at Element AI and other tech companies while preparing to publish work from his master’s project and look towards his Ph.D. research. In addition to his research, he runs machine learning boot camps and works for a data science re-skilling training program, the Explore Data Science Academy, that expects to enroll 1,000 students in the fall.

Wilhemina Adoma Pels

Wilhemina Adoma Pels is pursuing a Ph.D. in mathematical statistics at Kwame Nkrumah University of Science and Technology, a university in Kumasi, Ghana. She said the MLSS event was a chance to meet a wider variety of scientists — she is one of only five Ph.D. students in mathematical statistics going into their second year in September, and the only woman.

“Being the only female, I have to strive for it,” she said. “MLSS was very, very inspiring. I met other women in the field, and now I know I’m not the only one. It really motivated me.”

The time in London with other scientists was a good reminder of why she’s pursuing AI research, she said. While the field has been embraced in the United States and Canada, she said some professors in Ghana have been slow to embrace the importance of AI and the shift in research it represents. At her school, there is no Ph.D. program in AI.

Pels said she’s still in touch with other attendees via a Slack group run by the organizers, and made connections with others from West Africa studying abroad and an Indian researcher studying in the U.S.

“I was stuck on a Monte Carlo simulation, validating something I’m doing, for two months with no one to help out,” she said. “They taught us, I talked with some others, and I was able to fix that problem in two days.”

Wilhemina plans to finish her Ph.D. in Ghana and become a lecturer. In addition to her research, she mentors other women at her university and volunteers to teach students and at the junior high school near her childhood home.

Mohammed Merghaney

Mohammed Merghaney recently finished his M.Sc. at the African Institute for Mathematical Sciences, part of the inaugural class of the African Master's in Machine Intelligence program. Part of that program included guest lecturers from around the world, and the students came from all over Africa.

“In our school, I got used to knowing people from different cultures,” he said. “At MLSS that was even more extended. You had many people with many stories, many backgrounds and different ideas. It was an amazing combination.”

Merghaney works in Bayesian machine learning, focusing on uncertainty estimates in deep learning. He said the MLSS agenda, which included tutorials and workshops on fundamental and applied research as well as bias and fairness, helped broaden his understanding of the field. And, while he cited optimization and graphical models as two areas of new interest, it was about more than the math.

“The fairness idea, it was so totally new to me,” he said. “You have to be aware of what you are doing and the effects on other people and other things. You have to understand the context.”

Back in his home country of Sudan, Merghaney said, many professors are set in their ways and don’t fully understand the potential for AI to upend traditional subjects in the world of mathematics and computer science. Pels and Zwane echoed similar sentiments about the academic structures in Ghana and South Africa.

“The students have the ideas,” Merghaney said. “I have done lectures, I have run courses, interacted with a lot of people who think AI can help solve our problems. For the next five or ten years, there might not be that much investment. But I am optimistic, and I do believe that AI can help make life better.”

Merghaney is starting a year-long residency at Facebook in December, then plans to pursue a Ph.D. in Europe. He also works with a startup in Sudan that’s using AI in customer service at a major regional telecom company.

At Element AI, we believe open collaboration is the shortest path to real impact, and that the best way to advance AI research is to work smarter, together. We’re pushing AI science forward in collaboration with our international network of academic fellows and the wider research community. As part of that goal, we’re committed to supporting the next generation of AI researchers.