Anima Anandkumar: Developing next-generation AI algorithms

Women Special

|

1999 was the year of prominent growth for the PC gaming market, redefined modern computer graphics and revolutionized parallel computing, and recently, GPU deep learning has ignited the modern AI. For the Da Vincis and Einsteins of this time, NVIDIA pioneered a supercharged form of computing loved by the most demanding computer users in the world — scientists, designers, artists, and gamers. Fueled by the insatiable demand for better 3D graphics, and the massive scale of the gaming market, NVIDIA has evolved the GPU into a computer brain at the exciting intersection of virtual reality, high-performance computing, and artificial intelligence. Powering this AI revolution is Anima Anandkumar, the Director of Machine Learning Research at NVIDIA (since 2018) and the Bren Professor of Computing at California Institute of Technology. She is a renowned name in the world of GPU computing. Her research is focused on tensor algebraic methods, deep learning, and non-convex optimization. Here’s her story of becoming a world-class AI expert.

Anima was born in Mysore, India. Belonging to a family of engineers and mathematicians, appreciably she studied electrical engineering at the Indian Institute of Technology Madras and graduated in 2004. In 2009, at Cornell University, she earned a Ph.D. in EE under the supervision of Lang Tong. Post-doctorate, she was a scholar at Massachusetts Institute of Technology, where she worked in the Stochastic Systems Group with Alan Willsky. Later, in 2010, she joined the University of California as an assistant professor and began her work on tensor decompositions of latent variable models. She joined Microsoft Research in New England in 2012, as a visiting scientist after which she achieved a National Science Foundation Career Award for investigating big data and social networks in 2013. She is also the recipient of the Sloan award. Soon, she joined as an associate professor with tenure at UC, Irvine and after almost a year, she was honored with Bren named chair professorship at Caltech in June 2017. From 2016-2018, she was a principal scientist at Amazon Web Services, where she worked on and the launch of Amazon Sagemaker, and represented AWS on various renowned platforms to discuss deep learning, until she finally came with NVIDIA as director of ML research in September 2018. The list of her accomplishments is never-ending. Recently, she was nominated to the World Economic Forum’s Expert Network consisting of leading experts from academia, business, government, and the media. She has also been featured in documentaries by PBS, KPCC, wired magazine, and in articles by MIT Technology Review, Forbes, Yourstory, O’Reilly media, and so on. She is the co-director of the Center for Decision Making, Optimization and Learning (DOLCIT) at Caltech and is on the scientific advisory committee for the Center for Autonomous Systems and Technologies (CAST). Anima has worked to democratize AI, promote ethical use and bring diversity and inclusion in the industry. Therefore, in 2018, she was presented with the good tech award by NYTimes for her efforts. 

“I believe that ‘diversity is the mother of creativity’. A diverse team of different dimensions will bring about new ideas. I encourage open discussions and make sure there is a healthy and respectful environment.”

Success and learning’s

The secret to her success Anima notes is due to her nature of being naturally drawn to challenges. “My mom likes to remind me of my childhood when every year, my family celebrated the Indian New Year or Ugadi. The tradition was to consume a mix of jaggery (cane sugar) and neem (an extremely bitter herb) and it symbolized accepting both happiness and challenges. As a kid, apparently, I only wanted the neem!” shares Anima. “I was naturally driven to challenges and still continue to look for new ones.” For instance, in 2012, she switched her focus to the problem of learning latent variable models. “This is a form of unsupervised learning, and it is considered as one of the hardest problems since there is no supervision or labels available at the time of training. The goal is to design algorithms that can automatically extract hidden or latent factors from data,” explains Anima. “As an assistant professor, this was a big risk. But I embraced it and enjoyed the whole journey. I invented a new class of tensor algorithms, in a seminal paper, along with an excellent group of collaborators. Prior to this, tensors were not even on the “radar” of AI researchers. They were not considered a viable tool for machine learning. Fast forward today, tensors are ubiquitously associated with machine learning.”

Anima understands that without risk, there are no rewards but she realizes the needs to be in healthy balance that comes with efficient planning. Another important factor that needs to be conquered is to build lasting relationships with potential partners for which she uses the trust-radar and crowdsource information from the community. “I recently learned about the “trust radar” from Prof. Alison Kluger at Stanford business school,” says Anima. “This includes empathy, communication, expertise, and transparency. Looking back, these principles have been my guiding force. I focus on the long term. I invest in communicating my vision and giving others the opportunity to voluntarily come on board to that. I commit to empowering junior colleagues/students to succeed. I emphasize on creating a healthy environment where diverse voices can thrive.” Therefore, when hiring people for her firm she looks for a combination of technical competence, depth, and breadth of expertise, good communication skills, and the team spirit. She emphasizes, “I believe that ‘diversity is the mother of creativity’. A diverse team of different dimensions will bring about new ideas. I encourage open discussions and make sure there is a healthy and respectful environment.”

Anima Anandkumar

Anima Anandkumar, Director of Machine Learning Research, NVIDIA

When a woman is not afraid, a lot of things happen

Although even after being an accomplished leader, being a woman, Anima is often mistaken to be a student or an intern because she does not fit the stereotype of a chaired professor or director. Several people have tried to cast her down by commenting on her capabilities but never did that make her lose her focus. She shares, “There is pervasive sexism all around us and #meToo movement is just the ‘tip of the iceberg’. Barriers for women and minorities continue to exist. I try to raise awareness and I speak openly about it on social media like Twitter.” Anima was a major contributor in starting the #protestNIPS to demonstrate the problematic behavior at the annual premier conference on AI which went by the name “NIPS”.  There was a lot of push-back and even gun-threats but that didn’t make her leave her ground. She tells Aspioneer, “The name was emblematic of “bro-culture” coming into our academic conferences and parties being held where scantily-clad women were displayed for the pleasure of the mostly-male audience. It was a hostile and unhealthy environment for women and minorities. I rallied the community for a name change and more. By raising awareness, we were able to get the name of the conference changed. We also enforced a code of conduct for all participants. I am hoping this is the beginning of a revolution in AI that promotes diversity and inclusion. I strongly believe this will be a critical ingredient for us to develop better AI techniques that promote fairness, interpretability, transparency, and robustness.” One of her major concerns is the depleting number of women enrolling in computer science field specifically since the 80s. She says, “We can’t expect that technology will always make the situation better for women. There is more centralization of data and decision making, and monopolies controlling different aspects of our daily lives. Activism will always have a role. #meToo has been a powerful movement. We need to continue fighting for what we truly deserve: full equality. I don’t believe in “lean in.” We need to “barge in” and take down the walls of patriarchy.”

Doing her part she has dedicated herself to life-long learning and growth. “I foresee that AI will keep me engaged all my life! We have barely made a dent on some of the deepest challenges in AI,” says Anima. “I am fascinated to explore more on this life’s great adventure.”

Share on email
Share on linkedin
Share on twitter
Share on facebook
Share on whatsapp

All designs displayed on this page are the intellectual property of © Aspioneer BizByts Marketing Pvt. Ltd.