My journey from a deep mind intern to a mentor

Former internship manager Richard Everett explains his journey to Deep Mind and shares his inspirational deep Mindder tips and advice. The 2023 internship application will open on September 16th. For more information, please visit https://dpmd.ai/internshipsatdeepmind.

What was your path to deep mind?

Like many people, I loved playing multiplayer video games that I grew up with. The interaction between human and seemingly intelligent computer-controlled players fascinated me, and I dreamed of a career in AI. This dream led me to hold a bachelor’s degree in computer science. A general (but not exclusive!) route to the industry. However, after working on several research projects with the professor, I decided to develop a taste for my research and continue towards my Ph.D.

Around the time I completed my PhD, a small startup called DeepMind was acquired by Google. When I saw their research closer, I quickly realized that I was inspiring my research, so I decided to apply for an internship in 2016. After a few interviews with engineers, researchers and program managers, I was not given the offer. However, after meeting a lot of great researchers, I decided to reapply the next year and get an internship. That experience led to a full-time offer and I have been here ever since, working on AI and supporting interns who have had the same experience.

Can you explain the interview process for your internship?

The interview process was thorough, but it evolved after applying. Today’s interns can expect the entire process to last just a few months. This includes technical and team interviews. In my application I listed researchers who were particularly interested in collaborating and were fortunate to be able to speak to them after a technical interview. I was very excited. This was a unique opportunity to talk about my past work, brainstorm potential internship projects with world-class researchers I have followed for years, and ask questions about DeepMind.

My recruiters were extremely helpful in guiding me through the process and providing resources to help prepare for the interview. I prepared for a technical interview by revisiting my first year undergraduate courses in mathematics, statistics and computer science. For example, a review of linear algebra, calculations, probability, algorithms, and data structures. I also tried to practice coding exercises and talk to you about what I was doing.

For team interviews, I reviewed the team’s recent work (papers, blog posts, articles, lectures, etc.) and thought about how my work relate to it. I also came up with a short list of questions I wanted to learn more, including the team’s collaboration style and how past internships worked.

What was it like when you joined full time?

It took me a long time to find the footing! With so many exciting projects going on and talking to amazing people, working at Deepmind often feels like a child in the world’s largest candy store. For interns, it is difficult to develop and focus one of a very large number of projects, especially within a limited time span. This was a challenge I found in my internship and today I enjoy supporting new starters who are experiencing the same excitement for the first time through this process.

Why did you take part in the internship program as a full-time employee?

After experiencing your own internship experience, you can relate to what our ambitious and current interns experience. It can all be nerve wrapping, stimulating, confusing and stimulating at the same time. After receiving so much support during my internship, I wanted to provide the same support to future interns. As a result, I am now part of several groups that are constantly looking to coordinate team internship programs and improve the program across DeepMind. You will also interview, mentor and manage interns, spend time with potential candidates and talk with potential candidates (e.g. Gracehopper, Neurip, and Research Calks).

What kind of work does an intern do?

It’s always exciting to see our interns decide to pursue with us. My team (Game Theory and Multi-Agent) worked closely with interns to co-develop projects that they could create their own, which has led to incredible projects over the years.

To highlight some public examples, interns designed new multi-agent environments (e.g. inspired by social deduction games between the US and assembly lines and inspired by assembly lines), developed infrastructure to study human agent interactions, used collaborative game theory for language models, negotiated team formation, addressed multi-agent inverse reinforcement learning, and maintained exbolic mostams to edit inverse reinforcement versions that reinforce innovative learning to reinforce game learning. learn.

How do you explain culture in Deepmind? And your team?

In short, they are kind and supportive. Over the years, we’ve heard dozens of interns and new starters make the same statement: “I can’t believe how friendly and supportive everyone is!” The time, energy and support that Deep Minders give each other is amazing, and this stretches all the way from company veterans to new starters on their first day. Everyone is happy to have coffee, chat, discuss work, share feedback, and partner together on projects.

As an example, one of my favourite projects at DeepMind (learning robust real-time cultural communication without human data) came from close collaborations with artists, designers, ethicists, program managers, QA testers, scientists, software engineers, research engineers and more. This diverse and collaborative culture has also been extended to internships, with intern projects typically including multiple collaborators and advisors (from roles, teams, and even offices) across the company. For example, some of the Game Theory and Multi-Agent Team interns work closely with Deep Minders in both London and Paris offices.

From left, a subset of the authors of the project: Ashley Edwards (RS, London), Milna Pissler (RE, Paris), Corey Mathewson (RS, Montreal), Alexander Zachel (Designer, London), Richard Everett (RS; London), and Edward Hughes (RE, London) (London).

Are there any tips for an ambitious deep mind intern?

For students interested in AI, there are plenty of easily accessible resources available to learn independently and in depth about the industry and the deep, from documents, blog posts and lectures to open source code, demos and tutorials. It’s easier to get stuck than ever! You can also participate in workshops and meetings. Many of them offer student discounts and mentorship opportunities (Deep Learning Indaba, Cooperative AI). For me, I found a love for AI research. I talked to the professor about their research, worked on projects with them, and reached out to other researchers in the field that excited me.

DeepMind is made up of kind, supportive and promoted people from every path of life, and our internship programme reflects that. Whether you are an undergraduate or doctoral student, studying subjects in technical, physical or social sciences and have AI/ML experience, you probably have an internship opportunity. We offer internships for a variety of teams in research, engineering, science, ethics, society, and operation.

After going through the process myself (twice) I can fully understand and be relevant how intimidating it is. I spoke to so many incredibly talented students. This mistakenly believes DeepMind is out of reach or their skills are inadequate and therefore they don’t even apply. If you are considering applying for an internship, my real advice to you is to do it. You have nothing to lose, and there is probably a lot to gain for both you and deepmind.