"'AlphaGo Father' Hassabis: 'DeepSeek Has No Scientific Progress'"

"China's Best Work... But Overhyped"

Demis Hassabis, the developer of ‘AlphaGo’ and Nobel Prize laureate, as well as CEO of Google DeepMind, described China’s artificial intelligence (AI) DeepSeek as "China’s best work," but also evaluated the DeepSeek craze as exaggerated.

According to US CNBC, CEO Hassabis stated at a press conference held on the 9th (local time) in Paris, France, ahead of the ‘AI Action Summit’ that DeepSeek "can perform very excellent engineering and is expected to bring changes on a geopolitical scale."

Demis Hassabis, CEO of Google DeepMind. Photo by Reuters and Yonhap News.

However, CEO Hassabis pointed out that from a technical perspective, DeepSeek is not a major change. He said, "Despite the hype, there has been no real new scientific breakthrough," and "It is using already known technologies in AI." He added, "It is somewhat exaggerated."

While US big tech companies are pouring huge amounts of money into AI model development, DeepSeek released an AI model with performance comparable to major big tech AI at about one-tenth the development cost and lower power consumption, shocking the world. In particular, this achievement sparked controversy amid the US blocking exports of cutting-edge AI chips to China, and Nvidia’s stock price plummeted.

CEO Hassabis currently assesses the AI industry as "being on the path toward Artificial General Intelligence (AGI)." He defined AGI as "a system that demonstrates all cognitive abilities possessed by humans."

He said, "We now think we are close to (AGI)," adding, "It may take about five years for such a system to be realized, which would be truly amazing." He continued, "Society needs to prepare for it and understand what it might mean," urging, "We must ensure that the entire society benefits from it, but also mitigate some risks."

CEO Hassabis is the AI developer who created AlphaGo, which played the historic Go match against 9-dan Lee Sedol in 2016. He developed the ‘AlphaFold’ series of AI models that analyze protein structures and predict drug interactions, for which he received the Nobel Prize in Chemistry last year.