I have always been keenly interested in the life experiences of minority communities, a curiosity likely stemming from my own personal history. As a child, I was a left-behind child for a time before joining my parents in the city where they worked. Initially, due to my inability to speak Mandarin, I became a target for ridicule by teachers and bullying by classmates. The factory school I attended in the city was somewhat marginalized, with only five students a year advancing to middle schools in the city. Upon passing the exam, I found most of my peers came from a few key primary schools, and it took years for me to integrate and adapt to the new environment. Shortly after moving to the city, I started to stutter when I talk, adding a significant psychological burden. My parents advised me to 'think clearly before speaking,' implying that stuttering was a mental issue. Most people around me believed that stuttering was 'contagious,' that children acquired it through imitation, as if it were some dreadful infectious disease. Although I now understand that stuttering is a hereditary neurological condition and not the fault or choice of the stutterer, living under its shadow has profoundly affected my social interactions and career choices. However, it has also endowed me with a greater ability to listen and empathize, especially with marginalized and voiceless groups. This experience as a minority was further amplified during my university years. As a female engineering student at Tsinghua University, I often felt a sense of discomfort and inconvenience that I later realized is a common part of minority experience.

After completing my PhD, I joined Meta as a machine learning research scientist. By chance, I became involved in accessibility technology, starting with projects related to assistive features. In 2016, I led the development of automatic alt-text, an AI image description tool that describes billions of images daily for visually impaired users. However, I eventually realized that the existing corporate structure in big companies is unsuitable to support me to do the things I truly care about: building technologies with marginalized groups so that everyone can thrive and achieve their dreams as who they are. Therefore, I decided to leave and founded AImpower.org, a non-profit organization dedicated to researching and developing technology products in collaboration with marginalized communities. Currently, we mainly work with the stuttering community, allowing members to design stutter-friendly voice and communication tools, such as psychological and communication aids for video conferences. This collective creation process also empowers community members - we don’t wait for solutions; we create our future.

我一直对少数社群的生活体验很关注，这大概来源于我的个人经历。我小时候曾当过一阵留守儿童，后来去了父母工作的城市里上学，一开始因为不会讲普通话，自然而然成为被老师嘲讽和被同学欺负的对象。我上的厂矿学校在城里也算得上边缘，一年只有五个人能上市里的中学。考上后发现大部分同学都来自几个重点小学，又花了几年时间融入和适应环境。进城上学后被发现口吃，也让我心理包袱很重。父母会让我“想清楚再讲”，暗示着口吃是一种心智上的问题。身边大部分人也相信口吃是会“传染的”，小孩口吃是“学来的”，好像口吃是一种可怕的传染病。虽然我现在明白口吃是一种遗传性神经多元症，并不是口吃者主观的过错或选择，但多年生活在口吃阴影下还是很大程度上影响我社交和职业的选择，也同时给予我了更强的倾听和共情的能力，尤其是对被边缘，被失声群体的共情。这种少数派的体验在上大学时越发被放大了，作为是清华大学的一名工科女生，时不时感受到当时还无法准确描述的不舒服与不方便。

博士毕业后我进入Meta工作，原本是单纯的算法工程师，当阴差阳错接触到无障碍技术，从辅助功能相关的项目做起。我2016年主导开发的AI图片描述每天为视障用户描述几十亿张图片。但后来我渐渐发现，现有的公司组织框架已经不足以支持我想专门为边缘用户开发更赋能、更平等的技术产品。于是我选择离职，创立了AImpower.org, 一个和边缘人群社群一起研究和开发社群需要的技术产品的非盈利组织。我们目前主要和口吃社群合作，让大家来以第一视角设计口吃友好的语音和通讯，比如视频会议的心理和沟通辅助功能。这个集体创造的过程，也是对社群成员赋能的过程——我们不再等待，未来是我们自己创造的。