Professor Yejin Choi Challenges Monopoly with SLMs
Joanne Chang Establishes Principles to Safeguard User Creativity
The weekly news magazine TIME has announced its list of the "100 Most Influential People in AI for 2025," with two Korean-American female scientists named side by side. The honorees are Professor Yejin Choi of Stanford University and Joanne Chang, Head of Model Behavior at OpenAI.
On August 28 (local time), TIME included them in the "Thinkers" category of its "TIME100 AI 2025" list. This category is reserved for individuals who not only develop technology but also contemplate the philosophical, social, and ethical direction of AI.
Yejin Choi, Researcher of Small Language Models, Achieves Selection for Two Consecutive Years
Professor Choi has made the list for the second consecutive year. TIME explained that she has consistently raised concerns about the limitations of large language models (LLMs) and the issue of AI monopolization by a handful of companies.
As a professor at the Stanford Institute for Human-Centered Artificial Intelligence (HAI) and an AI research lead at NVIDIA, she is exploring the potential of "small language models (SLMs)" as an alternative to the high-cost, high-energy structure of large AI models. TIME evaluated her work as "presenting the possibility of alternatives to the existing inefficient model paradigm."
After majoring in computer science at Seoul National University and earning her Ph.D. from Cornell University, Professor Choi held faculty positions at the State University of New York and the University of Washington before joining Stanford earlier this year. In 2022, she was awarded the MacArthur Fellowship, establishing herself as a leading researcher in AI ethics.
Previously, in a conversation with Microsoft founder Bill Gates, she emphasized the need for transparency and decentralization, stating, "AI has become so complex that it is now a 'black box,' and power is being concentrated in certain companies."
"AI Developers Are Not Judges"... Joanne Chang's Principles
Joanne Chang, who was also selected, is responsible for designing the behavioral policies of OpenAI's ChatGPT and the image generation model DALL·E2. She studied applied mathematics and computer science at Stanford University, gained experience in product development and AI-related work at Google and Dropbox, and joined OpenAI at the end of 2021.
TIME described Chang as someone who "defines her work as helping users achieve their goals," and quoted her philosophy that "AI developers should not become judges who decide what people can do."
During the initial testing phase of ChatGPT, she observed that the AI repeatedly refused to answer even simple questions. Based on the principle of "maximizing creativity while preventing harm," she designed the framework for behavioral policies.
This year's list includes global leaders driving the AI industry, such as Elon Musk (xAI), Sam Altman (OpenAI), Jensen Huang (NVIDIA), Masayoshi Son (SoftBank), and Mark Zuckerberg (Meta). TIME classified these figures into four categories: "Leaders," "Innovators," "Shapers," and "Thinkers."
The selection of Professor Choi and Head Chang in the Thinkers category is especially meaningful, as they are recognized not only for their technical achievements but also for their perspectives on the ethics and social role of AI.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
