How to Ensure AI Benefits Society: A Perspective from a Tech Leader
As the COO of a leading AI lab, ensuring that our technology has a positive impact on society is a top priority. Recently, at the World Economic Forum, I had the opportunity to discuss the importance of equity through technology. This got me thinking about my own journey and how we at DeepMind approach the challenge of developing technology that benefits everyone.
It all started in 2000 when I visited an orphanage in Lebanon where my father grew up. I installed computers in their first computer lab and trained the students and teachers to use them. This experience made me realize that without efforts from the technology community, many people would be left behind. It fueled my passion for building technology that can have a lasting impact on society.
When I joined DeepMind, I knew they shared the same focus on social impact. We believe in pioneering responsibly, which means being mindful of both the positive and negative impacts of AI. AI has the potential to address major challenges like climate change and healthcare, but it’s crucial to design AI systems that don’t amplify human biases, such as in hiring and policing.
To achieve this, we need to continuously question our assumptions, invite discussion and debate, and seek diverse perspectives. At DeepMind, our mission is to solve intelligence and benefit humanity, and pioneering responsibly is at the core of this mission.
One way we practice pioneering responsibly is through open conversations about responsibility within our organization. We have a multidisciplinary leadership group that advises on the risks and social impact of our research. This has helped establish a shared vision and trust among team members.
Another aspect of pioneering responsibly is embracing a kaizen philosophy, which focuses on continuous improvement. We apply this mindset to optimize our processes and systems. For example, at Coursera, we continuously improved our course structure based on learner feedback, ensuring a better learning experience for everyone.
We also recognize the importance of seeking external input and building relationships with experts in fields like security, privacy, and ethics. We engage with diverse communities impacted by our technology and listen to their needs and concerns.
Pioneering responsibly also means prioritizing initiatives focused on ethics and social impact. At DeepMind, we are committed to making AI systems more equitable and inclusive. We have published research on decolonial AI and queer fairness in AI, among others. Additionally, we strive to increase diversity in the field through scholarship programs.
In conclusion, ensuring that AI benefits society requires a responsible approach. It involves open conversations, continuous improvement, seeking external input, and prioritizing ethics and social impact. At DeepMind, we are dedicated to pioneering responsibly and making a positive difference in the world through AI.