Naver will lead the global AI ethics discussion by holding various events under the theme of AI-related ethics.
Naver Clova announced on the 22nd that it participated in "ACM FAccT," the largest international conference in fairness in computer engineering, and held workshops and tutorials under the theme of hyper scale AI-related ethics with international experts in attendance.
Naver plans to get ahead in the development and application of hyper scale AI technologies through "HyperCLOVA" and secure global leadership in the field of AI ethics.
According to Naver, “ACM FAccT (ACM Conference on Fairness, Accountability, and Transparency)” conference, which marks its fifth anniversary this year, deals with interdisciplinary research on fairness, accountability, and transparency in computer science. This conference is considered the largest international conference in the field of AI ethics.
Naver explains that this year's "ACM FAccT" conference was held both face-to-face and online simultaneously at COEX in South Korea from the 21st to the 24th.
On the 21st, the first day of the conference, Naver held a workshop (CRAFT) under the title "Fairness, Accountability, and Transparency in Hyperscale Language Models." Experts in the field of AI ethics at home and abroad participated to discuss ethical problems that may arise in the process of developing hyper scale language models and applying them to services for actual users, and to explore solutions to solve these problems.
The workshop had active discussions on the current, future, and ethical issues of hyper scale AI. In the first part of the workshop, the presentation was given on legal and ethical issues and fairness and transparency related to hyper scale AI.
In the second panel discussion, experts in various fields related to AI ethics participated including KAIST Professor Cha Mi-young who served as a moderator, Seoul National University Professor Ko Hak-soo, KAIST Professor Oh Hye-yeon, and Seoul National University Professor Park Sang-chul.
On the 22nd, the second day of the conference, Jeon Sang-hyuk, leader of Naver AI Lab, will hold “Tutorial on Shortcut Learning in Machine Learning: Challenges, Analysis, Solutions” along with Song Kyung-woo, a professor at University of Seoul and Chung Yong-han Ph.D at Purdue University. AI models tend to prefer easier methods for data learning, which often causes problems. The conference will analyze the problems that arise and suggest solutions.
"This ACM FAccT workshop is a good example that shows Naver's global leadership in the field of hyper scale AI ethics," said Ha Jung-woo, director of Naver Clova AI Lab. "As AI ethics is a matter that can be tackled by creating social consensus along with technical efforts, Naver will work hard to develop 'AI for people' by giving much thought into it with experts."