Bangladeshi-origin Dr Rumman Chowdhury has been named in the Time 100 AI list for her contribution to the field of Artificial Intelligence.
She is a data scientist and social scientist operating in the field of applied algorithmic ethics.
“For the past six years, I’ve been creating cutting-edge socio-technical solutions for ethical, explainable and transparent AI,” she writes in her LinkedIn bio.
She was previously the director of the META (ML Ethics, Transparency, and Accountability) team at Twitter, leading a team of applied researchers and engineers to identify and mitigate algorithmic harms on the platform.
Completing her MS from Columbia University in 2006 and PhD from the University of California in 2014, she has been in this field for over a decade.
In early August, Rumman co-organised an event where some 4,000 hackers gathered in Las Vegas to break chatbots from OpenAI, Google, and Anthropic, reads Time Magazine. Hackers who were able to persuade the AIs to act in ways that broke their own rules—giving the recipe for anthrax, say, or spewing racist screeds—were rewarded with points. The winner found several vulnerabilities, including one where a chatbot revealed a credit card number it had been told to keep secret.
Read more: Bangladesh Brand Forum Awarded the Most Outstanding Creative Communications of Bangladesh
The practice takes pointers from hacker culture, where prizes for identifying security flaws by stress-testing computer programs are commonplace. The thinking is that by incentivizing many people to try as hard as possible to break chatbots and other forms of AI, developers can find and fix problems that could be much more dangerous if left to be discovered after their release to the public.
