Agricultural Bank of China Chairman Outlines Three Main Risk Categories for AI Large Language Models on June 18

According to Gu Shu, chairman of the Agricultural Bank of China, AI large language models face three main risk categories during practical application, as he stated at the Lujiazui Forum on June 18. The first risk stems from massive parameters that reduce model interpretability—with parameters reaching trillions of units, non-linear matrix operations create opaque decision-making mechanisms that are difficult to explain. The second involves probabilistic generation compromising accuracy; instead of linear reasoning, models generate outputs based on statistical probability patterns derived from training data, easily producing self-consistent hallucinations when evidence is insufficient. The third risk emerges from autonomous reasoning and decision-making capabilities that bypass traditional software constraints, amplifying process uncertainty and outcome unpredictability.
Disclaimer: The information on this page may come from third-party sources and is for reference only. It does not represent the views or opinions of Gate and does not constitute any financial, investment, or legal advice. Virtual asset trading involves high risk. Please do not rely solely on the information on this page when making decisions. For details, see the Disclaimer.
Comment
0/400
No comments