中国科学院上海先进研究院开发的ScienceOne-Yuheng碳核算大模型，将中国2022年温室气体排放量比联合国模型低17.7%，同时将美国的排放总量上调15.2%。该模型考虑生产和消费两方面，减轻了包括中国在内的出口国的责任。
中国科学院上海先进研究院（CAS）开发的ScienceOne-Yuheng碳核算大模型，对2022年全球碳排放进行了重新计算。该模型显示，中国温室气体排放量比联合国常用模型低17.7%，而美国总量则上调15.2%。
CAS于4月8日表示：“全景碳核算将成为推动全球气候治理向更高精度和共识迈进的核心驱动力。”该模型旨在建立“更公平、更具科学依据的全球碳核算和责任分担框架”。
研究院副院长兼研究团队首席科学家魏巍表示，该模型“代表了我们理解和管理全球碳排放方式的范式转变”。
这一方法强调消费端排放，旨在为出口国如中国提供更全面的碳责任评估。
由 AI 报道 AI 生成的图像 事实核查
A new scenario framework for CMIP7—the standardized set of emissions pathways used by many climate modelers and referenced in IPCC assessments—concludes that CMIP6’s highest-emissions pathway, SSP5-8.5 (and its earlier counterpart, RCP8.5), has become “implausible” given recent energy-cost trends, climate policy developments and emissions patterns.
Chinese authorities have issued new emission rules targeting strategic sectors including AI data centres to meet 2030 climate goals. The move comes amid the Iran war heightening energy security needs. The rules also call for greener digital infrastructure.
由 AI 报道
A US Congressional commission concludes that China’s open ecosystem has narrowed performance gaps with top Western large language models. The report highlights the compounding force of open-source models and manufacturing dominance.
Alibaba chairman Joe Tsai has credited China’s advantages in artificial intelligence to investments in its power grid and open-source models. He stated that such massive investments have provided ample supply and low costs for the energy-intensive AI sector. China’s industrial depth highlights vast potential for AI applications, Tsai said.
由 AI 报道
A district in Guangzhou has launched a platform to connect businesses with artificial intelligence providers. Initial projects are valued at more than 13 million yuan. Officials say the move aims to accelerate the commercial use of AI.