Around the Table with Tina Huang

Around the Table is a three-question interview series from the Make Room email newsletter. Each edition features a conversation with a peer in the national security community to learn about their expertise and experience in the sector.

Tina Huang is a policy program manager at the Stanford Institute for Human-Centered AI. The views expressed are those of the author and do not reflect the position of Stanford University or the Stanford Institute for Human-Centered AI.

Where do you see diversity, equity, and inclusion (DEI) fitting into the AI sphere?

AI is primed to transform society and no sector will go untouched. The second, third, and fourth order effects of a new AI tool might not be obvious to one person, but painfully apparent to another. Additionally, there’s a misconception that just because AI is “math” it cannot be biased, but in reality, AI (particularly the subset of AI called machine learning) learns from the data we feed it. This data often reflects decades of past human processes, behaviors, and representations. So, if you really consider U.S. and world history, it shouldn’t be a surprise that the biased and discriminatory AI systems we have experienced thus far isn’t a reflection of the algorithm, it’s a reflection of humanity. DEI efforts in the AI sphere are crucial to ensuring AI is fair and equitable. Unfortunately, those negatively impacted by AI typically may not have a voice or aren’t involved in the development or deployment of these technologies. DEI efforts that focus on bringing individuals with disparate backgrounds (whether that’s race, religion, sexuality, etc) to the table at every stage of the AI lifecycle—and encouraging them to speak up at any point they are concerned—may mitigate future harm against already marginalized communities.

You've been involved in developing DEI policy recommendations for grad schools. What prompted you to get involved with that and what have you learned?

After experiencing, witnessing, and hearing about the microaggressions (and in some cases blatant aggression for my peers) occurring in my program, I knew something, anything, needed to be done. I understand that nobody and no institution is perfect. But I also believe that we all must measure and be measured by our willingness to do better. I worked with a group of current and former students, who all felt the same responsibility and desire to improve our program, to raise these concerns with our administration. We provided concrete, actionable recommendations to create a more diverse, equitable, and inclusive environment for students from all walks of life. Our administration was cooperative, receptive, and took steps to fulfill our recommendations. As straightforward as this may sound, it was far from it. These recommendations took a year to develop, and in that year I developed and grew relationships with new peers. My greatest lesson came from when we would share our own experiences in the program whilst learning about the adversities that someone different from ourselves often faced. It gave me a deeper understanding of how to recognize various forms of discrimination and how to be a better ally to those around me moving forward.

What resources are most useful for people just starting out in their careers?

Mentors. I cannot express how much my mentors, from school and work, helped me early on. They always gave me the extra push to apply for positions or programs after I had convinced myself I wasn’t qualified or ready. I traveled to London because my mentor encouraged me to apply for a summer research grant and connected me with the scholar I eventually worked with. I applied to and accepted my first job out of grad school that set me up for my career in tech policy because my mentor assured me that I was qualified for the position. There are so many opportunities I’ve seized, pushing me out of my comfort zone, because my mentors were always cheering me on and telling me that I can do it. There’s no better advice I can give than to find good mentors and maintain those relationships.