Market should not be dominated by only a few AIs models

Lee Hae-jin. [Photo by MK DB]

No single artificial intelligence (AI) model should dominate the market in order to for AI to be even more beneficial to humans, according to Lee Hae-jin, founder and global investment officer (GIO) of South Korea’s internet giant Naver Corp.

“Unlike search engines where users can choose information from diverse results with a single keyword, AI that provides immediate answers leaves no room for choice, which can have a significant impact, especially on children and teenagers,” Lee said, during the AI Seoul Summit on Tuesday.

He emphasized that this must be very carefully considered in terms of AI safety, quoting a passage from George Orwell’s novel “1984” – “Who controls the past controls the future. Who controls the present controls the past.” He also pointed out that if a few AI systems dominate the present, the answers they provide would be the sole source moulding the viewpoint of history and culture from the past, ultimately influencing the future as well.

“There should be many AI models that reflect diverse perspectives and understand the cultural and environmental contexts in different regions,” Lee said.

He noted that multiple viewpoints in history should be preserved, and various AI models should protect the diversity of cultures worldwide, so that children and teenagers can develop an objective historical perspective and viewpoint.

From this perspective, Lee announced that following Naver’s release of the AI Ethics Code in 2021, it plans to unveil a follow-up measure, the Naver AI Safety Framework, in June 2024.

“Naver will respect and understand the culture and values of each region and support the development of diverse and responsible models, so that many global countries can secure their own sovereign AI in any form via technology,” he said.