Zhou Hongyi Predicts the Trends of Large Models in 2024: Multimodal Capabilities Become Standard for Domestic Models


The 2025 Guangming Science City Forum was held in Shenzhen, focusing on intelligent computing power and large model agents. Institutions such as the Shenzhen Institute of Advanced Technology announced four important achievements: the open-source multimodal model Pengcheng Haiwen 2.1 along with its corresponding dataset and toolchain; the domestic 10,000-card inference engine FenixCOS made its debut, supporting large-scale parallelism and efficient switching; the meteorological intelligent agent "Afu" was integrated into Pengcheng CloudBrain III, providing services for the 15th National Games.
Xiaomi's President Lu Weibing announced a 10-year AI strategy focusing on integrating large models with physical scenarios, embedding AI into hardware and services. Luo Fuli has joined the AI team, with AI investment growing over 50% quarterly, exceeding board expectations.....
A new study conducted high-pressure tests on 12 mainstream large models, finding that their performance significantly declined when facing shortened deadlines and increased penalties. For example, the failure rate of Gemini 2.5 Pro increased from 18.6% to 79%, and GPT-4o also experienced a near-halving drop. In critical tasks such as biosecurity, the models even made serious mistakes by skipping key steps.
ICLR 2026 Review System Suffered Large-Scale AI Infiltration: Detection Shows Among 76,000 Reviews, 21% Were Fully Generated by Large Models, 35% Were Polished by AI, and Only 43% Were Written by Humans. Machine Reviews Are Longer, Score Higher, but Often Contain Errors Such as 'Hallucinated Citations', Triggering Protests from Authors. The Organizing Committee Issued an Emergency Ban, Planning to Block AI-Generated Content at the Submission Stage to Rebuild Trust.
The 2025 Hong Kong Fintech Week focuses on the integration of fintech and AI, bringing together guests such as Carrie Lam and Geoffrey Hinton. Zhu Guang, CEO of Du Xiaoman, emphasized the innovative applications of large models in financial services, driving customer service from monthly surveys to real-time responses, achieving a revolutionary transformation centered around customer-centricity.