Life at DeepMind
Avishkar Bhoopchand, a research engineer on the game theory and multi-agent team, talks about his journey to DeepMind and how he is working to raise awareness of deep learning in Africa.
Learn more about Deep Learning Indaba 2022the annual meeting of the African AI community, to be held in August in Tunisia.
What does a typical day at work look like?
As a research engineer and technical leader, no two days are the same. I usually start my day by listening to a podcast or audiobook on my way to the office. After breakfast, I focus on emails and admin before jumping into my first meeting. These include one-on-ones with team members and project updates, as well as diversity, equity, and inclusion (DE&I) working groups.
I try to carve out time for my to-do list in the afternoon. These tasks might include preparing a presentation, reading research papers, writing or reviewing code, designing and running experiments, or analyzing results.
My dog Finn keeps me going while I work from home! Teaching him is very similar to reinforcement learning (RL) – the way we train artificial agents at work. So I spend a lot of time thinking about deep learning or machine learning in one way or another.
Where did your interest in artificial intelligence come from?
During a course on knowledgeable agents at the University of Cape Town, my lecturer showed me a hexapod robot that learned to walk from scratch using RL. From that moment on, I couldn’t stop thinking about the possibility of using human and animal mechanisms to build systems that could learn.
At the time, machine learning applications and research weren’t really viable career options in South Africa. Like many of my fellow students, I ended up working in the financial industry as a software engineer. I learned a lot, especially in designing huge, hearty systems that meet user requirements. But after six years, I wanted more.
At that time, deep learning was starting to take off. I first started doing online courses like Andrew Ng machine learning lectures on Coursera. Shortly after, I was fortunate enough to receive a scholarship to University College London, where I earned a master’s degree in computational statistics and machine learning.
What is your involvement in Deep Learning Indaba?
In addition to DeepMind, I am also a proud organizer and member of the steering committee Deep Learning Indabaa movement to strengthen machine learning and AI in Africa. It started in 2017 as a summer school in South Africa. We expected about 30 students to come together to learn more about machine learning – but to our surprise we received over 700 applications! It was amazing and clearly showed the need for a connection between researchers and practitioners in Africa.
Since then, the organization has grown into an annual celebration of African AI with over 600 participants and local IndabaX events in almost 30 African countries. We also have research grants, dissertation awards, and complementary programs, including a mentoring program—which I started during the pandemic to keep the community engaged.
In 2017, not a single publication by an African author working in an African institution was presented. NeurIPSleading machine learning conference. AI researchers across the African continent were working in silos – some even had colleagues working on the same topic at another institution down the road and didn’t know. Through Indaba, we built a growing community on the continent, and our graduates forged recent collaborations, publishing papers at NeurIPS and all the major conferences.
Many members have gone on to work at leading tech companies, founded recent startups on the continent, and launched other amazing grassroots AI projects in Africa. While organizing Indaba is a lot of difficult work, it is worth it when you see the accomplishments and growth of the community. I always leave our annual event inspired and ready for the future.
What brought you to DeepMind?
DeepMind was my dream company, but I didn’t think I had a chance. I’ve struggled with imposter syndrome from time to time – when you’re surrounded by intelligent, capable people, it’s simple to compare yourself on one axis and feel like a fraud. Luckily, my wonderful wife told me I had nothing to lose by applying, so I sent my CV and eventually got an offer as a research engineer!
My previous experience in software engineering really helped me prepare for this role, as I was able to rely on my engineering skills in my day-to-day work while also developing my research skills. Not getting your dream job right away doesn’t mean the door to this career is closed forever.
Which projects are you most proud of?
I recently worked on a project aimed at giving artificial agents the ability real-time cultural transmission. Cultural transmission is a social skill that humans and some animals possess that gives us the ability to learn information by observing others. It is the basis of cumulative cultural evolution and the process responsible for extending our skills, tools, and knowledge across generations.
In this project, we trained artificial agents in a simulated 3D environment to observe an expert performing a recent task, and then copy and memorize that pattern. Now that we have shown that cultural transmission is possible in artificial agents, it is possible to apply cultural evolution to aid generate artificial general intelligence (AGI).
This was my first time working on RL at scale. This work combines machine learning and social science, and I had a lot to learn on the research side. At times, progress towards our goal was also snail-paced, but we got there eventually! But really, I’m most proud of the incredibly inclusive culture we had as a design team. Even when it was tough, I knew I could count on my colleagues’ support.
Are you a member of a peer group at DeepMind?
I’ve been really involved in a number of diversity, equity, and inclusion (DE&I) initiatives. I really believe that DE&I in the workplace leads to better outcomes, and to build AI for everyone, we need to have representation from diverse voices.
I am a facilitator of an internal workshop on the concept of allyship, which is using one’s position of privilege and power to challenge the status quo in favor of people from marginalized groups. I am involved in various working groups aimed at improving the integration of the research engineering community and diversity in hiring. I am also a mentor in DeepMind Scholarship Programwhich has partnerships in Africa and other parts of the world.
What impact do you think DeepMind’s work could have?
I am particularly enthusiastic about the potential for AI to have a positive impact on medicine, especially in terms of better understanding and treating disease. For example, mental health conditions such as depression affect hundreds of millions of people worldwide, but we seem to have confined understanding of the causal mechanisms behind them, and therefore confined treatment options. I hope that in the near future, general AI systems will be able to work with human experts to unlock the secrets of our minds and aid us understand and treat these diseases.