In a leap for AI, researchers unveil s1-32B, a model that enhances reasoning with a mere 1,000 curated questions. Outshining OpenAI’s o1-preview on math contests, it scores 59.6% on tough science problems using a simple "budget forcing" technique. Fully open-source, it challenges industry norms with efficiency and transparency.

China’s local governments are rapidly adopting the DeepSeek AI model, transforming operations with faster document processing and audits. In Shenzhen’s Futian district, 70 AI units slash task times, but public reactions range from praise for efficiency to fears of job cuts among the nation’s 8 million civil servants.