What we learned from DeepMind and the Office for Artificial Intelligence in the UK about AI and building a more inclusive workforce on International Day of the Girl 2021. 

On October 11, 2021 Technovation hosted an event with Lila Ibrahim, the Chief Operating Officer of DeepMind, Sana Khareghani, the Head of the Office for Artificial Intelligence in the UK, and Michelle Muchilwa, a 2021 Technovation Girls Regional Winner fighting plastic pollution in Kenya’s Lake Victoria. These three inspirational speakers spent an hour discussing Artificial Intelligence, the importance of mentorship, and the future of work. Watch a full replay below, and keep reading for selected highlights from their conversation.

On getting started in AI fields 

As moderator, Michelle kicked the panel off with a question a lot of Technovation alumnae have—how do people get started working in artificial intelligence…and what is it actually like to work in AI once you’re in?

Both Sana and Lilia’s paths to careers in AI took some surprising turns. Lila spent 30 years of her career in technology, but hadn’t really considered working in AI until 2017, when DeepMind approached her. Despite her initial skepticism (“why would they want to hire me? I don’t have a background in artificial intelligence”), Lila was intrigued. As she tells it: “I spent a lot of time talking with the company and talking with the scientists and the engineers and understanding what was needed…the more I learned the more I felt a moral calling. I thought ‘how can I bring my three decades of experience of bringing new technologies to benefit humanity and do that for advanced AI?’ ”

Lila’s work is also helped by her depth and breadth of experience, even though her background is broader than just AI. “Sometimes not being the expert in artificial intelligence is where I bring the most value because I can ask questions…and sometimes those are the most important questions in the room because I can bring that diversity of perspective with the experience that I have over the decades.

Sana’s career path has also taken her to surprising places. Sana is currently the Head of the UK Government’s Office for Artificial Intelligence, where she helps put in place the foundations to help people adopt AI and data technologies. Sana didn’t start her career with an eye on AI, although she did earn her degree in computer science. After working at several startups, Sana eventually found herself doing work on facial recognition technology. “I got into AI through the back door basically and much like Lila, I very soon figured out that actually my power was in being able to translate what the developers meant to the rest of the company.” From there, Sana earned her MBA and eventually found her way to the policy side of AI work. Like Lila, Sana asks a lot of questions—“I force a lot of people to speak until everybody understands them, and often I ask so many questions that people get up and draw on a piece of paper for me. I always find that once they get up and start drawing on the whiteboard we’re on our way to discovery.

We love these answers because they showcase how many ways there are to find yourself in STEM and AI roles and they remind us that learning is a lifelong process—and most importantly, that a diverse range of experience and the ability to ask questions are two essential ingredients for success.

On why AI needs more diversity

That range of experience and the need for greater diversity became a major theme of the conversation. At several points all three of the speakers pointed out ways that the future of AI—and all of technology—depend on making sure STEM fields are more diverse. Sana explained that “the way to make technologies like AI solve real world challenges is by making sure we get the representation right…but we don’t have that right now. We don’t have the right representation of people developing AI solutions.” 

Lila also explored this same idea while discussing whether or not AI is a threat to employment. As Lila walked us through the ways that she sees AI technology as a way to create more jobs, she continually reinforced the necessity of making sure the right people understand and have a say in how technology is applied in their work and their communities. “I think this is really what’s important—there may be an underlying technology like artificial intelligence but it’s really how you can apply it locally both globally and locally to solve real problems that different communities are facing. Again this goes back to needing diversity of perspective in order to understand and solve the problem.”

AI and Ethics

And of course, the question about diversity and range of experience touches on building ethical AI too—it’s critical to consider purpose and potential impact of technologies on individuals and communities from the very outset, and when you have more people with different experiences and perspectives, you can better consider the possible effects of the tools you’re building.

Lila shared that DeepMind’s process of building ethical AI includes starting these conversations early by gathering people from different teams to talk about the potential benefits and risks of the technology. “You need to be thinking about it from very early on so that you can make sure you’re mitigating the risk and you’re partnering with [people who] may be impacted. I have to tell you that this sounds easy [but] it’s really really hard to do because we all have different expectations of what’s right or wrong. The thing that we really have to do is allow the time to ask the questions, and create a culture where it’s okay to have these conversations.” 

Sana added that from a government perspective, the objective is the same, about looking ahead for potential policy shortfalls, and that it’s absolutely not negotiable to consider these questions. “When it comes to AI technologies, safe and ethical AI, or responsible AI…there is no other bet. That is the full and total and only bet. We don’t want ‘move fast, break things and worry about it later’—this is too important and too powerful a set of technologies to not be thinking about these questions from the outset.”

How you can get involved and support girls creating big solutions with technology—including AI.

As the speakers mentioned again and again, emerging technologies (especially AI) are shaping the future. But right now too many people—especially women and people of color—don’t have a voice or influence in creating that technology. To help shift that, Technovation is focusing on including more AI resources and lessons in our curriculum, and for the 2021-2022 season, teams can choose between building a mobile app prototype or an AI-powered tool to solve community problems.

Registration for girls and mentors opens on December 8, 2021. You can learn more and sign up for an email reminder for registration.

Companies interested in engaging their employees or funding Technovation Girls can learn more and get in touch here.