Turing Award Winner's Joint Article: AI Will Scale 100x in 18 Months

Recently, threeTuring Award winner Hinton, Bengio, and Yao, in a joint article, "Managing in an Era of Rapid Development AI Risk. This is a consensus paper (共识论文) that features, in addition to AI researchers, scholars in public governance, such as Xue Lan of the Institute for International Governance of Artificial Intelligence (IIGAI) at Tsinghua University, as well as Nobel Prize-winning economist Daniel Kahneman.

Hinton argues that companies are planning to scale up AI models computationally 100-fold in 18 months, and that no one knows how powerful these models will be, nor are there rules about how they can be used.

"Considering the continuous investment in AI research and development and the trend of self-evolution, the possibility of general-purpose AI surpassing humans in the next ten to twenty years cannot be ignored. If applied correctly and fairly, advanced AI could solve long-standing problems for humanity, such as disease, poverty, and the environment. At the same time, however, powerful AI brings with it enormous risks that we are far from ready to address," the paper reads. the paper reads.